CNC-2116 - Up and running, sort of (2015-03-21)

[Yvan256]

After seeing crashwg's CNC machine and especially the results he got when he milled that 4" wooden token, I've decided to build my own CNC machine too. Seeing dexxy's MechMate cutting panels for his Neon Copy was also inspiring.

For the last five years or so,  I've collected easy-to-do ideas about linear slides, motor couplings, anti-backlash nuts, etc. Just like an arcade cabinet, there's a lot of topics to cover and different ideas depending on the available budget and tools.

My problem is that I have many projects on hold because I can easily design things in Sketchup but I have a hard time actually building them. It's easy to make a Neo-Geo side panel in Sketchup, it's another thing to be able to cut it properly with all those angles and curves. Having a small CNC would make the job of cutting the required parts so much easier. I'm not planning on doing RCBs (routed circuit boards), so even a so-called low-precision machine would be just fine with me.

So, I'm trying to build a really simple, very extremely low-cost 3-axis CNC milling machine: CNC256.

• Cutting area of 24 x 20 x 2 inches (60.96 x 50.8 x 5.08 cm - I know Canada is supposed to be metric, but being so close to the USA, almost everything from wood to bolts is almost always in feet and inches). I could have gone for 24 x 24, but the machine is already going to be wide enough as it is. Big enough for desktop arcade cabinets and big enough for the panels of a future project, which at last count needed over 50 precisely-cut panels.
• I've bought four stepper motors (1.8 degrees, 24v) at an electronics surplus store for around 8$CAD each, I think. It's been too long, I can't even remember which year I bought them. (update: I contacted the store, they were only 4$ each) Why four? At that price, and given that it's a surplus store, I wanted to have at least one backup motor.
• I'm making my own stepper controllers to reduce the cost. So far I've got (good) preliminary results with an Atmel ATtiny85 and four TIP120 for my first tests. I plan to use one ATtiny84 (the ATtiny85 doesn't have enough pins to control the motor and receive signals) and one ULN2803 (two lines wired in parallel for each motor winding to allow 1A per chip) for a total cost of around 4$CAD per axis (plus taxes and shipping, required pcboard that I'll make myself, etc). It won't have extra-smooth motion or power-saving features, but at around 15$CAD to control all three motors this is a hard deal to beat. The cheapest controller for 3 motors seem to be around 35$ on eBay which is more than twice the cost of my controllers. And by building it myself, and with a microcontroller, I control and know everything from A to Z, and I'll be able to update the software later if I want to (microstepping, etc).
• For the "CNC computer" itself, the thing that drives the motor controllers, I plan on using Grbl instead of an actual x86 computer. With the external clock, power supply, etc it's going to cost about 10$ for the "computer". I'll still need a way to transfer g-code files to it... probably via an SD card, a small LCD display, small joystick and buttons (I plan on using arcade parts for that). The goal is to have an autonomous machine that doesn't require an external computer connected to it.
• Inverted V-groove skate bearings linear slides (idea by larry104) - I dropped that idea because the nuts grind on the aluminium angle and it's near impossible to adjust them properly.
• Fake V-groove skate bearings on aluminium angle, idea taken from Ro-Bot-X's Weblog, who modified an idea from TinHead, who modified the idea from the blackToe... The plan is to go with the Route-X design, change the aluminium angle when it wears out (if it wears out enough to need replacing that is, after all I'm not planning on using the machine more than a few hours per week) and upgrade to real v-groove bearings later and possibly steel angle much later on. That should allow easy upgrades, one at a time to lower costs, and without needing to re-cut any panel.
• For the motion, I'm going with regular threaded rod (low-cost) with home-made HDPE nuts. My father has donated enough HDPE blocks to make 12 nuts - three for the machine, nine backups!)
• I planned on using the Dremel tool I already have for routing, but regular 1/4" routing bit won't fit and according to most people a Dremel is just too under-powered for such a job. Seeing the videos on YouTube vs proper trimmers/routers, I have to agree.
• I'm recycling a lot of particle board panels from older/abandoned/failed projects. So far I had enough to make the whole Y and Z axis.
• I also plan on adding a mini Shop-Vac with a home-made cyclone dust separator similar to that one. My father has both a full-size and mini shop-vac so I'll be able to compare the two and see if the mini is enough for the job.

Bill of materials (prices in Canadian dollars with taxes):

• 32 skate bearings (ABEC5): 30$
• Four stepper motors (NEMA 23, 24v, 1.8 degree per step): 16$ (three for the machine, one backup)
• Two 8 feet long, 1/2 inch wide, 1/8 inch thick aluminium angles: 36$
• Particle boards: 0$ (already had those, recycled from older projects)
• Three Motor controllers: 13$ (not bought yet)
• Threaded rod: 10$ (not bought yet)
• Nylon nuts: 0$ (home-made, HDPE plastic blocks donated by my father, not done yet)
• Various bolts, nuts, washers: around 10$
• Power supply for the microcontrollers and motors: unknown (not bought yet)
• Computer: unknown (not bought yet - I know the ATmega328 costs 4$, but the other parts are a bit unknown right now)
• Shop-Vac mini: 40$ when on sale (not bought yet)
• Cyclone system of the shop-vac: 2$ (plastic pipe connectors, one angled and one straight. I plan on finding the plastic bucket and tubing for free)
• Total so far/planned: around 166$ (rounded up)

Please note that since some parts came from surplus stores or some parts are different, I used average prices for those parts. I'm also going to need more wood panels to complete the whole thing.

I still need to get a small trim router, but I'm setting the budget for that at 100$CAD maximum. I'll be checking for units on sale.

Attached is a cutting size capability comparison between standard size/popular kits sold on the internet (the two models of blueChick, the two models of blackToe and the Mechmate).

[drventure]

Looking forward to this thread. I've wanted to do the same thing, but don't really have space.

Maybe after you're build, I'll find space 

[Yvan256]

Some new parts arrived today. I already had 16 bearings, but due to the design I'll need more. Local stores only sell extremely high-end skate bearings, so my brother brought me this pack of 16 bearings when he visited. Adds 14.80$CAD to the build cost, or 0.93$CAD per bearing. Still cheap to replace when they fail.

Thanks Daniel!

[syph007]

Glad to see you start this thread.  Im in Canada too and have been wondering how hard sourcing parts will be.   Can I ask where you found your steppers?

[Yvan256]

Glad to see you start this thread.  Im in Canada too and have been wondering how hard sourcing parts will be.   Can I ask where you found your steppers?

I got my steppers at Accès Électronique. They have four stores, all in the Quebec province. Never mind their website, their stores usually have a much bigger selection. Since it's a surplus store, the selection varies all the time.

The inventory number for my motors is 51925. I added two photos to my first post. For an idea of scale, the square part of the motor mount is 56mm wide/tall.

I emailed them to ask if they had any left, and the price. They don't have this motor for sale anymore which isn't surprising since they're a surplus store (am I glad I bought four), but the price was only 4$ each. I'm sure most of their stores currently have similar motors at around the same price.

If I had known how good they were, I would have bought a dozen.

Tip: try to get 12v motors, you'll have less trouble finding a power supply for your motors... i.e.  power supply from old PCs.

[Yvan256]

Looking forward to this thread. I've wanted to do the same thing, but don't really have space.

Maybe after you're build, I'll find space  

A small CNC that can cut 10x10 inches panels wouldn't be much bigger than a big desktop laser printer.

[Yvan256]

Here's a small preview of the design, the X and Z axis are pretty much done and I only need to finalize some details about the Y axis. The goal is to keep it simple to cut and build to minimize the errors and alignment problems.

Yes, the bearings will be grinding the aluminium angles, but since I will only be using the machine a few hours every month at most, it shouldn't become a problem for a relatively long time. This design also allows easy replacement of the dual skate bearings with v-groove bearings without having to re-cut any panel.

[nox771]

Glad to see you start this thread.  Im in Canada too and have been wondering how hard sourcing parts will be.   Can I ask where you found your steppers?

If local sources don't work for you there are a couple places I know of:

http://www.kelinginc.net/index.html  - I got some heavy duty steppers from this place (I'm building an aluminum extrusion CNC).  They have the drive electronics also.  Good place for heavy duty steppers - not cheap, but durable.

http://www.allelectronics.com/make-a-store/category/400/Motors/1.html  - I ran across this place recently, they are always sending me mail (I'm in the US btw), but I never really checked them out.  Finally one day I did - out of their whole site the motor section looked most interesting to me.  Turns out they have a lot of small motors, and quite inexpensive also.  Some look pretty good too, like these:
http://www.allelectronics.com/make-a-store/item/SMT-119/4-LEAD-1.8-DEG-STEPPER-MOTOR/1.html
http://www.allelectronics.com/make-a-store/item/SMT-363/4-WIRE-STEPPER-MOTOR/1.html
I haven't bought from these guys yet, but if I were to make a tabletop sized CNC I might try them out.

[Yvan256]

Some look pretty good too, like these:
http://www.allelectronics.com/make-a-store/item/SMT-119/4-LEAD-1.8-DEG-STEPPER-MOTOR/1.html
http://www.allelectronics.com/make-a-store/item/SMT-363/4-WIRE-STEPPER-MOTOR/1.html

Just to simplify your life, unless you're buying the stepper drivers, get motors with 5 or 6 wires. The 4-wires motors need a completely different driving method. And as I mentioned above, try to get 5v or 12v motors so you can easily find power supplies for them.

[Yvan256]

Trying to optimize the panels on a standard sheet of 4x8' while also trying to minimize the number of cuts required. Panels are colored for easier comprehension. Also shown is the aluminium angles and 3' threaded rod.

[crashwg]

Looking good.  My only concern would be the wear on the aluminum angles but you've already acknowledged that will be "wear" parts and need to be replaced.  Also, I didn't see any mention of bearings for the lead screws.

[Yvan256]

Looking good.  My only concern would be the wear on the aluminum angles but you've already acknowledged that will be "wear" parts and need to be replaced.  Also, I didn't see any mention of bearings for the lead screws.

The bearing setup shown above is only because it makes the whole thing a lot easier to plan and assemble. However, I'm currently testing the same bearings+angle method used on your machine and a lot of other CNCs. It's an extremely popular method but I'm searching for the easiest, most cost-effective way of doing it. My aluminium angle is only 1/2" so I would need to buy yet another 8' length only for the bearing supports. It's not exactly cheap and I would only use a small fraction of it. Using square tubing would make assembly easier, but I also don't have that.

I do have six smaller bearings that fit my 1/4 threaded rods. I can't remember where I got them, but they were in my box of parts that I've accumulated for the last five years with the goal of building this CNC.

What trimmer or router are you using on your machine?

[crashwg]

What trimmer or router are you using on your machine?

THIS ONE  It's a little bigger than most people use on these home made CNC machines but another builder of the "book machine" did it and was happy with the results so I bought one too.  I upgraded the collet system on the router though to these precision ones. Which allowed me to hold smaller bits which was essential for the mini coin doors.  THIS is where I purchased my carbide end mills by the way.  Prices are amazing, product is awesome.  Only downfall is the $75 minimum order. Unfortunately or fortunately depending on how you look at it, it's pretty easy to meet the minimum with carbide even at their low prices.

[Yvan256]

This one. It's a little bigger than most people use on these home made CNC machines but another builder of the "book machine" did it and was happy with the results so I bought one too.

Hitachi M12VC, 3.3 kilograms... it's quite heavy, I hope your machine isn't flexing in the middle?  

From what I've read, most people seem to be happy with the Bosch Colt, which only weight 1.5 kilograms. Then again, from what I've read, the M12VC is also the quietest router in that price range.

[Yvan256]

I just got a Hitachi M12VC from my parents! The cost of this project just got a lot lower!  

Thank you both!

This thing is really quiet at 8000 RPMs! 

[mountain]

Sweet! That is the router I use on my CNC. Very quiet, cheap, and easy to refurb.

[Yvan256]

Sweet! That is the router I use on my CNC. Very quiet, cheap, and easy to refurb.

Not sure how much is costs in the USA but here in Canada it sells for 200$CAD, it was on sale for 140$CAD just before Christmas. Not what I'd call "cheap", however I don't know how much similar routers from other companies cost.


Also, I'll have to change the name of my machine since "CNC256" returns too many results. Maybe "CNC5161", which is the cutting envelope converted to centimeters. Currently at only 10 results on Google.

[Yvan256]

Looking forward to this thread. I've wanted to do the same thing, but don't really have space.

Maybe after you're build, I'll find space 

Would you have enough space for a small desktop CNC?

I'm working on my own design with a 6x4x3 envelope and removable tools. So far I'm planning on those four tools: multicolor pen plotter, mill for PCBs and MDF, hot wire foam cutter, hot glue 3D printer. And probably ABS/PLA 3d printer tool later, unless I build a Tantillus first.

[Yvan256]

Okay, so I backtracked a bit on what I planned to do for a CNC.

First of all, the Hitachi M12VC is a very nice router, however it's also very heavy at least compared to a small Dremel. That means the whole machine has to be stronger to hold it, not bend in the middle, etc. I also don't have the room for the size of CNC machine I initially planned to make.

So I did a few tests on MDF and particle boards with my Dremel 395 using a small 1/16" carbide bit made to cut PCB panels and the results are very good, although you need to cut slowly and not very deeply for good results. I'm not making a commercial production machine here, I only want something to help me cut wood and plastic for my various personal projects. Dremel also makes 1/8", 3/16" and 1/4" routing bits (Dremel #650, #652 and #654).

So I went looking for smaller CNC machines such as the Mantis I mentioned above, with the idea of using my Dremel on it. But then I kept making the design bigger to be able to cut some parts for MVS-99-6 and other small projects, then a bit bigger for some other parts, etc. In the end I had something as big as my laser printer. Enough is enough!

So, with that in mind, I went back to the design of my CNC machine. Small enough but still capable of doing what I wanted to do in the first place!

The goals for the "new" machine are:
- make it with the parts and materials I already have
- keep the design simple so it's easy to make and adjust
- use skate bearings to keep costs down but in a configuration* that makes it possible to upgrade to real V-groove bearings at a later date
- use cheaper step motors** in order to save the NEMA23 motors for a possible future bigger, stronger and faster CNC machine which will use the M12VC.
- small enough to be carried by a single person
- small enough to fit on a desk or table
- small enough to fit in the back seat of a car
- and last but not least, I want to be able to cut bigger panels*** in multiple passes

* I will use the same skate bearing configuration as TinHead but with a single washer in the middle and ridding them on the edge of the 1/8" aluminium angle. It works surprisingly well and from what I've read the aluminium will deform with use but only up to a point. Again, hobby CNC usage, not commercial. I don't think I'll need to replace the aluminium angles any time soon.

** Yes I know those motors are only 24 steps, but with 1/4-20 threaded rods 5/16-18 threaded rods the resolution will not be a problem. That makes 480 432 steps per inch, or a theoretical 0.002083 0.002315 inch per step. For reference, 1/32" is 0.03125". I changed from 1/4 to 5/16 threaded rods because then I can use a regular skate bearings to hold the rods in place. And 18 threads per inch instead of 20 also means it's going to be about 10% faster.

*** Now, the "able to cut bigger panels in multiple passes" idea requires some explanation. By "bigger" I mean panels bigger than what the machine can cut in a single pass. If a picture is worth a thousand words, how much is an animation worth? See the first attached image (click on it to see the animation). I cut the big panel into three equal parts to make it easier to see what's happening between each frame. Bender will probably remember why I want to be able to cut bigger panels. No, that project is not dead. ;-)


As an added bonus, those indexing pins will allow me to cut both sides of a panel (cut, flip over, cut), so having pockets or angles on both sides of a part won't be a problem.

The first frame shows the machine by itself, with its cutting area of 18 inches wide (left to right) by 10 inches deep (front to back).
The second frame shows the machine with its back and front extensions and a 18x30 inches board.
The third and fourth frames show how I will move the board between each step, held in position by four indexing pins in the table of the machine.

So to cut a "3x panel", I put it in position #1, let the machine cut the first third of the panel, move the board into position #2, let the machine cut the second third of the panel and then move the board into position #3 and let the machine cut the last third of the panel.

The back and front extensions will have holes to bolt them on the CNC machine when in use and hinges so I can fold them flat for transport or storage.

The second image is the machine with five parts from my MVS-99-6 project, just to test things out with the cutting area of a single panel.

[Yvan256]

After the usual Sketchup plans, here's the next step for most projects around here: a pile of cut panels.

Low-quality photo courtesy of my Nintendo DSi.

[Yvan256]

This project is progressing slowly, but at least it's moving forward. Sorry about the non-metric measurements, but almost all the building materials here in Canada use the imperial system so it's easier to just use imperial everywhere. And even when metric hardware is available, it's usually nearly twice as expensive as imperial. With that said, let's get back to the topic.

I started cutting the router mount for the Hitachi M12VC yesterday. In the first photo, you can see that I cut 5x5" squares out of 5/8" MDF. Then, I drew lines to find the center. I used two methods at once to be certain I was in the middle. First I drew four lines, using a block a bit shorter than 2.5". That made parallel lines with a little square right in the middle of my 5x5" part. Then I drew the usual "X" lines from the corners that cross in the middle. Since the "X" happened to be right in the middle of the little square, I was 100% sure that I was centered. Or at least that I was completely off-center by exactly the same amount, twice, on two parts. 

Then I carefully punched the middle with a very small and sharp nail and drilled a progressively bigger hole, in 1/16" increments, to make sure it was always centered. Once I got to 1/4", I switched to a 7/8" hole saw bit to make big round holes. This bit makes holes that are exactly the same diameter as a roller skate bearing (as shown in one of my first posts of this project), it's a really tight fit.

In the second photo, you can see how I used a 1/4" router bit on a drill press with a jig with two skate bearings and rotated the 5x5" plate many, many times, probably around 1/32" depth on each pass, to drill a 3.27" diameter hole. A drill press doesn't turn nearly as fast as a router and has a much bigger play at the bit, but at least it's easy to increment the cutting depth.

Why go to such lengths and not use a circle cutting add-on with the router I already have? Because the only one I had access to couldn't make a hole that small, not to mention that the M12VC is a fixed base plate router and incrementing the cutting depth at each pass would have been troublesome. And the smallest hole I can make with my Dremel router attachment is around 3.5".

The third photo is the router inside the two mounting plates. There's still work to be done on those plates, but at least the router fits snugly inside them.

[rbarr110]

Good luck with the build, they are very fun machines to play with.  It took me about 2 years of slowing collecting parts and building to get mine done.  Mine has been FULLY functional for close to a year now and am just starting to really put it to work.  Once I started pricing putting one together, there wasnt much difference from a smaller machine to the larger...so I went larger, a 4'x5' cutting area.
A quick video of my first movement.


Looking at your pricing list, where did you find cheap $13 motor controllers? Most of the controllers I find are anywhere from about $40 - $130 depending on what you need.   I personally went with the Gecko G540 for the built in driver board and 4 motor controllers...and amcontrolling the Nema 23's.

[Yvan256]

Good luck with the build, they are very fun machines to play with.  It took me about 2 years of slowing collecting parts and building to get mine done.  Mine has been FULLY functional for close to a year now and am just starting to really put it to work.

I've been reading and collecting parts for the last 5 years, so I know what you mean. From what I've read, I'm guessing the learning curve of all the necessary software is the hardest part. I've already got half a dozen projects planned for my machine, can't wait to have some fun too! 

Once I started pricing putting one together, there wasnt much difference from a smaller machine to the larger...so I went larger, a 4'x5' cutting area.

I wish I had the budget and the room for such a huge machine. Unfortunately I need something small enough to be mobile, to carry around from my apartment to my father's shop or to a friend's house.

Looking at your pricing list, where did you find cheap $13 motor controllers? Most of the controllers I find are anywhere from about $40 - $130 depending on what you need.   I personally went with the Gecko G540 for the built in driver board and 4 motor controllers...and amcontrolling the Nema 23's.

I'm building my own controllers from ATtiny microcontrollers and IRLZ14 logic level mosfets. So far it works fine with the can-style, 12V stepper motors I got for 4$CAD each.

[selfie]

Nice work! The pass through cutting of the larger sheets is a great idea, once you get your head around the programming it's really easy.

I'm in the market for a new CNC router too but mine gets used for paid work 8 hours per day so my budget is a bit higher 

Do you have plans for software?

To run codes I use Mach3 http://www.machsupport.com/
and Vcarve to program codes  http://www.vectric.com/

A tip when you get around to you HDPE nuts. The link says to cool the HDPE in water, this gains you nothing over letting it cool naturally but increases the chance of creating backlash in the nuts with the unnecessary movement.  I would secure the threaded rod in place and heat the rod as described but instead of quenching in water I would cool the rod in the same method as heating.

[Yvan256]

Nice work! The pass through cutting of the larger sheets is a great idea, once you get your head around the programming it's really easy.

I'm in the market for a new CNC router too but mine gets used for paid work 8 hours per day so my budget is a bit higher 

I've read good things about spindles, especially water-cooled ones. Not sure if it's in your budget range, but it's really worth looking into from all the comments I've read about these.

Do you have plans for software?

To run codes I use Mach3 http://www.machsupport.com/
and Vcarve to program codes  http://www.vectric.com/

I have zero budget as far as software goes. I was planning on designing things in Sketchup, then export/convert to another format used by LinuxCNC. Going with Sketchup just because I'm used to it and it's easy to make parts, group them, move them around in a 3D project, etc. I'm talking mainly about arcade parts and arcade panels here, however, not 3D signs with text.

A tip when you get around to you HDPE nuts. The link says to cool the HDPE in water, this gains you nothing over letting it cool naturally but increases the chance of creating backlash in the nuts with the unnecessary movement.  I would secure the threaded rod in place and heat the rod as described but instead of quenching in water I would cool the rod in the same method as heating.

I was planning on using two blocks with holes to hold the threaded rod level and to a precise height, and use two clamps, one on each side of the HDPE halves sandwich. All that would be left to do is heat the threaded rod, the rest would be "automated", so to speak. Less risk of movement. The stuff I have is much tougher than HDPE though, I think it's Delrin or something. And while I'm at it, I'll be making more than three blocks, to have ready-to-use spares.

[selfie]

Delrin will last MUCH longer than HDPE and won't be any harder to form.

A water cooled spindle would be nice as they are a lot quieter than the air cooled ones but they are too hard too get in my power range ~9HP

[Yvan256]

Another small step done today. I drilled more holes to attach the back and front panels to the bottom plate of the base of the machine. I didn't have enough bolts for all the holes, however it's quite sturdy even with four bolts and the top panel missing. The bolts are not countersunk because this is the underside of the machine.

I can't wait to complete the base of this thing, I'd like to have at least one axis completed before the end of the month!

The six panels, seen on top of the reversed base, will go underneath the top panel of the base to reinforce it from back to front. There is 10 panels in the base of the machine and a total of around about 18~20 panels for the whole machine. So once the base is complete, the machine will be about 50% complete. Or only 33% complete if we go with the three axis needed, but from what I have read the base is usually the longest part to build.

And yes, that is a skate bearing you see in the front. Two bearings and four nuts will hold a 5/16"-18TPI threaded rod in place.

Once again, the low-quality photo is courtesy of my Nintendo DSi.

[Yvan256]

A shop vac is a very crucial part of a CNC machine, because of all the dust made while cutting.

So, imagine my surprise when last wednesday (my birthday), I got this little thing. Thanks Daniel!

This Shop Vac may be small, but boy is it loud! But it's also very strong, just like a regular-sized one!

Now all I have to do is built a little dust cyclone to go with it!

The name seems vaguely familiar for some reason... What's even funnier is that the colors are the same. 

edit: it doesn't matter if the capacity of the Shop-Vac itself is small, because the dust will be collected by the... eh... dust collector. 

[Nephasth]

I got that same mini vac for a wedding present. It's great. I can hook it up to my router table and really keep the dust down, love it for sucking up debris when doing engine work as well.

[Yvan256]

Does anyone know where to buy a Black and Decker RTX in Canada for a reasonable price? I really don't think my machine will be able to support the weight of the Hitachi M12VC and the Dremel is just a bit underpowered. I mean, it would work but I'll only be able to cut about 1/32" deep at a time. To cut a 5/8" thick panel would take 20 passes. It's still better than cutting by hand for complex shapes, but that's quite slow.

The RTX has almost twice the amps of the Dremel so I guess it would be a good compromise.

The problem is, the RTX is listed at 30~35$USD on USA websites (Target, Wal-Mart, Amazon) but I've only seen it on Amazon.ca for 52~70$. 

[Yvan256]

Another small update. I had problems with the particle boards, so I've switched to MDF. I'm also unsure about the whole "make bigger panels with indexing pins", so I've changed the size of the machine back to one that would allow me to cut the panels for my MVS-99-6 desktop arcade machine.

Here's the progress on the Z axis. It's a variation on TinHead's design from his Valkyrie CNC, made to reduce the width of the whole carriage. By using a "T" configuration for the front plate/rails assembly, instead of putting the rails on the side of the front plate, I was able to keep the whole thing at only 3.5 inches wide.

It's missing four mounting holes on the front plate, a top plate with the motor and the threaded rod and nut, but this is more or less complete for the Z axis. It's also the X axis carriage. Some of the holes for the bolts+bearings assemblies are oblong, to allow for precise alignement with the rails.

Included is three photos, showing the two parts alone and assembled.

[Bengaz]

You lot should work for NASA!!!

[menace]

You (and crashwg) have inspired me into my next project  .  I like how your linear bearings look compared to the standard DIY variety.  Did you regain much work space doing that versus the >< arrangement?  I'm trying to make the machine size small but maximize the work space so anything that gives me more travel is desirable.  I understand there is about a 4" loss in the x and y due to how the frame and router line up and my ultimate goal is to crank out some bartops.  So a table of 24" x 48" works out to a working space of 20" x 40"  that seem about right?

With regards to you stepper drivers--was this a circuit you just knew how to build or were you following instructions from someplace.  It seems to me the electronics are the most expensive part of the build and I'm hoping to trim a few bucks from the budget.

Failing building my own, any thoughts on this combo:  http://www.ebay.ca/itm/200806516142?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649

Keep up the good work!

[Yvan256]

You (and crashwg) have inspired me into my next project  .

First off, I'm glad to see yet another arcade builder embark on the CNC quest! If anything, we'll shortly be able to see more arcade-related CNC cutting videos on YouTube, I'm tired of watching 3D plates and router mounts videos. 

I like how your linear bearings look compared to the standard DIY variety.  Did you regain much work space doing that versus the >< arrangement?

The space gained is probably not that much, all things considered. Also, after playing with the Z-axis manually, I have to agree with tinhead about the aluminium angle being too soft for this configuration. His latest machine still uses a weird setup but it seems to be much better, at least according to him.

I still had to try using the edge of aluminium angle just to see how bad it was, because it makes the whole thing a lot easier to build. It might end up better with real v-groove bearings. However, with my fake v-groove bearings, this setup it didn't last very long because after moving it a few dozen times manually it already had a small play in the setup. I've also read that once the aluminium edges are compressed that it's fine after that, but you do need to "break in" the edges" with a program to run your machine on all three axis from zero to max, to wear/compress them evenly on the whole surface. For that reason, I went back (again) to the tried-and-tested setup used by people all around the world. I only have two six inches bearing assemblies made so far, but it runs a lot smoother than the edge rolling setup. And in that configuration, the wearing of the aluminium angle will be minimal since the bearing rolls parallel to the surface of the angle.

I'm trying to make the machine size small but maximize the work space so anything that gives me more travel is desirable.  I understand there is about a 4" loss in the x and y due to how the frame and router line up and my ultimate goal is to crank out some bartops.  So a table of 24" x 48" works out to a working space of 20" x 40"  that seem about right?

A 24x48" table with a working space of around 20x40" sounds about right to me if your Z axis assembly is around 4" wide and your X axis assembly is about 8" deep. However, those are rough estimates, I'd need to see your plans to make sure.

You will lose a lot in the Y axis because of the X axis assembly. But you can't make this assembly too narrow because you'll lose structural strength. As for losing working space in the X axis, you can work around that a bit by making your Y rail go "outside the box". See how the Y rail of the blackFoot is assembled (first photo)?

The best thing would be to make the plans in Sketchup or similar, down to the last detail. That way you can really squeeze out every last 1/16" possible and know in advance what's needed, the dimensions of each part, etc.

With regards to you stepper drivers--was this a circuit you just knew how to build or were you following instructions from someplace.  It seems to me the electronics are the most expensive part of the build and I'm hoping to trim a few bucks from the budget.

It's more or less an already known circuit to which I'm going to try and add an Atmel microcontroller to lower the cost instead of using a pre-built controller chip. The IRLZ14 transistors will do the heavy lifting and the microcontroller will do the logic between the commands of the computer and the transistors. However, it's going to be crude at best with no power control or microstepping, so it's more of a temporary solution. And there is now very cheap TB6560 controllers on eBay, so I'm probably not even going to bother making my own at all. Since surplus motors are 24v, I'll have to try and get three used 24v laptop power supplies from the local recycling center. I hope I'll be lucky enough to find three power supplies of the same model.

Failing building my own, any thoughts on this combo:  http://www.ebay.ca/itm/200806516142?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649

Actually, for all the included stuff, it's a pretty good price. Those motors will probably be more than enough, the controller is the one I have on my watch list and even the power supply and parallel cable are included. Frankly, if I had the money I'd be tempted to order one of those kits myself!

What do you plan to use for the milling tool? Dremel 4000? Bosch Colt? Dewalt 611? Hitachi M12VC? The weight and power of your tool will decide of a lot of factors for your machine, i.e. the structural strength needed, etc. I'm still new at this, but my one tip would be to never underestimate structural flex caused by weight.

[Yvan256]

As you can read from my reply above, I've decided to switch back to the usual bearing assembly used by hundreds if not thousands of people worldwide. In building the Z-axis to test things out, I found out that the aluminium angle edges wear out too fast and too unevenly. Real v-groove bearings and breaking in the machine would probably work but I don't have nearly 150$CAD for bearings alone.

So, yesterday I built the first two bearing assemblies, six inches long each. I did have access to my father's drill press, so the holes for the bolts are at a perfect 90 degrees. The holes were tapped manually but with a cut bolt inside the drill press, so the threads are at a perfect 90 degrees too.

These assemblies roll extremely smoothly on the sides of the aluminium angle. I would also not be afraid to put the weight of the M12VC on these, as opposed to my previous setup.

Sorry about the uneven lighting in the animation, the DSi isn't exactly a high-end camera. 

note: the aluminium angles do not touch, it's an optical illusion because of the angle of the camera. There's about 1/8" space between the "tops" of the angles.

[Ryglore]

Oh man. I must catch up on this thread later. So much goodness to read since I've come back!

[menace]

Thanks for the feedback!  I started cutting mdf this weekend and managed to scrounge enough old roller blades from kijiji to come up with 30 5/16" bearings--total cost was about $10 for those   I'm going to pull the trigger on the electronics and motor bundle as I have yet to find a place to get everything together locally.  My next stumbling block is the cross dowels--they are 1.39 ea. at home deport and I can't believe thats the best price for these--The build your own cnc guy sells them for .50 so I'm sure they can be sourced somewhere.

As far as the router goes, I plan on using my rotozip with its handles removed.  It's 23,000 rpm and screams like a banshee but I think I can dial that back a little.   It cuts mdf so fast that if you don't keep it moving it'll start to burn

 I should really start a build thread and stop hi jacking yours 

[Yvan256]

Thanks for the feedback!  I started cutting mdf this weekend and managed to scrounge enough old roller blades from kijiji to come up with 30 5/16" bearings--total cost was about $10 for those   I'm going to pull the trigger on the electronics and motor bundle as I have yet to find a place to get everything together locally.  My next stumbling block is the cross dowels--they are 1.39 ea. at home deport and I can't believe thats the best price for these--The build your own cnc guy sells them for .50 so I'm sure they can be sourced somewhere.

As far as the router goes, I plan on using my rotozip with its handles removed.  It's 23,000 rpm and screams like a banshee but I think I can dial that back a little.   It cuts mdf so fast that if you don't keep it moving it'll start to burn

 I should really start a build thread and stop hi jacking yours 

Good idea for the bearings, I bought mine new, didn't think about getting old roller blades at flea markets. I agree about the electronics+motors+power supply kit, if you can afford it.

As for the cross-dowels, I haven't found any place that even know what these are, let alone a place to buy them. So I went with the "cheapo" idea instead.

I'm guessing you'll need to be careful with the cutting speeds if the rotozip can burn the material. Or maybe you're trying to cut too deep in one pass?

Please do start your own thread if you have something to show, another CNC building thread would be nice to read! 

[menace]

Not sure if you have lee valley near you..

http://www.leevalley.com/en/hardware/page.aspx?p=44238&cat=3,41306,45375

Not as cheap as el cheapo but not 1.39 either

[Yvan256]

Not sure if you have lee valley near you..

http://www.leevalley.com/en/hardware/page.aspx?p=44238&cat=3,41306,45375

Not as cheap as el cheapo but not 1.39 either

Nope, closest one is Ottawa. Then again that means no duty or brokerage fees either.

Since I'm using 5/8" inch thick particle board/MDF, I would need the 16mm long cross-dowels, which don't have a centered hole. I'm not even sure why they make them that way.

Even in 100's quantities, it's still at least 4 to 5 times more expensive even without shipping compared to simply using a washer+nut. Ugly or not, if it costs about 5 cents for a nut and 1 cent for a washer, it means around 6.50$ for 100 nuts+washers bought locally vs around 45$ for 100 cross-dowels shipped, if it only costs 10$ for shipping. Given the weight of 100 of steel cross-dowels, I doubt it would be as low as 10$.

The price difference is almost enough to buy that TB6560 controller on eBay, since a lot of Chinese vendors offer free shipping.

Thanks for the link though, I might have future projects that would be better off with cross-dowels for a neater finish.

[Yvan256]

Sorry for the lack of updates but this whole project is driving me nuts, in a good way. People who have a CNC machine probably know what I mean. 

I've done a few tests on my old Roland MDX-3 and found that it can cut MDF pretty easily but I have to go extremely slowly and with extremely thin cuts. We're talking about 0.005mm deep cuts at something around 30 cm per minute. Doing a 20cm hole in 3/8" MDF takes around 15 to 30 minutes... I forgot to time it so let's round the value to "forever".

The reason it's so painfully slow is because this machine wasn't designed for such hard materials and the spindle motor is barely up to the task. Using a 3.175mm end mill surely doesn't help but it's all I could find locally.

I still don't know how to use Mach3 or LinuxCNC but since the MDX-3 is basically a printer that you send text files to, I did the test cuts with a PHP program I wrote. I plan on making the whole thing available as a "dirt simple 2D" program. I'll try to have it able to export both g-code as well as Roland's own format. I'm also writing a "2D Roland Web Simulator" to have a top-view of paths for 2D-only Roland Modela files. It's great for debugging.

Seeing the results with such a weak spindle, however, has led me to rethink my bigger machine. The first version will use my Dremel 395, I already bought the bits #115 and #650 for resurfacing and milling, respectively. The planned upgrade is to switch to a Black & Decker RTX.

Since both the Dremel 395 and the Black & Decker RTX are extremely lightweight, the whole CNC will not require to be built like a tank, helping keep its weight down and helping with the portability factor. Ideally I'm trying to have the whole thing designed as an oversized briefcase so it's easier to carry around.

Given all this new information, the goal is again to have a milling area of 20x24x1 inches, the shorter Z axis should make it easier to keep it strong and as flat as possible to keep it "briefcase-like", with removable Dremel/RTX. Yes I know, a 24x28x8 inches briefcase is going to be huge...

Attached are test cuts done on a variety of materials such as old membership cards, hardboard and recycled plastic containers. As you can see from the photos, some tests went better than others. 

But at least, I'm having fun! 


2012-12-17-4.jpg is a NEMA 17 motor mount gone terribly wrong.

[TopJimmyCooks]

subscribe 

[Yvan256]

The MDX-3 was busy yesterday after my last post, cutting a test part out of 3/8" (~10mm) MDF.

The motor was taken from an old 5 1/4" floppy drive.

[tightwad]

I love your progress and I am working on re-working my mini-cnc machine.  It has never had the accuracy I wanted and the Z axis had play in it which wrecked havok cutting anything more than foam.  I needed to see various bearing options.  I found great prices on the 608ZZ bearings on ebay...I think I paid about $0.50 each for the ones I bought in the past.

[Yvan256]

Yesterday I noticed that the horizontal cuts were not at 90 degrees. The MDX-3 had some slack in the horizontal axis assembly so I had to disassemble part of it just to access the bolt that had gotten loose. Really bad design as far as maintenance is concerned.

After tightening the bolt and putting it all back again the results were much better. I have not measured it but at least just looking at it you can't tell if the edges are straight or not. A great improvement!

I asked the neighbors below if the noise of the machine was annoying and their reply was "what machine?" I told them that if it ever was a problem to simply come and tell me. The fact that the machine is sitting on five layers of oversized bubble wrap sheets seems to prevent the vibrations from reaching the table it's on.

Since I can run the machine all day from my apartment, I plan on making the new bigger machine completely with the MDX-3 using smaller plates that bolt together. The bigger machine, once assembled, will be able to cut continuous replacement panels for itself if needed. The MDX-3 can only cut parts about 150x100mm, so that's quite a challenge indeed. The more I design and build, the more I'll see if it's possible. The number of panels will be high to say the least.

So this morning I started to design my new machine, starting with the Dremel 395 mount. I plan on designing and cutting the machine from the inside-out, going from the tool holder to the Z axis (top/bottom), then the X axis (left/right) followed by the Y axis (back/front).

The photos below are the first version of the bottom Dremel mount holder but I will have to cut it again. The two small round holes were supposed to be for cross-dowels but these were made for 5/8" panels and I'm using 3/8" MDF because of the cutting height of my end mill bit.

I'm currently cutting bolt+washer+nut assembly test plates so that the next Dremel plates will have the proper assembly holes.

[Yvan256]

Here's the test for the bolt+washer+nut assembly method. The hole for the bolt was just a bit too small and the rectangular hole for the nut and washer just a bit too wide and tall. Still works fine but I'll fine-tune the measurements when cutting these holes on actual parts.

The MDX-3 is currently busy cutting the upper panel for the Dremel 395 mount, I'll post an update in the next few hours.

[Yvan256]

Here is more pictures of the Dremel 395 mount.

The first image is my web-based "design" software, currently written in PHP and HTML5 (using Canvas), showing the part to be cut. The lines are extra-wide because they represent the cutting diameter of the end mill (1/8 inch or 3.175mm). The light grey line represents the cutting area of the machine.

The other photos are the MDX-3 about one minute before it finished cutting the plate. I still need to get a proper cutting plate with rails and clamps.

The last three photos show the Dremel inside the plate and a small cut I had to do because I incorrectly measured the diameter of the body. Still some mistakes to correct so I'll be re-cutting these two plates tomorrow with the required corrections as well as the back plate for them. In 24 hours I should have a proper Dremel 395 mount with mounting holes, ready to be attached to my yet-to-be-designed machine.

I don't mind re-cutting all the plates again because the CNC does all the work.   

It's really amazing to design something on the computer and holding the part in your hands a few minutes/hours later. It's kind of the reverse of "Tron", taking something from the digital world and bringing it to the real world.

[Yvan256]

Good thing I was still making new joints tests and the Dremel 395 mount was on hold, because yesterday my brother gave me this as a Christmas gift!   

The RTX is rated at 2 amps compared to the 1.15 amps of the Dremel 395. If you judge by the numbers it means it should be almost twice as powerful but from the torque alone you would think the RTX is almost three to four times better. It's also a lot quieter at the slower speeds, which is extremely good. And it's more or less the same weight so I won't need an over-engineered gantry to support it.

Step motors: check
Electronics to drive the step motors: check
Linear slides: check
End mill bits: check
Powerful and lightweight spindle: check!

[Yvan256]

After looking at a few dozen types of CNC joints, I choose the type with tabs because it allows the parts to align themselves in all three axes. A simple manual drilling in one part is required but centering is done by a CNC'ed hole in the other part in front. Kind of hard to mess it up. 

Can't wait to finish my CNC machine and start making actual arcade projects, I know it must be boring reading about all these changes after over a year since I started this thread.

Well, here's a few photos, feel free to leave replies, suggestions and ideas!

[crashwg]

That's pretty neat.  I'll have to remember that method for machine #2.

[Yvan256]

While designing the new CNC machine, I had to find a way to hold those LM8UU linear bearings for the Z-axis. The most compact way to mount them is parallel to the sheet material (MDF or otherwise) so I had to create something simple, compact and easy to mill. Easy for my current little CNC anyway. People with 3D printers make clamp-style holders but with flat sheet material you need to think differently.

I made a rectangular hole, 4mm deep, with four holes for cable ties. The bone-shaped rectangle is to make sure the inside corners are actually square without the diameter of the cutting tool.

The more I move the small plate by hand, the more I think those linear bearings are worthless. Instead of a smooth glide I feel some resistance. The skate bearings assembly feels a hundred times better by comparison. The size and weight difference between the two assemblies, however, is far from negligible. Using the LM8UU will allow me to build a very compact and lightweight Z-axis.

And since I'm starting with the Z-axis assembly anyway, I'll be able to decide if it's good enough once I have it working with a stepper motor and the RTX. Maybe with the stepper motor and the screw moving the plate it won't make much of a difference compared to the skate bearings. So many people use the LM8UU linear bearings with their 3D printers that they can't be that bad.

[Yvan256]

Since I can only cut panels 140x90mm with the MDX-3 (150x100mm minus the diameter of the tool all around the perimeter), the design of the bigger machine has to be made from a lot of small panels.

This is what I've come up with for the sides, to fit within the 140x90mm limits and to still get something bigger and yet sturdy enough for the bigger CNC. Twelve panels, 20 bolts, 20 nuts and 40 washers are required for each side. The material used will be 3/8" MDF since the MDX-3 seems to have no trouble cutting it, albeit slowly.

Am I the only one who thinks it kind of looks like a building with windows, from the game Rampage?

[nemesismachine]

If the cable ties ever start to loosen or eat into the MDF, you could always go with pillow block bearings. You can order them online, or a lot of tractor supply stores carry them. I used a few for a wheel polisher I made.

[TopJimmyCooks]

Wow - you're going to cut 10 pieces of MDF and ~20 bolts.  I can't help but think that four full sized pieces would give better, cheaper, more rigid results.  Do you have a friend with a tablesaw? 

It does look like a good solution to cutting pieces only with the smaller CNC.  I don't want to be super negative, but if you're working MDF, you have to have a way to cut larger pieces.  It comes 48 or 49" x 96 or 97". 

[CoryBee]

Love this build!

[mcseforsale]

Subscribed...

AJ

[Yvan256]

Wow - you're going to cut 10 pieces of MDF and ~20 bolts.  I can't help but think that four full sized pieces would give better, cheaper, more rigid results.  Do you have a friend with a tablesaw? 

It does look like a good solution to cutting pieces only with the smaller CNC.  I don't want to be super negative, but if you're working MDF, you have to have a way to cut larger pieces.  It comes 48 or 49" x 96 or 97".

I think the result will be at least as strong as a single piece of 5/8" MDF, if not more. Using the two small "tabs test" panels as a guide, I can tell you that it's extremely rigid. The goal is to prevent warping from side loads and I think it will be more than enough for that goal.

I will be able to use the finished machine to cut new parts for itself once it's assembled, to gain some space in the Y axis (curved sides). None of the people I know have a scrollsaw and I don't want to cut complex shapes by hand. This is why I'm building the CNC in the first place.

Also, the sides need holes for the motor, the rod, bearings, etc. All the people I know aren't equipped to drill half-depth round pockets so going with CNC'ed panels removes that problem.

The base and surface of the machine will be cut with a tablesaw (5/8" MDF) since it's only plain rectangles and I have someone who will help me cut a big sheet of 3/8" MDF into 140 small panels for my MDX-3... I just calculated that the pile of panels will be more than 1.3 meter high. 

My biggest problem at the moment is finding a place where I could buy a small 28x14cm piece of slatwall, all they have is 4x8 feet sheets. 

[tightwad]

at what linear speed and depth are you able to cut?  what kind of bit do you use?  How do you secure the piece to the table?

[Yvan256]

at what linear speed and depth are you able to cut?  what kind of bit do you use?  How do you secure the piece to the table?

Are you talking about the Roland MDX-3? The speed is set to "5" (I think it means 5 mm per second, although it seems faster than that to me). The cutting depth is 0.02mm.

The bit is a 1/8", 1-flute straight end mill (the website says 3/4" for the cutting length but it's wrong, it's around 1/2"). I could probably go faster or cut deeper on each pass with a 1/16" bit, but most places I find online don't have 1/16" bits with 1/4" shanks that can cut 1/2" deep. And the places that do have those bits sell them for 40$ or more, plus shipping, etc. It's not worth the cost.

I have a 1-slot slatwall MDF panel instead of the flat plastic sheet that comes with the MDX-3. To secure the piece I use bolts and big washers and the piece is held on two sides by the edge of the washers which are at about 45 degrees once they hold the piece. I'm trying to get a larger piece of slatwall and I plan on making better clamps to hold the panels in place.

Cutting the smallest panel in the side panel assembly shown above will probably take 2 to 3 hours, the bigger ones will probably take 5 to 6 hours. Doesn't really matter, I'm not in a hurry to complete it. I'm just happy to be able to use a small CNC to cut parts for my bigger CNC.

The machine I'm building will use a Black & Decker RTX with a Dremel #650 bit.

[Bender]

I'm just happy to be able to use a small CNC to cut parts for my bigger CNC.

  That my friend is how the robot apocalypse starts

I love it!

[Yvan256]

I'm just happy to be able to use a small CNC to cut parts for my bigger CNC.

  That my friend is how the robot apocalypse starts

I love it!

Nah, we're safe, I won't let the bigger CNC machine cut another even bigger CNC machine. 

[Yvan256]

Yesterday I bought a 3/8 inch thick, 49x97 inches MDF panel and cut it into smaller panels that can fit inside the Roland MDX-3. I ended up with 120 panels of 17x11cm and 40 smaller panels roughly 9~10x11cm.

Here's a photo of the piles of panels next to the Roland MDX-3 cutting the very first panel of the bigger CNC machine, sitting on five layers of big bubblewrap sheets to eliminate the vibration. The old IBM ThinkPad 760XL is sitting on top of the three piles of 17x11cm panels.

I'm still making the cutting files manually in Google Sketchup by measuring the coordinates of the tool and typing them inside my custom MDX program made in PHP. It's slow but at least there's no costly software involved.

The second photo is an attempt of taking a close-up of the tool cutting one of the holes in the panel. I think I'll need to tape a flashlight on my Nintendo DSi.

edit: the third photo shows cutting progress two hours later.

edit: the fourth photo shows the finished panel four hours and twenty minutes later. Did I mention patience was required to cut MDF with the MDX-3? 

[kahlid74]

I like where you are going so far.  I too am building a CNC machine but I decided with a 4x6 cutting area and a mobile base that folds up so the machine can be moved around.

Keep up the good work!

[Yvan256]

I like where you are going so far.  I too am building a CNC machine but I decided with a 4x6 cutting area and a mobile base that folds up so the machine can be moved around.

Keep up the good work!

Thanks!

Do you mean 4x6 cm, 4x6 inches or 4x6 feet?

I like the folding/mobile idea, do you have a thread for your build?

Update on my build: the MDX-3 is currently busing cutting the second panel of the bigger CNC machine. Same as the first one above except with two holes for bolts.

[kahlid74]

I like where you are going so far.  I too am building a CNC machine but I decided with a 4x6 cutting area and a mobile base that folds up so the machine can be moved around.

Keep up the good work!

Thanks!

Do you mean 4x6 cm, 4x6 inches or 4x6 feet?

I like the folding/mobile idea, do you have a thread for your build?

Update on my build: the MDX-3 is currently busing cutting the second panel of the bigger CNC machine. Same as the first one above except with two holes for bolts.

No thread yet, I'm procuring all of the parts now.  It's a 5 feet by 6.5 feet totaling a cutting area of 4 feet by 6 feet.  I wanted to be able to mill a side of a regular arcade cabinet or as close to the standard size as possible while still being mobile.  The base is 11ga steel 2" square tubing welded together.  There are two cross beams that will stabilize the whole getup.

The design I'm using is on Joe's CNC board which you have to pay to join.  Here's a picture of the guy's machine I'm modeling mine after:
It's Dkohfeld's CNC machine and he is actually a frequent of these boards as well including Mountain, whose CNC machine is the stuff of legend.

[Yvan256]

No thread yet, I'm procuring all of the parts now.  It's a 5 feet by 6.5 feet totaling a cutting area of 4 feet by 6 feet.  I wanted to be able to mill a side of a regular arcade cabinet or as close to the standard size as possible while still being mobile.  The base is 11ga steel 2" square tubing welded together.  There are two cross beams that will stabilize the whole getup.

The design I'm using is on Joe's CNC board which you have to pay to join.  Here's a picture of the guy's machine I'm modeling mine after:
It's Dkohfeld's CNC machine and he is actually a frequent of these boards as well including Mountain, whose CNC machine is the stuff of legend.

That's a pretty neat idea to have it fold that way, takes a lot less room when not in use. 

Mine is going to be big enough to make desktop cabinets yet small enough to carry by myself. It won't be lightweight but not heavy enough that I can't carry it and get it in the back seat of my car.

[Yvan256]

Yesterday's panel now has a slightly modified clone... the holes are for #10-24 bolts.

A third clone should be completed in a little more than four hours.

You can see the precision in the second photo, with the two panels inserted into one of the test assembly plates. I can't even slide a thin sheet of paper vertically between the two parts. 

edit: it's triplets!

[Yvan256]

I needed to modify the plate holder and make new clamps to hold the bigger 17x11 cm panels. Here's a photo in case someone else with a Roland MDX-3 needs to see how I've done it. There's a part bolted at the bottom that holds the panel in place since I'm leaving the whole holder in the machine at all times. Two bolts+washers+MDF spacer assemblies are on the sides of the panel to be cut, I simply tighten those to hold the panel in place against the holder.

Back on topic, the MDX-3 is now cutting the plate that will hold the X-axis motor (Sketchup screenshot shown below, second image). It should be ready by the end of the day tomorrow. I'll update with a new photo once it's completed, assembled with the two smaller plates shown above as well as the motor itself with the coupler and the threaded rod, as seen in the third image.

After three days of cutting I'll be happy to see some progress!

edit: even if I have to go extremely slowly to cut MDF with the MDX-3, I'm beginning to understand the feeling of owning a CNC.   

edit2: I really under-estimated the cutting time, so I had to stop it for the night. The CNC always returns at the same position when starting, so I'll just send it another file to cut the outer edge. Attached is a photo of the progress so far.

edit3: I added two bolt holes for another plate in my design, so I cut those two extra holes before cutting the outline, which is what the MDX-3 is doing right now. It was already half-done so I would expect to have the plate ready before lunch, in roughly an hour and a half.

[Yvan256]

Talk about slow... it takes about 5 hours and 30 minutes to cut the outer perimeter of a big 14x9 cm panel. 

I don't want to push the MDX-3 to cut faster and deeper, however, because of the noise and the risk of breaking the machine.

It's currently busy cutting a 4th small panel with tabs as shown above, but for the moment here's two photos of the completed panel that holds the stepper motor for the X axis of the big CNC. I still need to drill holes into the small panels with tabs to be able to bolt those panels together but that will come later once more panels are ready for assembly.

The first photo is from the "outside" and the second photo is from the "inside" with a home-made fuel line coupler.

[Yvan256]

I started building the Z-axis today, the MDX-3 is busy cutting the outline of the top panel, right below the motor, shown in the assembly below. It should be done cutting the panel in less than two hours. The motor is a NEMA17 from an old floppy drive. The sticker says 1.8 degree but manually I count 50 steps. I'll need to test it to know more, once this whole assembly is completed.

The first two images are the 3D models of the complete Z-axis assembly, apart from the missing bearing for the threaded rod in the bottom plate and the plate that will hold the RTX.

The third image shows the MDX-3 cutting the top plate of this assembly.

I just calculated the cutting speed at roughly 3.4 mm per second.   

edit: the plate is cut. The fourth and fifth images show the finished plate, and the plate upside-down with the motor, bolts and 8 mm rods. I should be able to cut the bottom plate tomorrow.

[Yvan256]

The bottom plate of the Z axis assembly is done. There's a notch in the corner because I'm still not using milling tabs to hold the part when cutting and I was away when the MDX-3 was done cutting. 

You can see the top and bottom panel in the photos below. The bottom plate has half-depth holes for the 8 mm rods just so they won't fall down when the machine is working, although they would probably hold in place by friction alone. I didn't take the picture with the rods in place because I need to cut them to the proper length.

I'm still missing the three vertical panels that will go between them, I should be able to start cutting those tomorrow.

The MDX-3 is currently busy cutting a long panel with tabs for the X-axis assembly, I'm hoping it will be done before bedtime...

[Yvan256]

One of the four long panels with tabs is now done, but I was getting tired of looking at this MDF-colored assembly... boring. 

Since the little Hang-On Shop Vac is red and black and the Black & Decker RTX is also red and black, I started painting the panels. I've decided the panels with their face visible from the side would be red and the panels with their face visible from the front, top or bottom would be black.

I've also decided to paint the side of the parts the same color as the panel itself instead of the face of the joined panel to get an interesting effect, to show off the assembly method of the whole machine instead of trying to hide it (i.e. the tabs will show up in black through the panels).

And here are the results so far. It looks better in real life than in those photos, I guess the close-ups really magnify the defects and the texture of the MDF.

For reference, the paint used is Crafter's Acrylic from DecoArt, Bright Red DCA22 and Black DCA47. I'll be applying two or three coats of gloss varnish DS13 after two coats of paint. In those photos, the red panel has two coats of paint and the black panels only one.

Photo 1 shows the left side of the X-axis assembly so far, upside-down to be able to take the picture.

Photos 1, 2 and 3 shows the "interesting effect" I was talking about, showing off the assembly method (tabs are in black) instead of trying to hide it by having all the side red.

Photos 3 and 4 show the bolt+washers+nut assembly method discussed in a previous post, from both sides. Since the MDX-3 can only cut perpendicular to the panel itself, I need to carefully drill the holes for the bolts, manually.

Photo 5 shows, once again, a close-up of the home-made fuel line coupler that attaches the motor shaft to the threaded rod.

[Yvan256]

Some parts require manual drilling on the edge of the panel. Messing up one of those hole would mean re-cutting a part and wasting another 4 to 12 hours of CNC time depending on the size of the panel. Some panels have zero holes to drill manually but others may have four or more, that increases the chance of messing up when doing things manually.

So right now the MDX-3 is currently busy cutting a drilling jig to make sure all my manual drilling is spot-on, every time. First image is the design in Sketchup, with the four panels to cut in the back, the complete drilling jig in the middle and the jig without the top panel on the right to see how it will lock in place for the panels with tabs. The big round hole in the middle of the two top panels is for a drill bushing insert (Lee Valley Tools Item 25K6220). The first photo is the MDX-3 cutting those four panels.

edit: the MDX-3 is done cutting all the holes in the four panels and is now starting to cut the outlines.


Progress on the Z-axis assembly. I still need to cut the back panel and the panel that will actually move the RTX up and down. The second photo shows what it looks like so far with the panels held in place by gravity, until my drilling jig is complete. I still need to apply varnish for the final touch.

[Yvan256]

Just to recap things since I've joined BYOAC, I'm building a CNC machine to be able to cut parts for my arcade projects.

But I need to apologize to everyone who reads my threads.

My MVS-99-6 and ColeCade projects have been on hold practically forever.

My CNC256 project is taking forever to properly start and my perfectionism means I keep starting over. Since these are personal projects I can allow myself to aim for perfection. There is no set deadline, no launch date to meet and no budget to speak of except my own spare time.

Having the MDX-3 to cut parts, however, is helping me to concentrate on the designs in Sketchup and have perfectly cut parts on my desk a few many hours later. This was the missing tool in my process.

But... the MDX-3 is old and was probably never designed to cut MDF with a 1/8" end mill bit, so the panels are taking forever to cut. What should be a 15-20 minutes job is taking 6-8 hours.

So I decided to make "CNC64", a smaller machine with the same cutting envelope as the MDX-3 so that I don't have to trash the 100+ small MDF panels that I have.

And since CNC64 will only be built using panels as big as it can cut, I guess it could become the "reprap" of the CNC world. I guess I'll build a second CNC64 with the first one just to prove the concept and to see how long it takes for it to "replicate" itself.

The smaller CNC64 will require fewer panels (current count is only 37 CNC-cut panel) and a simple rectangular table saw/hand cut base with not much precision required and thus will be ready in a much shorter time than CNC256. Without counting the base and without optimization of cutting multiple smaller parts on a single 6x4" panel, the MDF cost is currently around 8$CAD for one CNC64 machine. That's without counting the machining or design time, however.

It also means I can practice some design ideas on a smaller scale and in the end I will have two faster machines to cut more panels at the same time, if needed. At the very least it will give me a more portable machine for when I only need to cut smaller panels.


Back to the topic of my previous post, the two side panels of the drilling jig would have required a drilling jig to make precise holes. The first hole I drilled on the side wasn't exactly centered and the second hole just split open the MDF because I tried to drill too fast. 

I also messed up the two side panels of the Z-axis assembly, once assembled the whole assembly wasn't even level. Not a good start for a CNC machine. 

So, from all this came a new joint method. The method I showed earlier is based on the joint method used by the first version of diylilcnc but since 3/8" MDF is a much thicker material I decided to combine the tab and the bolt coupling in a single, compact part with no post-machining drilling to do. Take the panels, join them, bolt them. The only way I can mess that up is by doing a mistake in the Sketchup design.

In the five photos below you can see that each joint requires six parts:

• A tab with a channel for the bolt and a hole for the nut and washer
• A hole for the tab and a hole for the bolt
• A bolt
• A nut
• Two washers (to make sure that the MDF doesn't get twisted and ripped apart while tightening the bolt)

The channel for the bolt is not cut all the way through, so there is an "inside" and "outside" for the panel. I'm doing it this way for two reasons:

• It looks cleaner from outside
• It makes a stronger joint because there is more material for the washer and nut to sit on.

For reference, the small test assembly shown in the last photo is only 30 mm wide (left to right), 20 mm deep (front to back) and 30 mm tall (bottom to top, without the bolt head). The tab itself is 15 mm wide and 3 mm tall but I'm switching to 5 mm tall for the final parts for an even stronger joint. This means the tab inserts more or less 50% into the thickness of the other panel (I measured the thickness of the MDF a around 9.7 mm).

I'm calling this joint "100% CNC-Cut Compact Joint", or "100CNCCCJ" for short.

... okay, I need a better acronym, not to mention that it probably already exists and already has a name.

Hopefully, CNC64 won't take too long to build, then CNC256 and then I will finally be back to posting about arcade projects!

[mcseforsale]

Cool...a place for your nuts!

Love the build so far.

AJ

[Yvan256]

Thanks for the feedback, mcseforsale.


The MDX-3 is has been cutting the first panel of CNC64 for about two hours from an estimated cutting time of around three hours. I need twelve of those plates for CNC64, which means roughly 36 hours to cut these simple assembly plates. 

Now you see why I'm building CNC64!

edit: It took the MDX-3 about 2 hours and 28 minutes to cut this small plate (shorter than my 3 hours estimate), meaning twelve plates will take around 30 hours, so 27 hours and 30 minutes to go for the other eleven plates. The rounded corners certainly didn't help the cutting time but I was getting tired of making rectangle plates. Nothing says "CNC" like rounded corners! CNC64 completion countdown: 36 panels to go.

edit2: here's a partial preview of CNC64 in Sketchup. You can see where those 12 plates go.

[Yvan256]

I forgot to mention that I modified the RTX by following these instructions (thanks cre8er!) because the speed dial was really hard to turn, I couldn't imagine turning that dial once the RTX would be in place in the CNC mount. After that small and easy modification, the speed dial is now very easy to turn. 


I now have four panels ready to build CNC64, those are small panels used to join other panels end-to-end. The holes are a bit smaller than the bolts to allow the threads to run into the MDF and prevent any movement once assembled. I had to dump the first plate with the rounded corners because the holes were 0.5 mm too big in diameter.

I switched back to rectangular plates to save about 10 minutes of cutting time per plate and because the MDX-3 was doing weird noises while cutting the rounded corners. I'll let CNC64 make rounded corners when cutting the parts for its clone.

I promise the next few parts will be more interesting to look at.

That's 4 plates done out of... 42.

[Yvan256]

(cheesy TV announcer voice) "Tired of looking at small rectangles with round holes? How about looking at a longer rectangle with holes and tabs?"

As seen in the Sketchup render in a previous post, this is one of the two side panels.

The first image is the 3D render in Sketchup, the second image is the 2D render in Sketchup with the milling bit circles all around the part to calculate the polygon outline coordinates. The third image shows the cutting outline path in my own web-based "CAD" program, if you can even call it that. The photo shows the MDX-3 cutting the panel.

edit: the last two photos show the beginning of the assembly. Don't mind the small part at the front of the panel in the 5th photo, it's just my test plate for the new joint I talked about. The MDX-3 is busy cutting the same panel again for the right side of CNC64. The next two panels to be cut will be either the other half of the sides or the front of the machine. Five plates done, 40+ to go. I'm also going to need longer bolts for assembling these small plates...

edit 2: the very last photo shows the six panels that are ready and assembled and the two sides roughly at the correct distance from each other if all parts were there. I also need to paint the other side of the long panels as well as the edges of all six panels. I know it's not an arcade cabinet, but I just can't wait to add more panels for the base! 

[Yvan256]

I'm not sure people still read this thread judging by the lack of comments, but then again there's not much to comment on. I apologize for those boring updates. 

The MDX-3 is cutting the first front corner panel that will connect with the long panel from yesterday.

edit: wow, 6400 views 

edit2: added a photo of the final part, the MDX-3 is busy cutting another for the other side.

edit3: another photo, the front left corner of the first CNC64 is complete! The other Sketchup screen capture shows a corner that will be added to the first CNC64 clone for added strength. I made this update while the second long panel was still being cut yesterday, we'll see if I need to cut them again to add these corners once it's complete.

[Drnick]

Don't apologise for Boring updates, At least you keep giving us updates.  Gonna be fun when you finally have 2 cnc 64s busy making up the big brother cnc256.

And on top of that, reading this is still better then staring at ping traces trying to work out why we keep losing connection to one of our DNS Servers.    I am sure that the problem is in the routing

[Yvan256]

Don't apologise for Boring updates, At least you keep giving us updates.  Gonna be fun when you finally have 2 cnc 64s busy making up the big brother cnc256.

I did not even think about the fact that there is going to be two CNC64's making the parts for CNC256!

And on top of that, reading this is still better then staring at ping traces trying to work out why we keep losing connection to one of our DNS Servers.    I am sure that the problem is in the routing

Data getting lost in tubes, you say? Better call Mario and Luigi!

[Yvan256]

Here's what all the six panels look like so far. At last, it's beginning to look like something!

I'm still looking for bearings for the threaded rods, the local stores are asking insane prices for those. The next panels will be the ones shown in the Sketchup image.

[Yvan256]

I just added 400 more parts to the project.

Okay, so I cheated a little. It's 300 bolts and 100 nuts. Those parts are needed, maybe just not all 400 of them.

I cleared a place on the workbench, it's now dedicated for building CNC64. Still the same 8 panels except this time they're all bolted together. There's also a photo of the next panel being cut, it should be ready in less than two hours about 15 minutes. I'll update this post once it's cut, cleaned and bolted to the right side.

edit: post updated with 3rd photo. It's almost funny that CNC64 is around the same depth, front to back, as MVS99-6.

[Yvan256]

I moved the MDX-3 to my parents' house so I'm able to cut panels about six times faster. I cut five panels yesterday, should have been six but I screwed up on a small panel, hence the bolts sticking out the front on the right side.

In my apartment I used the settings "speed 15, milling depth 1" to cut down on the noise, but at my parents' house I use the settings "speed 5, milling depth 10". A lot more noise, the cuts aren't as clean on the edges but the precision still seems spot-on. The MDX-3 has an automatic speed controller, if the spindle can't cut as fast as I ask it will slow down the cutting speed to compensate.

The front panel should say "CNC64" but I cut too deep and the middle dot of the "6" just got ripped away by the router bit. 

I need 19 more panels to complete the base of CNC64, not counting the moving platform for the Y-axis (front to back).

[Yvan256]

I had to re-cut the front panel with the bearing. I also still need to recut the two long red parts on the left and right which are not connected to the front in the photo and another small panel like the ones in the front. Those small setbacks aside, it's progressing incredibly fast since I've moved the machine to my parent's house. I still need to paint all of those panels, too.

I'll need to cut a longer threaded rod and make it fit the bearing at the front, cut the 8 mm smooth rods and build the moving platform but the Y-axis is nearly complete.

I can't wait to build CNC256 so that my MVS99-6 project can progress as fast as this build! We create the design and let the machine cut the parts. Everyone should have a CNC machine! 

[Yvan256]

Since CNC64 uses threaded rods to move the axes I'm going to need a controller to move them manually. Instead of going for a ready-made and boring controller, I thought I'd design one in line with the kinds of projects I have in mind for the bigger CNC. Cutting a second controller for CNC256 will also means one less thing to design for the bigger machine.

I still haven't found proper buttons for the up/down movements and maybe a spindle on/off button, but here's what it would look like so far. The best part is that I designed it so the MDX-3 can cut the parts required.

I'm not sure about adding small T-molding because the MDF is 3/8" thick (9.7 mm) and I'm having a hard time finding any in black, I've only found Polished Brass at Suzo-Happ.

So far I'm thinking about using the Grayhill 04A-B01 joystick for the X/Y movements but I need suggestions for the buttons, especially from those who built micro/mini cabinets using the 04A-B01. Your buttons seem to be the right size compared to the joystick. I only use Digi-Key for electronic parts because ordering from U.S.A. suppliers always means brokerage fees and duty charges making the whole process incredibly more expensive than the parts themselves.

Since there is no holes to attach the joystick I was thinking about holding it between the top plate and a smaller MDF block underneath, with a hole cut inside to fit the joystick body.

[Nephasth]

Planning on taking any orders when you've got your big CNC up and running? I could use a few stand alone CPs (flat packed) for the Troubleshooter down the road...

[Yvan256]

Planning on taking any orders when you've got your big CNC up and running? I could use a few stand alone CPs (flat packed) for the Troubleshooter down the road...

As long as someone has precise plans and it fits inside the cutting area of the CNC machine, I'm willing to take custom orders for various parts. What's the size of your CPs and which material and thickness are we talking about? I don't think I'd be able to machine the top of an aluminium briefcase directly, not necessarily because of the aluminium but because of the thickness of the briefcase itself. Unless you can send only the "top panel" part of the briefcase, if it's at all possible.

[Nephasth]

Lap size, 5/8", mdf or ply

[Yvan256]

Lap size, 5/8"

The cutting area for CNC256 is planned for 24x19 inches and my main target is 5/8" MDF/particle board/plastic, but I'll be trying 1/8" aluminium too.

[Nephasth]

I'll be trying 1/8" aluminium too.

Oh really?

[Yvan256]

I'll be trying 1/8" aluminium too.

Oh really?

I've been tempted to try 1/8" aluminium on the Roland MDX-3 for a long time. At the slowest possible speed and lowest possible cutting depth, I'm pretty sure it would work.

Keep in mind it's not a laser, so internal corners will have a 1/16" radius because of the 1/8" cutting bit. To cut aluminium, however, I could buy special bits. I tried cutting PCBs with a "diamond chipbreaker" bit and the results were nothing short of amazing. If it can cut PCBs I'm pretty sure aluminium wouldn't stand a chance.

Let's continue this discussion via PMs for your custom CPs request.

[Yvan256]

Small update on the MVS-style controller for CNC64/CNC256.

I still need to find buttons of the proper size, ideally in red, yellow, green and blue. Apart from the holes diameters and positions for those buttons, the plans are more or less complete for the controller casing.

For such a small part, it's actually quite complex because I wanted flush edges in the front and back and a minimum of bolts to assemble it.

[Nephasth]

What size do you need the buttons to be?

Here's some 20mm buttons in the colors you're looking for: http://www.paradisearcadeshop.com/en/95-sanwa-20mm-snap-in

[Yvan256]

What size do you need the buttons to be?

Here's some 20mm buttons in the colors you're looking for: http://www.paradisearcadeshop.com/en/95-sanwa-20mm-snap-in

I already have a yellow, green and blue SDM-20, but the "ring" of the body has a diameter of 22.8 mm and the plunger a diameter of 14.1 mm.

From my Sketchup plans and the Grayhill joystick I would need 13.5 mm rings with 10 mm plungers. I found 10 mm bouton caps in the proper colors at Digi-Key (AT422C, AT422E, AT422F and AT422G) but the buttons that are supposed to go under these caps are way too expensive, I'll probably make something from cheaper buttons.

I'm also thinking about using an old SNES gamepad, I think the buttons are small enough and they're already the right colors, I would only need to make a custom PCB to go underneath.

Below is images comparing the Sanwa SDM-20 and the NKK AT422* caps. You can see the SDM-20 is way too big compared to the joystick. I'm not aiming for an arcade-perfect, scaled-down version of a real MVS control panel but it just looks wrong if the buttons are bigger than the joystick. There's also the small detail that the buttons are physically overlapping themselves.

[Yvan256]

This is what I have so far, the photo is missing the two long vertical plates shown in the Sketchup image because I'm unable to make parts float in mid-air in real life. The machine is currently disassembled because I had to paint some panels.

I'm thinking about possibly redesigning the back plate where the motor is attached because it's extremely fragile due to the big hole. The redesigned panel would fit NEMA17 and maybe NEMA34 mounting holes, so I would make a NEMA34 "adapter" plate for my low-cost Hurst motors. Putting the motors and their couplers "outside" the machine would also allow more flexibility in choosing parts for the machine.

It's only missing a few parts before I can start testing the Y axis! The short threaded rod in the picture is 1/4" but it's too small for the 608ZZ "roller blade/skate" bearing, so I need to make my own couplers to go from the 1/4" motor shafts to 5/16-18 threaded rods and a small holding plate for the bearing in the front. That's what the two unused holes next to the bearing are for. The temporary electronics are ready for the motors, I simply hard-coded a cutting pattern into an ATmega328 so I can test the motors without needing a whole computer. I'm planning to use Grbl for computer-free operation and for a more compact and portable system without having to rely on obsolete parallel ports.

The design for the rest of the machine is almost complete but I want to make sure everything fits together in Sketchup before showing the whole machine.

[Yvan256]

There's already four panels ready for the upper part of the base assembly, but in the meantime here's a photo of the fully painted machine so far. And yes, I need to cut that threaded rod a bit.

The second image shows the nearly-complete design of the machine. Don't mind the two motors floating in mid-air.

[kahlid74]

Looking great so far man.  Keep on getting!

[Yvan256]

Sorry, no photo for this update because of a technical problem. Something about gravity.

Here's the progress so far, Sketchup-style. The two big panels at the back and the bearing plate for the X-axis are ready to be painted.

[Yvan256]

Here's a few photos of the last panels, some of them half-painted. Also included is the first panel cut in "expanded PVC plastic sheet", given to me by a local sign shop. I have enough of these to make the whole RTX mount assembly. The cutting depth wasn't set correctly for this new material so I had to finish cutting the last 0.2 mm by hand, I'll fix that mistake for the next panels.

The new MDF part is an LM8UU holder, I'll need three more of these to go with the still-to-be-cut Y-axis bed.

Also shown is the RTX with the lower part of the mount assembly. The upper part and the back of this assembly will be cut next. I like this PVC stuff a lot more than the MDF, but given its cost I have to stick with MDF for now. But if I have enough PVC panels I'll be making the mini-MVS-controller out of it.

I've only been playing with the Roland MDX-3 for a few weeks and all I can say is that having a CNC router is addictive. I design on the computer one day and the next one I have the real part in my hands. I still find that amazing. 

[Yvan256]

The left side panel of the upper part of the base has been cut, as well as the two big panels of the back. Only one more side, two smaller back panels and four joint panels are needed to complete the base.

The side panel has a lot of holes because I made it compatible with my Hurst motor, NEMA17 and NEMA23. Just in case... 

[Yvan256]

I cut a lot more panels since last time, here's the first update: the RTX mount is complete, minus the holes for the bolts at the front which need to be drilled manually. The whole mount is made of expanded PVC panels.

I'll be updating this post in a few hours to add photos of the completed and painted base.

edit: Added a third photo. It's already too old but at least it shows a bit of progress. The machine is actually quite more advanced than this, the base is completed and painted. I've already cut some parts of the Y-axis bed and some parts of the X-axis carriage.

edit 2: if anyone found this post searching for the diameter to use for the upper part of the RTX mount, I used 49.4mm and it seems to fit perfectly. There's some empty space because of the non-circular diameter of the body but it doesn't move in any direction.

[Yvan256]

The base being completed I have started cutting parts for the moving gantry and platform.

I've moved the "CNC64" inscription to the top right of the machine. The last photo shows the beginning of the moving platform for the Y axis.

You might notice some holes without bolts because I made a few mistakes about their position. I'll try and drill them manually since the machine is already bolted together and there's already other bolts in place to keep things aligned.

[jammin0]

This thing is looking pretty slick.  I can't wait to see what you come up with when you have a faster, bigger machine.  From concept to design to machine it's pretty sweet to see what you come up with.

[Yvan256]

This thing is looking pretty slick.  I can't wait to see what you come up with when you have a faster, bigger machine.  From concept to design to machine it's pretty sweet to see what you come up with.

Thanks. It's actually quite hard to concentrate on building CNC64 when I'm thinking about CNC256 at the same time. And after CNC256 is built and working properly, the real fun will begin. 

[Yvan256]

I only cut five panels this week-end, four of those are for the X-axis assembly. Two more panels and this assembly is complete and I estimate less than ten panels to complete the whole machine. It's almost too easy to build things when you have access to a CNC... or maybe I'm just not good enough with hand tools.

Here's two more photos of the progress so far. The X-axis assembly is quite sturdy but those LM8UU linear bearings are noisy as hell when moving the whole thing by hand, though the noise is lower when I move it slowly. Maybe I'll try bronze bushing for CNC64 Mark II.

[CoryBee]

Oh my jesus......

[Nephasth]

Looking sharp.

[Yvan256]

Here's the Z-axis, made of four more panels of expanded PVC, four LM8UU linear bearings, ten bolts and four nuts. It's only missing the delrin/acetal nut for the threaded rod in the back (the yellow part in Sketchup), otherwise it's complete.

[Santoro]

!

Subscribe.

[Yvan256]

Yesterday I cut the last panel for the X-axis assembly, which was the top panel.

I made a huge mistake on the tool support assembly... Given the vertical range of the Z-axis, the router bit would never have touched the material to cut. I'm not sure what went wrong in my Sketchup calculations. Maybe I shouldn't have used a Dremel 395 model for my measurements. 

Since I'm running low on expanded PVC panels, I cut a new one out of MDF for testing. Now that I know it's okay I'll have to cut a new one out of PVC. I'll keep the MDF one for the first CNC64 clone.

I'm only missing the motor mounts, delrin/acetal nuts and some paint coats on the last MDF panels, otherwise both the X-axis and Z-axis are finished. I plan on manually cutting straight lines just to test the power of the Black & Decker RTX with the Dremel router bit #650.

I'm not sure if I already mentioned it, but in case someone is wondering, I swapped the chucks of the Dremel 395 and Black & Decker RTX.

[floriske.nl]

PFf, respect for this build! Takes a lot of patience this way (building a cnc to build a cnc that is)

I hope you get enough accuracy out of this machine to build the new one with even better accuracy.

I myself thought about building myself out of extrusion profiles at first:

https://dl.dropbox.com/u/72592758/cnc/Proef%203.jpg

But after calculating the costs I came in contact with Stepmores in China and most likely I'll be buying a SM-6090 of them for less instead:

http://i45.tinypic.com/290uagi.jpg

[jammin0]

PFf, respect for this build! Takes a lot of patience this way (building a cnc to build a cnc that is)

I hope you get enough accuracy out of this machine to build the new one with even better accuracy.

I myself thought about building myself out of extrusion profiles at first:

https://dl.dropbox.com/u/72592758/cnc/Proef%203.jpg

But after calculating the costs I came in contact with Stepmores in China and most likely I'll be buying a SM-6090 of them for less instead:

http://i45.tinypic.com/290uagi.jpg

What price on that chinese cutter?  I got a quote for about $2500 US.

[floriske.nl]

Please don't let this topic end up being about Chinese routers  I know how these things can get out of hand quickly 

But just to answer your question: $ 2350 for the base version - $ 100 for usb control - $ 100 for a water sink - $ 100 for a spray cooling system. Comes with standard led light, limit switches and even a free calibrator if you ask for it. (Prices are FOB, exclude tax, import duties and shipping)

Now back to Yvan256's build 

[jammin0]

Right you are.  Back to the regular scheduled programming.

[Yvan256]

Back to the regular scheduled programming.

Tonight at 11:00, we discuss the lack of news on Yvan256's CNC64 project.

[Yvan256]

Instead of trying to cut nuts from an HDPE cutting board, I'll first try simple t-nuts. If it doesn't work for a CNC maybe it will work for something else later...

First two photos are the t-nuts, last two photos are the temporary control circuit. There's a lot on the protoboard, only the Atmel microcontroller on the upper right and the four IRLZ14 logic-level power MOSFETs in the bottom left corners are used for one motor. I plan on milling the final board with CNC64 itself.

[jammin0]

Dang.  I love how you're building the controllers yourself.  It's gotta be pretty rewarding.  Very cool.

[Yvan256]

Dang.  I love how you're building the controllers yourself.  It's gotta be pretty rewarding.  Very cool.

The main controller should be an ATmega328 running grbl, but I plan on trying to code the stepper drivers from an ATtiny84 and four IRLZ14 power MOSFETs for each motor. The goal is a lower cost and a tougher driver that's more immune to overheating.

[Yvan256]

I've been working on the prototype for the stepper drivers. An ATtiny84 will control a stepper motor via four IRLZ14. Only a few parts and wires left do to on each ATtiny84/IRLZ14's group. As you can see, I've cleaned up the wiring a bit.

After that, I need to code a small test program to see if all three motors are being controlled properly, then I need to program the actual code to convert the direction/pulse signals from the ATmega328 running grbl.

It doesn't matter if it's a CNC mill or an arcade cabinet, these builds are 50% hardware and 50% software.

[Yvan256]

Warning: this post is heavy on technical details, only has one photo and involves mathematics!

(insert "Get on with it!" clip here)

After I had installed everything (motor, threaded rod and stepper driver), I had problems with the torque... I knew that a surplus stepper motor that only cost four dollars wasn't going to deliver extreme torque and speed, but I was completely disappointed. The motor was missing steps when trying to move the X-axis without any force applied to the cutting tool!

Then I looked at my AVR code again. I was driving the motor in wave drive mode instead of full step drive. 

In full step drive, the torque is better and I'm pretty sure that once I apply dry wax on the threaded rod it's going to be better. Hopefully that metal-on-metal grinding noise will also go away. 

The motor is only 48 steps per rotation, but combined with the 1/4-20 threaded rod it means 960 steps per inch, or 26.4583333 microns. That translate to 0.0264583333 millimeter per step, compared to the 0.01 millimeter per step of my Roland MDX-3, or around 38% of the precision. That may seem low, but when you consider the cost of CNC64 vs the cost of the MDX-3 (even its cost when bought used on eBay), it's still amazing.

And if that's not precise enough, I just found out that putting vinyl tubing on an 8-32 threaded rod allows it to fit inside the fuel line motor coupling. That would mean 48 steps * 32 = 1536 steps per inch, 16.5364583 microns or 0.0165364583 millimeters, 60% as precise as the MDX-3. Of course, the higher the thread count, the slower the machine will be.

I could also "overvolt" the motor and limit the current, but then we get into costlier electronics. When you consider that in my setup, the cost for a stepper motor and a stepper driver is below 10$, it's hard to beat. But you also need to accept the limitations.

I'm still not sure if it's going to be enough for the Z-axis, I may have to go with the 8-32 rod for that. But the speed of the vertical axis is not much concern for me since my designs are mostly "2D".

[Yvan256]

Yesterday I cut two replacement panels for the machine. These new versions have holes for a T-nut and two bolts to hold it in place. I removed the "teeth" of the T-nuts with pliers and flattened the edges in a bench vise. These modifications allow the T-nuts to sit completely flat with the material. The small recessed edge of the middle hole is to compensate for the curved edge of the exterior of the T-nut.

Photos of the T-nuts in the new panels for the X-axis and the Z-axis. The Y-axis is lagging behind because I have not yet finished the design.

I do know that with simple T-nuts, I'm going to get backlash. Will it be small enough that I can ignore it? I'm not sure yet. From what I've read, the best way to test backlash is to cut a circle. I plan on doing a ballpoint pen holder so that I can do tests without cutting any material. It will be especially useful for testing printed milled circuit boards. My guess is that "cutting" a circle with a ballpoint pen will make it easier to check for backlash.

[Unstupid]

I do know that with simple T-nuts, I'm going to get backlash. Will it be small enough that I can ignore it? I'm not sure yet. From what I've read, the best way to test backlash is to cut a circle. I plan on doing a ballpoint pen holder so that I can do tests without cutting any material. It will be especially useful for testing printed milled circuit boards. My guess is that "cutting" a circle with a ballpoint pen will make it easier to check for backlash.

The best way to test is to CUT a circle not draw one.... you wont be subjecting your rig to any torque with a ball point pen.

[PL1]

I removed the "teeth" of the T-nuts with pliers and flattened the edges in a bench vise.

Why not get this type that use wire brads through the flange holes to prevent them from twisting?  Lowes carries them.




Scott

[Yvan256]

Why not get this type that use wire brads through the flange holes to prevent them from twisting?  Lowes carries them.



Scott

I did try to find these nuts locally, without any luck. That's why I went with hacked T-nuts instead.

Also, there's no Lowes in Québec.

[Yvan256]

The best way to test is to CUT a circle not draw one.... you wont be subjecting your rig to any torque with a ball point pen.

Indeed, but if it can't draw a circle with a ball point pen I expect it won't be able to cut one either. I wouldn't limit myself to drawing one, it would only be the first test. I also need an excuse to make a ball point pen attachment.

edit: I'm not sure how to merge my last two replies into a single one. 

[PL1]

there's no Lowes in Québec.

Amazon has them, too.   


Scott

[Unstupid]

http://www.widgetworksunlimited.com/  for the plotter pen... also felt > ball point! 

[Typefighter01]

Install two T-nuts either back to back or one on either side of your "panel" (one on either side would probably look best, but I can't tell from the pics if there is enough depth before they bottom out on each other). This way you can just turn the backlash out by turning them in different directions and locking them down with the two bolts. DIY anti-backlash nut. Nice going BTW

[Yvan256]

Amazon has them, too.   

Scott

Amazon Canada doesn't have the same huge catalog as Amazon USA, though I haven't checked for that kind of part in particular. After all, my Dremel router bits do come from there after all. I'll need to talk to my friend to see if he's ordering more stuff from Amazon soon. I don't feel like paying 15 dollars in shipping for 2 dollar parts.

http://www.widgetworksunlimited.com/  for the plotter pen... also felt > ball point!

Yikes, 40 dollars for a plotter pen... that's nearly half of the cost for CNC64 so far! I'm going with a "hacking-style" plotter pen... BIC ball point pen with DIY spring system if I can't find one with a built-in one. And by felt, do you mean something like a Sharpie?

Install two T-nuts either back to back or one on either side of your "panel". This way you can just turn the backlash out by turning them in different directions and locking them down with the two bolts. DIY anti-backlash nut. Nice going BTW

Thanks! And strangely enough that's exactly what I was thinking today!

However, I still plan on trying home-made acetal/delrin/HDPE/etc plastic nuts, using a slotted threaded bolt as a home-made tap. At the very least, I should try to find T-nuts or similar in brass instead of whatever those current T-nuts are made of, the friction is just too much for those low-power motors.

edit: I just did some tests with the Z-axis and I'm quite certain that without the friction from the T-nuts, they would be powerful enough at a lower speed.

[Yvan256]

The Z-axis was having a hard time, so in the meantime here's a video for the X-axis movement.

I'm not sure how to embed a video from Vimeo directly in my post, so here's the link:
http://vimeo.com/user17022509/cnc64-2013-03-12

[Yvan256]

Yesterday I was able to cut the four panels required for the new design of the Y-axis assembly of the machine. Even though the old design already had six panels cut, the design was getting too complex for no good reason.

I was also able to cut yet another replacement panel, this time the reason being that I had forgotten about the end stop microswitches. To make it a bit more arcade-related, I'm using D44 Cherry microswitches. I still need to replace a few other panels again to add the microswitch mounting holes, cut those 8mm rods and paint a few panels, but otherwise the project is nearing completion. Then I'm back to my CNC256 project so that finally after that one I can get back to my arcade-related builds. 

Also, as you can see, the new panel with "CNC64" and the top two panels of the Y-axis assembly have rounded corners but unlike all the previous panels with rounded corners, this time it's one of the new function of my "Dirty Easy CNC" which made them for me. I've made a function that can output different radiuses for each four corners. I just wish I hadn't lost nearly two hours because PHP's sin and cos functions take radians instead of degrees as input... 

[Yvan256]

Not much of an update for now. The two new top panels are cut, painted and installed, with their Cherry microswitch. The machine now has both home and end stop switches on its X-axis. I've also painted and assembled the four panels of the Y-axis assembly.

I also finally ordered the router bit to be able to cut the T-slots (16J60.72) into the spoilboard that will go on top of the Y-axis assembly, along with a high-speed steel router bit (86J01.01) so that I can try cutting aluminium on the MDX-3. Yes, that small machine takes 1/4" router bits... crazy, right?

The reason I'm going with slots cut directly into the spoilboard is because it makes the whole thing easier and lighter and there's no danger if the router bit ever hits the slots, unlike aluminium rail slots. It also means I can cut many more spoilboard with T-slots, for example a new spoilboard for the MDX-3, the first CNC64, the CNC64 clone, CNC256, etc. With aluminium rails you need to buy rails every time you make something new.

I also ordered T-slot nuts and bolts to use with the rails of all the machines and a bushing adapter so that I can use Dremel bits on the MDX-3. I smell milled circuit boards in the near future.

The last photo is something completely off-topic, as if posting the build of a CNC machine on an arcade forum wasn't already off-topic enough (hey, the home and end stop switches are arcade-grade Cherry microswitches, what more do you want?). I was getting tired of the fact that the batteries in my calliper box shuffled around so I cut a little PVC holder for them. 

[Yvan256]

I received my parts from Lee Valley and cut two T-slot beds. I also made a new Y-axis motor mount plate since my small, used NEMA17 motors seem to be a bit stronger than the bigger "tin can" stepper motors.

If you like MDF dust, you really should get one of those T-slot cutting bit! 

In the next few days I plan to cut a new MDF bed with T-slots for my Roland MDX-3 as well.

I just thought about cutting T-slots into the bed for CNC256... that's going to make a boatload of dust. 

[Yvan256]

Funny how things turned out. I wanted to complete the Y-axis assembly before going any further, to test with the step motors and all... and it's practically the last thing to get finished.

Yesterday I cut a new replacement side panel for my new motor mount plate (that's why it's missing in the first two photos), one motor mount plate and also cut six printer/scanner rods. Cutting those rods is hard, it's eaten almost a quarter of a Dremel cutting wheel to cut six ends! I've also upgraded the Roland MDX-3 to a new T-slot bed. It's so much easier to clamp boards down now!   

Here's a few photos to go with this update. The first two are the Y-axis assembly without and with its T-slot board.

The third photo is a PVC mounting panel for my used NEMA17 motors and room for the stepper driver board. These motors don't have threads so the 3mm bolts need to go all the way through the motors and I can't find bolts long enough locally to go through both the motor and the MDF. The wires of the motors are also very short so I'd rather have the stepper driver board right next to them.

The space for the drivers is huge because I made the schematic for 2mm traces, just to be sure I can't mess them up while milling them and because I'm still coding everything by hand, so square uniform traces are easier to do. The 4th image is a Sketchup screen capture of that board. Top drawing is the board viewed from the top with most of the components in place, bottom drawing is the bottom of the board with the traces. I could probably have gone with smaller traces in some areas, but I made a function in PHP to have grid-aligned holes and traces so that's the result.

[Yvan256]

I was planning on using the Roland MDX-3 to mill the circuit board, however the 1/8-to-1/4 adapter just doesn't work with the press-fit spindle of the MDX and I can't cut circuit traces and drill IC holes with a 1/8" router bit...

So I had to turn, once again, to laser-printed circuit boards. Here's a photo of my fifth attempt, the only one successful after four failures.

I read a good tip on YouTube:
1. heat up the copper clad board (CCB)
2. instead of going table, CCB, paper, iron like most people do, go table, paper, CCB, iron (with the copper facing down on the paper with the toner-printed circuit, of course)
3. heat up CCB again, pressing everywhere
4. reverse, iron evenly all over the circuit
5. soak in hot water

This reverse process made the paper remove itself completely after a few minutes, and zero toner was left on it, it all transferred to the copper! 

I was a bit rough on the traces afterward, that's why there's a small bit missing in a middle trace.

edit: second photo shows the PCB with the traces etched and cleaned, third photo is the PCB with all the components minus the power and signals connectors and the jumper wires. I still need to cut the PCB outline and solder the components, otherwise it's complete.

[rablack97]

I etch boards myself....

I'm confused on the heat up CCB......

Heat it up how.....

Do you heat it up then iron?

Heat it up again? How?

Then when you say reverse, do you mean flip it over and iron the paper side?


If you use good printer paper and have a good laser printer, the toner will come off the paper no issues...

The biggest mistake folks make in this method is either the iron isn't hot enough or they are too impatient and dont iron long enough.

One sign i use is if i can see the traces through the photo paper when ironing, or the paper starts turning a little black.

Then i soak in cold water, and the paper peels right off.

Next time use a little soap and light scouring pad, to remove the rest of the paper off, rub lightly and be patient it will come off....

[Yvan256]

I etch boards myself.... I'm confused on the heat up CCB...... Heat it up how..... Do you heat it up then iron? Heat it up again? How? Then when you say reverse, do you mean flip it over and iron the paper side?

The usual method is to put the CCB on the table (copper face up) and put the photo paper (face down) on it and then iron the whole thing out.

The method I picked up from a comment on YouTube said to put the photo paper (face up) on the table and then drop the heated CCB (copper face down) on it, heat up a bit again on the non-copper side, then flip over and just iron out as usual. I've found that it leads to better and faster results.  I still have two boards to do, so maybe I'll try this method again and the non-reverse method, keeping all other steps as similar as possible and compare the results.

If you use good printer paper and have a good laser printer, the toner will come off the paper no issues...

The biggest mistake folks make in this method is either the iron isn't hot enough or they are too impatient and dont iron long enough.

One sign i use is if i can see the traces through the photo paper when ironing, or the paper starts turning a little black.

Then i soak in cold water, and the paper peels right off.

Next time use a little soap and light scouring pad, to remove the rest of the paper off, rub lightly and be patient it will come off....

I'm using Epson inkjet glossy photo paper (ran out of Staples inkjet photo paper), but my fifth attempt gave good results. My iron is at maximum, maybe I was too impatient on my first four attempts.

Thanks for the tip about watching the traces through the paper while ironing and for cleaning the board. Maybe my next two boards won't have damaged traces.

Do you have a trick for cutting out the boards? Scoring with a knife takes forever. 

[rablack97]

For cutting the boards, you can use these...Once the copper is gone, the material is easy to cut. Although it also depend on the material your core is made out of.  Some cut easier than most, i've had no issues with the paper cutter though....



Or my favorite if you can find one cheap is this....



I'm looking at one of these in the future though....step it up a notch...



I'm still not following, the heat it up part....How are you heating it up before ironing....

Honestly, patience is the key, you can flip it upside sideways....etc doesn't matter.  Ironing from the underside isnt really transferring heat any better, just seems like wasted energy.

This is what i do..

Tape photo paper to copper
Place copper on thin towel, to preserve table (Mrs. kills you if you dont)
Place iron directly on photo paper, hold for about 30 secs pressing down hard (this is the key pressing down hard and even)
Never lift the iron off the board, and use the tip to hit all of the edges
Stop every 30 seconds or so and press down hard....
When the paper starts turning black or you start really seeing the outlines of the traces then you should be good


Also, get you some liquid tin.....it tins your traces and protects them, and it makes for easier soldering.....

[Yvan256]

I have similar cutting scissors but it just cracks the edge of the board while cutting, but I do have a slicing paper cutter, so I'll try that!

A tile cutter would be nice but as soon as CNC64 is completed and working, I'll be making the boards with it. Holes with drill bits, traces with specialized V-carve bits and board outline with a "chip breaker" bit. I've already tried the chip breaker bit on a random piece of PCB and the result is extremely clean, almost like a commercial board. Being controlled by a CNC means complex board shapes will also be possible.

Thanks again for the tips, I'll see if I can find liquid tin.

[rablack97]

I have similar cutting scissors but it just cracks the edge of the board while cutting, but I do have a slicing paper cutter, so I'll try that!

Yep that's the material i'm referring too that's more gummy...There is a material that cuts much easier....

The paper cutter should help you out...or if you have a band saw, i've had luck with those too....

[Yvan256]

I'm still not following, the heat it up part....How are you heating it up before ironing....

Sorry I forgot to answer that question. It's just about heating up the copper board with the iron before applying the photo paper to it.

[Yvan256]

First off... wow, over ten thousand views! 

You guys have even more patience than me!   


Well, back to the topic I guess... When you use surplus parts, you don't really get to choose much of anything as far as specifications are concerned.

My NEMA17 motors don't have threaded holes, so I had to figure out a way to mount them and the only local store which had bolts to fit the mounting holes didn't even have bolts that were long enough to go through the motor and the MDF.

About a month ago, the local sign store gave me some thinner expanded PVC panels and it was the perfect thickness for the length of the bolts, so here is two of the motors installed on their mounting plate, bolted to the CNC machine. I'm only missing couplers and I have two axes ready to go! Well, except that I need to make more driver boards.

Speaking of driver boards, the last photo is of my custom stepper driver board which I still need to cut properly. But guess what? I made two mistakes in the design. I forgot one trace for the ground of the ATtiny84 when powered from the power connector and I forgot a ground line for the step/direction communication pins. The first mistake would be easy to patch with a simple wire underneath the board but the other mistake would require a weird split connector but I do have a ground pin on the ISP header for the ATtiny84...

Considering the cost (around 8$CAD in parts) and the trouble of making the board versus buying a stepper driver on eBay, I'm not sure even the first CNC64 clone will be using my custom boards. The drivers on eBay have gotten extremely cheap, there's a lot to choose from at around 6~12$ so it's not even double the price of a home-made solution.

Of course the cheaper drivers are weaker than my IRLZ14, however those small NEMA17 motors aren't exactly powerful either. Even the all-in-one boards (from parallel port to stepper drivers) are in the 40~50$CAD range now. With the popularity of 3D printers, everything seems to have come down in price over the last few years.

[Yvan256]

I just finished the second stepper driver. The board is much cleaner, there's no missing traces, all connections are there, etc.

Nothing exciting for now, but once the third stepper driver is complete, I'll be able to do some tests with something like TurboCNC!

[Yvan256]

I wrote the code for the ATtiny84 used in my low-cost stepper driver. Nothing fancy, it's a full-step driver, no micro-stepping or current limitation here. The code is only 220 bytes, however, so even an ATtiny24 would only be 11% full. I'm using ATtiny84's because that's what I have on hand. They're socketed, though, so I'll be able to replace them with ATtiny24's the next time I order parts.

I'll be posting the schematic, a PDF for toner-transfer (single side PCB with only 6 wire jumpers on the top), parts list, source code and compiled code in the next few days.

In the meantime, it's fun to finally see more than one motor turn, I have an ATmega88 driving two motors for testing purposes. The next step will be a pre-programmed part (a CNC64 part, that is) and the next step after that, probably TurboCNC talking to those stepper drivers.

edit: forget that, the prices on eBay are so low right now that you can almost buy two L298N driver boards already assembled for the cost of one of my own driver. 

[Yvan256]

I'm still having problems with the Z-axis. There's too much friction on the threaded rod and the whole thing may be too heavy for the little surplus stepper motor.

The X and Y axis, however, seem to work fine even with the fuel line couplers holding in place without anything but friction.

So I re-traced the ATtiny24 stepper driver board to only have square lines everywhere and I hand-coded the paths required in the ATmega88. But I forgot to mirror the circuit before doing all of this. I'm glad I tested on a white piece of masonite first!

There's a few problems aside from the mirror image. My cutting bit is kind of big to cut circuits, some traces didn't cut "all the way through" because the masonite is probably not flat enough, a corner simply teared off the surface and I calculated everything for 18 turns per inch instead of 20. The weird-looking rectangle at the upper left is when I thought my stepper motors were 48~50 steps per rotation, turns out they're 200!

But it's still a nice first cut anyway in my opinion! 

[Yvan256]

Here's my third attempt at cutting the circuit board for the ATtiny24 stepper driver. The second attempt was the same exact cut but in a white hardboard.

The copper board is warped a bit on the edge, which is why the leftmost trace isn't cut all the way through. I'll try double-sided tape for the next cut instead of holding it in place with bolts and washers.

Because my Z-axis isn't done yet, I have to cut traces by going in a single line, so the bit often has to go backward in traces already cut. You can clearly see the backlash on some of the traces (look at the front of the filled "P" at the upper right), I guess I'll have to make those delrin nuts after all. 

If it had been a CNC64 part cut in MDF, I probably would never have noticed the backlash.

edit: my double-sided tape sucks, it can't hold the copper board to keep it flat.
edit2: nope, double-sided tape is fine after all, I was just being a bit stupid earlier...
edit3: double-sided is more than fine, it's actually extremely hard to remove from the polycarbonate plate after the CNC is done cutting.

[Yvan256]

So, it's been a month since my last update and some people might be wondering why.

First, as a follow-up on my last few posts about milling circuit boards. Last week, I've read about OSHPark on SparkFun. And with the pricing of OSH Park including free shipping to Canada,  I don't see myself ever doing another PCB/MCB ever again unless I absolutely need a board as quickly as possible. Even then, I'll simply use the laser-printed sheet+ironing solution. I've read that a solution of regular table salt, oxygen peroxide and vinegar can be used to dissolve the copper so I'm not stressed about finding ferric chloride locally either.

Secondly, and that's the biggest problem, is that the spindle motor on my MDX-3 was getting too slow and could barely cut anything anymore. The prospect of having to pay over 80$ to fix a 10$ motor wasn't at all appealing and I always hated the "ratcheting motorcycle" noise the motor was doing because of the way it was using friction on the spindle to make it turn.

Long story short: I've fixed and upgraded the spindle motor myself. The funniest part is how I did it.

The result sure isn't pretty, but I've fixed it with:
- one bolt
- part of a metal plate from an old HP printer
- a "12VDC PCB drill motor" from eBay (under 4$CAD, free shipping)
- three LEGO parts*! 

* Yellow Band
* small pulley
* Cut-up yellow plastic tubing to fit the pulley on the motor shaft (can't find the part number)

The best part of that story is the MDX-3 spindle is now belt-driven (a bit more quiet than before) and seems to be about two or three times as powerful than before, if not more. And it's not because it's a 12VDC motor vs the original 5VDC motor because I need to power it with only 5VDC too otherwise it's way too fast and noisy!

And if the need ever arise, I can replace the spindle motor for about 3$ and I already have a spare one in stock.

Attached are photos of the hilarious fix, which includes a piece of expanded PCV I used to hold the spindle in place to replace the original little metal plate. I'm sure the Roland company would not be proud. Who knows, maybe they'll be embarrassed enough to send me a free MDX-15 (wink, wink)!

edit: as a test run, the MDX-3 is cutting a single 24mm hole in 10mm MDF as I'm writing this. The noise level is much, much lower than before.

edit2: the 24mm hole in MDF was successful, so now it's cutting hard plastic for the first time. The same 24mm hole, in a CD-R.

[Typefighter01]

The best part of that story is the MDX-3 spindle is now belt-driven (a bit more quiet than before) and seems to be about two or three times as powerful than before, if not more. And it's not because it's a 12VDC motor vs the original 5VDC motor because I need to power it with only 5VDC too otherwise it's way too fast and noisy!

Looks like you created some torque mutiplication with your pulley diameters...nice upgrade

[Yvan256]

I tried to cut a new panel for CNC64, but it seems my "repair" didn't work as well as I thought. 

This situation is a bit frustrating, because MDX-3 has a working X/Y/Z table but no spindle and CNC64 has a perfectly good spindle (the Black & Decker RTX) but is still missing a working Z axis.

edit: After careful verification it seems the LEGO silicon band isn't up to the task of transmitting the power to the spindle. I've hacked the original spindle motor assembly back into place.

[Yvan256]

I've fixed the MDX-3 (again) and cut the first two replacement panels for CNC64. The threaded rod would give amazing precision but there's too many drawbacks: threaded rod/nut friction, backlash and slow speed.

So I've decided to switch to a belt drive system. After all that's what the MDX-3 uses and it's cut quite a number of MDF panels so far. A while back I went to the local recycling centre and got two very long MXL belts for free. I got more than enough for both the X and Y axis. My surplus motors already have MXL pulleys on them, so it's a perfect match.

One of the upshot of switching to a belt drive for X and Y is that it will now be fast enough for other applications. The Z axis is designed so that I can attach various things to it, the first one being the Black and Decker RTX and the second would be a 3D printer extruder. I'll probably try with a glue gun extruder first just to keep the cost down and to try an automated "rotating barrel" system for auto-loading the glue sticks.

Here's some photos of the new panels for the belt drive with another NEMA17 motor and a S2M belt, I'll be using my old beige/black ones with the larger MXL belts for CNC64. I just need to find longer 3mm bolts to attach the motors to these new panels.

[Yvan256]

The MDX-3's spindle is no more, my hack didn't hold. I'll try to put back the belt back in place but it's a mess for the mounting holes right now. It would almost be easier to build a brand new Z-axis platform.

On the bright side, the Z-axis of CNC64 is now working if I keep the speed extremely slow. Not much of a problem for the Z-axis, in my opinion.

I still need to make those acetal/delrin/HDPE nuts, I may have a little time this week so stay tuned.

[Yvan256]

Since my MDX-3 can't cut new parts to convert CNC64 to belt drive, I have to make it work with my 1/4-20 threaded rods.

So I spent the last two hours or so cutting thin slices of yellow delrin and doing threads in them. Strangely enough, the best result I got was drilling 1/4" holes to make 1/4-20 threads otherwise there was too much friction and I was afraid my little NEMA17 motors wouldn't be up to the task. I still need to cut them into nice rectangles and drill holes to attach them to the MDF panels.

The friction is so low, I'll probably be able to speed up my Z-axis by at least 25% if not more.

edit: nope, 1/4" holes wasn't a good idea. I stripped the threads quite easily just trying to push it manually. Good thing I didn't waste time trying to install them on the machine. It looks like it's going to be nylon nuts for the time being, unless I can repair the MDX-3 to finish cutting the new panels for the belt drive.

[Yvan256]

The MDX-3 has been repaired once more and is back at work cutting parts for CNC64. Threaded rods are too much trouble for my low-power motors, I'm definitely switching to belt drive, including the Z-axis if I can.   

Sorry, no photo for now.

I think I'm going to do a micro-cabinet that can fit in the MDX-3 cutting area, getting tired of posting about non-arcade projects. 

edit: the MDX-3 has cut four small parts and one medium part. I'm calling the repair a success.

[Yvan256]

I've been cutting a second copy of CNC64 for a friend, in exchange for a stepper motor driver controller board.

We're still waiting for delivery but it does open up the option of going with threaded rods right now with full/half stepping and upgrade to belts later with 1/16 micro-stepping and a few new CNC64 panels. It will also drive the steppers in bipolar mode, which makes them stronger than unipolar.

And since we're going to need remotes for the first threaded rod versions, I've been busy cutting panels for my MVS-style CNC remote. It's designed to use the popular Grayhill 04A-B01 to control the X and Y axis. I've even planned for a tiny 5x7 dots LED display in the "credits" area to be able to display either two digits or scrolling text. A 5x7 dots display on its side becomes 7x5, which is enough for two digits with a 3x5 font and will also allow "smooth" text scrolling. I used the control panel of MVS-99-6 as a starting point, that's why there is three white buttons.

The usual red, yellow, green, blue, player start, main menu and pause game buttons will be replaced with CNC remote functions (Z up, Z down, panic stop, etc). I'm also used to the tiny control panel of my MDX-3 so I'm probably going to copy some functions from it (i.e. start position and view position).

[Typefighter01]

Loving the MVS-style remote...too cool!!!!

[Yvan256]

Had to cut the back and front panels again because I forgot the side tabs the first time around. 

The MDX-3 is busy cutting the top panel, which is also the last one. I'm leaving the bottom open to have access to the buttons, etc. With no bottom I'll be able to glue everything together with carpenter's glue, no holes or bolts to ruin the look.

But it's going to take a while to cut that panel, with all those holes and pockets.

edit: it took 1 hour and 30 minutes to cut all the holes and it's probably going to take another hour or so to cut the outline of the panel.

[Yvan256]

It's done! All I need to do now is to round up the front edge, order red and black vinyl, try to find 3/8" black T-molding, order a printed vinyl overlay for the top, order the joystick and the buttons, make a small board with a microcontroller to emulate a USB keyboard...

Ok, it's not really "done" yet. But at least I'm done cutting panels for this mini-project.   

The first photo shows the little joystick retainer panel, the other two shows the controller with the 5x7 LED display. I'm glad I added 0.25mm padding all around otherwise it would not have fit in the hole.

edit: added a 4th photo with the edge of the top panel rounded at the front.

[Typefighter01]

Your remote has to be "done" now, no ?

What microcontroller are you using for the usb keyboard emulation?

[Yvan256]

Your remote has to be "done" now, no ?

What microcontroller are you using for the usb keyboard emulation?

I don't have that tiny joystick and I don't want to pay 8$ shipping for the joystick alone, so the remote is on hold for now. I also don't have small buttons for 10mm holes, the best "arcade-looking" buttons I can find on eBay (in white, red, yellow, green and blue) are 12mm, so I might need to cut the remote again anyway. I can make it maybe 10-20% bigger than it is right now.

In other news, I'm building another CNC64 for a friend and in exchange he's ordering some parts in double, one kit for him and one kit for me. That kit includes a newer model of those popular Chinese 3-axis stepper driver, the TB6560-3V2. I have to say, I'm really impressed for the price which was about 55$CAD each, shipped. It's a new version with big ceramic resistors, I had zero problems using it so far, I highly recommend it. Each axis can be setup for 1/1, 1/2, 1/8 or 1/16 micro-stepping. I'll be ordering a second board for my bigger CNC256 machine.

I've been testing my 103H5208 stepper motors that I got from the recycling place and I just can't believe how strong these tiny motors can be. If you tried to hold their metal pulleys while it's spinning you'd probably rip the flesh from your fingers. They're scary strong.

And since a slow CNC is a boring CNC, and since those 103H5208 motors already have tiny S2M metal pulleys on them, I decided to go with a belt drive after all. I should have the Y axis finished before friday. And because the motors are really strong, I've also decided to go with a belt-driven Z-axis as well. Cutting 3D plaques will be lightning-fast or at least as fast as the Black & Decker RTX can cut them. It sure won't be the CNC itself that will be too slow.

For now I'll be testing with MXL belts that I got from old printers and scanners, because my S2M belts are still on their way from China.

Enough with the boring life story, here's some photos! As you can see, the way I mount the motor for a belt drive got reinforced from the last time I posted a photo of that setup.

edit: I added an image from the eBay auction comparing the older vs newer driver board.

[Yvan256]

I just ordered twelve pushbuttons, 12mm, with the proper Neo-Geo MVS colors. I got mine from a different seller, but it's the same buttons as shown in the photo below.

I should be getting 2 red, 2 yellow, 2 green, 2 blue, 3 white and 1 black buttons, if the seller reads the instructions attached to the eBay auction and/or the eBay message I sent to him directly after buying an auction for 12 black buttons.

I'll have to cut a new, slightly bigger controller for my CNC since the one I cut earlier was made for 10mm buttons , so if anyone is interested in the one I made a few posts ago, send me a PM and we'll discuss dimensions (precise Sketchup model available), price/trade and shipping.

If these buttons are good enough, I'll be using them in a future nano-MVS project.

edit 1: my special order of mixed colours was confirmed via email by the seller

edit 2: I bought and cut another 4x8', 3/8" thick MDF sheet today. Luckily the place where I bought it let me use their wall saw to cut all my stupid little 7x5 panels, see photo below. The floor at the bottom of that saw probably had a kilogram of MDF dust in the end.   

I now have (more than) enough panels to continue working on my belt drive modifications for CNC64 and for CNC256.

edit 3: the remote only needs to be 10% bigger to accommodate for the size of the new buttons.

edit 4: Digi-Key just increased the price of the 04A-B01 Grayhill joystick from around 16$ to 26$CAD, anyone know other sources for Canadian buyers? 

[PL1]

edit 4: Digi-Key just increased the price of the 04A-B01 Grayhill joystick from around 16$ to 26$CAD, anyone know other sources for Canadian buyers? 

Allied Electronics has the Grayhill 04A-B01 joystick here -- two thumbs up for this company and their outstanding customer service.   

According to their shipping page you can select CANCON (Canadian Direct Ship) or USPS shipping.
EDIT: Shipping page info outdated.    Sorry, Yvan56.


Scott

[Yvan256]

Allied Electronics has the Grayhill 04A-B01 joystick here -- two thumbs up for this company and their outstanding customer service.   

According to their shipping page you can select CANCON (Canadian Direct Ship) or USPS shipping.

Scott

Their website is not up-to-date, here's the reply I got from Allied Electronics:

As indicated on our website, UPS WW Saver is the most cost-effective shipping method. It is $19 and includes overnight shipping as well as all customs and brokerage fees. It is the equivalent of our old Cancon method.
 
DHL and USPS have been removed as shipping options for Canada.

So either I save 10$ on the joystick but pay 19$ for the shipping and hope there really isn't going to be any fees from UPS (I wouldn't count on it, last time I used UPS there was 30$+ in brokerage fees) or I pay 10$ more on the joystick and only pay 8$ for shipping, knowing I'd get it this week with no extra shipping fees. In either case, it's way too expensive considering the item. I could get a Seimitsu or a Sanwa joystick for that price less than that.

So instead of paying insane prices for what amounts to a cheap version of a joystick, I'm looking into continuing my own micro-joystick project, which would look like a smaller version of a Seimitsu LS-32. I still have enough 2.8mm thick, expanded PVC panels for such a small project and I already have the small microswitches. The expanded PVC should be strong enough at this size. It will make a nice side project and if enough people are interested I could start selling parts for them, including tiny custom-cut, 3/8" thick MDF control panels.

[Yvan256]

Today I received the S2M belt I ordered from eBay on august 26th. It only took two weeks to get something from Hong Kong. And no shipping fees, no duties, no problems.

I'm a dumbass, however, since one meter will only be enough for either the X or Y axis. I'll order two more meters as soon as I got one axis working with it.

The top of the Z-axis is ready for the belt drive, as you can see in the second photo. I'm using S2M belts because it would be pointless to remove the perfectly good metal pulleys from my three 103H5208 stepper motors.

[Yvan256]

Those LM8UU linear ball bearings are of really low quality. 

That is all.

edit: turns out the LM8UUs are fine, it's the rods I'm finding in printers and scanners which are a problem. Some are 7.99mm, others are 8.02mm, etc. I found that out while trying to fit the bronze bushings over some rods and measured them again with my digital calipers. 

[Yvan256]

Yesterday I ordered self-aligning bronze bushings from TechPaladin Printing.

I also have two CNC-related surprises in progress, I'll post about it once I have more than random painted panels to show. 

[Typefighter01]

  Can't wait 

[Yvan256]

Alright then, here's a teaser. A new pile of panels and my low-cost thien dust separator.

[Typefighter01]

The scale of that dust separator should work great with the size of CNC you are running. Looking good.

[Locke141]

Looks good.

I am loving all the DYI CNC builds on here.

I just got a small CNC to work at my job. They are a lot of fun.

[Yvan256]

Today I finally received the buttons. I'm not complaining about the shipping time since shipping was free.

The problem is, somebody made a mistake when I asked about getting coloured buttons instead of black buttons as stated in the eBay auction.

Here are the 12 buttons. Find the mistake.

 

[Slippyblade]

Did they seriously send you 12 colored AND 12 black?  For the same price?  Niiiiice.

[Rick]

Did they seriously send you 12 colored AND 12 black?  For the same price?  Niiiiice.

Look closer! Obviously, they included a free keyboard! SCORE!

[Nephasth]

Did they seriously send you 12 colored AND 12 black?  For the same price?  Niiiiice.

Happens a lot with Chinese vendors. I ordered around 20 momentary buttons like these, but they sent me 20 locking buttons instead. So I emailed them about it, and they sent me 20 momentary buttons for no additional charge and I didn't have to ship the others back. I've had this happen with other various electrical parts, like LEDs, too. It's a pain to wait sometimes, but it's always cheap, and they almost always make things right. YMMV.

[Rick]

Happens a lot with Chinese vendors.

E-Bay, Alibaba or other?

[PL1]

Did they seriously send you 12 colored AND 12 black?  For the same price?  Niiiiice.

Happens a lot with Chinese vendors. I ordered around 20 momentary buttons like these, but they sent me 20 locking buttons instead. So I emailed them about it, and they sent me 20 momentary buttons for no additional charge and I didn't have to ship the others back. I've had this happen with other various electrical parts, like LEDs, too. It's a pain to wait sometimes, but it's always cheap, and they almost always make things right. YMMV.

Same thing happened during the KADESTICK build.

Ordered 2 position DPDT On-On rocker switches but uxcell (the Hong Kong vendor) accidently shipped 3 position DPDT On-Off-On switches.
 
Once notified of the error, their top notch customer service team quickly sent replacement switches that got here several days before the delivery estimate on the original shipment.    


Scott

[Yvan256]

Bronze bushings

The bronze bushings from TechPaladin.com arrived yesterday. I now understand the "self-aligning" feature of those parts. They glide so smoothly on the printer/scanner rods that the LM8UU linear ball bearings seem really inferior by comparison.





New assembly plates

I was really tired of having bolts and nuts sticking out everywhere inside and outside the machine that I'm replacing the assembly plates with new ones that feature custom-cut countersunk holes for the bolt heads and the nuts. Yes, hexagonal holes. 









Threaded rod drive

Previously I talked about threaded rod drive versus belt drive, I tried making CNC64 bigger to be able to at least cut the control panel for MVS-99-6, etc. The problem is, the base of CNC64 is too low-profile to convert it into a belt drive CNC but I also didn't want it to go to waste. So I tested with a #6-32 threaded rod and nylon nuts and the result is nothing short of amazing. Using a nut on both sides of the moving platform, I can't feel any play and it turns extremely smooth (plastic on metal). I had to make tiny plates to hold the nylon nuts, using the same countersunk hole method as the new assembly plates.

So the goal is to complete CNC64 using this new method, along with its first copy. Oh, didn't I mention this already? There's another CNC64 machine that's nearly complete. It's not a clone since it was cut by the MDX-3, but it's a trade between me and my friend for the stepper driver board I mentioned in an earlier post.







A surprise

If the second CNC64 machine is not a surprise, then what is it, you ask? The surprise is what I currently call CNC86: a bigger CNC machine similar in design but with a belt drive instead of threaded rod. It's also bigger, big enough to cut all but five of the MVS-99-6 panels (sides, bottom, back, top). And those will be able to be cut in two passes, with the side panels being cut in bottom/top parts, assembled later. The cutting area is a bit bigger than 8x6 inches but I had to round down since "CNC 8.74016 6.37795" doesn't sound as catchy. I'm using inches for the names since otherwise it would be called "CNC222162" which is just plain bad (222x162mm cutting area, for those wondering).

Edit: how about CNC2016? (20x16 cm) But then I would need to rename CNC64 and call it CNC1510 instead...





There's one thing that I need to mention again: I hate painting. 

For those who wanted an update on the little MVS-style remote, don't worry. I'm just stuck at finding bolts locally to complete the custom mini-joystick that will be used inside it.

[Rick]

It all looks amazing! You should be extremely proud!

Now, onto the critical question. Because it looks like you're using MDF, and you've painted it to the point of expansion, do you think this is going to cause any issues regarding precision? I don't think it would cause an issue, but as a guy who's waffling between cutting up MDF and spending buttloads more on baltic or russian birch plywood, I figured I'd "ask the source".

Again - LOOKING GOOD!

[Yvan256]

It all looks amazing! You should be extremely proud!

Thank you, but the credit only goes for the design, my old Roland MDX-3 did the rest of the work. 

Now, onto the critical question. Because it looks like you're using MDF, and you've painted it to the point of expansion, do you think this is going to cause any issues regarding precision? I don't think it would cause an issue, but as a guy who's waffling between cutting up MDF and spending buttloads more on baltic or russian birch plywood, I figured I'd "ask the source".

Again - LOOKING GOOD!

I only applied 2-3 coats of acrylic paint on the panels, they haven't expanded one bit. If you're asking my opinion between MDF and baltic/russian birch plywood, let me tell you that once I have used all the little panels from the second MDF sheet that I bought, I will never buy MDF ever again. High-grade plywood is the way to go if you can afford it.

One thing you should check out is MDO plywood. I can't find that stuff locally but people who tried it seem to like it a lot. You can go on http://buildyourcnc.com to see what MDO plywood looks like.

[Yvan256]

Okay, so those bronze bushings really are precise. More precise than my MDX-3 is at aligning two holes, apparently.  The LM8UU didn't seem to complain about my rods alignment but the bronze bushings always seem to jam near the end points near the plates supporting them. I have checked my measurements three times and made a series of four test plates. Nothing seems to help. 

So I'm modifying my approach to have bronze bushings on one rod and Lucite plates with elongated holes on the other rod. After re-checking my old Roland MDX-3, it seems all its axes are build this way, bronze bushings on one rod and a simple two-pins plastic "fork" riding on the other rod. To avoid bending/torsion, the force is applied on the side with the bronze bushings. Most printers and scanners are also built this way, two bronze bushings on the belt side and the other side just rides directly on a little wheel or on plastic.

Since Lucite is an extremely hard plastic, it takes around 4 hours and 45 minutes for the MDX-3 to cut a single plate. It's going to take six of those plate. One down, five to go...

The edge is not clean at the bottom because my CAD program still lacks tabs.

Edit: I have now added "ramping cuts" for some functions of my CAD program, the MDX-3 can now cut one of those little Lucite plate in 30 minutes and one of those 2x2 holes MDF assembly plate in only 16 minutes.

Edit 2: ramping cuts are nice but cutting too much material puts stress on the structure of the MDX-3 and the cuts end up all wrong. Back to the old cutting speeds. 

[Yvan256]

Small update to my CNC64/CNC1510 project.

I made new Lucite plates with rectangular holes:



Going with rectangular holes seem to result in better precision (too much play in all axes of the MDX-3), I probably have about a quarter of the thickness of a paper sheet of lateral play with my new Lucite plates and the X-axis assembly slides very smoothly. The other axes are next but I want to complete the X-axis before continuing, so...


I also made motor/threaded rod couplers from expanded PVC:



The new couplers are great for three reasons, compared to my previous fuel hose and clamps couplers:
- they're more compact
- they're cheaper to make
- they look better

Since there is almost zero force needed to move with the bronze bushings + lucite plates combo driven by dual nylon nuts, I'll try to use the low-cost unipolar motors on CNC64/CNC1510 with my home-made IRZ14 stepper drivers.

I'll keep the fancy 3-axis micro-stepping board for my bigger, belt-driven CNC86/CNC2016.

[Yvan256]

Some panels still needs a coat of paint, otherwise the Z-axis is completed and 100% working. It's also using four 8mm bronze bushings. The MDX-3 isn't precise enough but it's always the same lack of precision, so after doing a few tests plus/minus 0.1mm of what I need, I can get usable parts that are aligned for the rods and bushings.

The new Z-axis is also a lot more compact and lighter than my previous one. The old surplus 4.00$CAD Japan Servo KP4M2-008 motor has no problem whatsoever moving the Black & Decker RTX up and down, so the other two motors won't have any problem with the X and Y axis either.

The green panels are discarded cuts of Sonopan sheets, given to me by a cousin. I've calculated that I have just enough to make a soundproof box for my belt-driven CNC. I'm still debating on how to do the actual box. I have enough hardboard but I don't think it would be strong enough for a box that size.

Now that the smaller and most problematic axis is done, I'll continue with the other two. I'll do the X-axis next since that's what the Z-axis is connected to. I already had the Y-axis working with lucite plates on one side to prevent binding/sticking though. I'm going to try switching to four bronze bushings there too.

Hopefully, at the current rate, I should have the first CNC64 up and ready before Christmas. If all goes well I would like to have the second one ready too, since my friend has been waiting since this summer.

[Yvan256]

I'm building a soundproof box for the Roland MDX-3 / CNC64. The local sign shop keeps giving me their plastic sheets offcuts and the last lot had relatively big pieces. Big enough that I was able to cut panels to make a small soundproof box for CNC64. What you see in the third image is the panels just sitting in place, held up by gravity. It took me about half a dozen attempts before it stayed up so I could take a photo. The problem is, I had enough PVC but only by having the bottom and back panels made from two panels. The fact that the top and the front are also cut in two for the opening door makes the whole thing look like a jigsaw puzzle.

The Roland MDX-3 will also fit if I cut about 40mm of the front legs and re-wire the parallel port to the top/back of the machine. I've wanted to re-wire the control panel at the top and bring it to the front for a while anyway so I might as well do everything in one step. But I'm not doing anything to the MDX-3 before CNC64 is up and running, just in case.

The soundproof box will open at the top+front, I'm just not sure if I'm going with an L-shaped door or two separate panels yet. I want to put a handle at the top but it would get in the way of an L-shaped door when opening it. It might be better to go with handles on the side, I'm still exploring options.

I still need to cut the sonopan and masonite panels but I did make a test cut on the sonopan yesterday while I was cutting the PVC panels. I was afraid it would just rip to shreds but as you can see in the fourth image, the cut is extremely clean.

There's still little "details" like letting wires in and hot air out, but I'm not at that point yet. I have enough sonopan that a soundproof air vent at the back/top is an option but I still need to design it. Hopefully I have enough PVC left to make it.

The current design, from the inside-out, is a sandwich of 3mm masonite, 19mm sonopan, 10mm expanded PVC (except the back and the bottom which are 13mm expanded PVC). I'm hoping the air bubbles in the expanded PVC will provide an additional layer of soundproofing. I have about 1cm of clearance inside the box versus the dimensions of CNC64, so if I need to add more layers to increase the soundproofing I should be able to add 3mm cork sheets and/or corrugated cardboard between the hardboard and the sonopan.

[Typefighter01]

Just curious, have you ever used Dynamat, and if so, how does it compare to Sonopan for sound proofing or deadening (price, ease of use, etc..)?

Your Z axis turned out pretty slick BTW.

[Yvan256]

Just curious, have you ever used Dynamat, and if so, how does it compare to Sonopan for sound proofing or deadening (price, ease of use, etc..)?

Your Z axis turned out pretty slick BTW.

No sorry, in fact you can say I never used sonopan either since I won't really know the "results" until the box is done. I'm using sonopan because it was free offcuts. What I can say about sonopan is that it feels like a thick hard foam panel made with paper fibers. It cuts real nice on a table saw so I'm giving it 10/10 for ease of use.

Thank you for the comment about the Z-axis, I'm really happy with the result too. But... I think I can shave 1~2 millimetres on each side, which would increase the X-travel by 2~4mm. It's not much so I'll probably only make those modifications for the second CNC64 though.

[Yvan256]

I've bolted all the PVC panels together and installed three right angle brackets in the cover, all I'm missing is the hinges.

After that I need to finish cutting the sonopan panels which I couldn't do because the sides weren't at 90 degres, I plan on cutting them manually from this point, test-fitting them one by one in the PVC box to be sure everything is alright before cutting the next one, etc.

The noise reduction from the PVC box alone is already quite good with the MDX-3, so I don't anticipate any problems once the sonopan is in place with the masonite to cover it all in the inside. I guess all those air bubbles in the expanded PVC do help a bit after all.

At this point I'm getting used to designing and assembling projects made from smaller panels, I've been thinking about a first revision of MVS-99-6 made from 150x90mm panels. 

[Yvan256]

Well, it had to happen. The MDX-3 is no longer fully functional. The spindle still works but one of the Z-axis assembly part is made of plastic and it's too loose to make precise cuts anymore. I'll have to design a completely new Z-axis assembly to make the MDX-3 work again. 

On a positive note, the parts required to finish CNC64 were already cut and I finished (temporarily) assembling the whole thing so that it could make its first cut, which was simply a 13x13mm square, about 5mm deep. It was supposed to be a 10x10mm square but I messed up my calculations. 

So, how did it go? I can summarize the result with two words: huge success! 

The horizontal cuts on the right were the actual first tests with only the Y-axis and Z-axis wired up, but they doesn't really count since this CNC has three axis and the square is the first cut to use all three. 

[Yvan256]

Since I haven't been able to move forward on my mini-joystick project (stupid small bolts), I had to find another way to control CNC64 manually.



The wires used by Nintendo were some kind of twisted foil around a thin rope and it's almost impossible to solder these things and finding a female socket is near impossible unless you take apart an old NES, so I rewired the gamepad with an old DB9 cable.

The NES gamepad is very easy to read, it's just a 4021 shift register and only require 3 pins on the ATmega328P!

Pressing left/right moves the RTX left/right, pressing up/down makes the bed move to the back/front and pressing B/A moves the RTX up/down.

Just to make things interesting I'll program it so that every press I do makes it move 2.54mm. That will allow me to manually drill a blank PCB with a fine drill bit and see if it matches up with a protoboard.

After a few manual tests, I will program the ATmega328P to make it cut the 3rd stepper driver board because the components are currently installed on the larger protoboard you see in the photo.

The keyboard is not part of the project.

[Typefighter01]

Cool idea 

Can you use the "select" and "start" buttons for other often used functions (pause, e/stop, zero all axis)?

[Yvan256]

After wasting a few hours with the pinout and wiring of the NES gamepad, trying to check which pins of the 4021 were connected to which colour of wires and mixing up the clock and data lines and not completely understanding how to "shift left" variables in GCC-AVR, I used another NES gamepad I had lying around.

It took what seems forever but I now have manual control of all 3-axis of CNC64. Well, I will have control, once everything is wired up. The wires of those surplus KP4M2 motors are rather short and I'm still missing my 3rd stepper driver board, but all the parts work. I have a tiny hope of completing the machine before 2014, except maybe painting the remaining panels which are already installed. I'm afraid that if I disassemble the panels to paint them, the machine won't work properly again once re-assembled. I'll wait until I can either finish the second CNC64 or repair the MDX-3.

Cool idea 

Can you use the "select" and "start" buttons for other often used functions (pause, e/stop, zero all axis)?

They could be used for that later, however I still need to add limit switches to all axis. But I'm thinking [Select] to go "home" so that I can install material / remove cut parts the same way "Home" works on my MDX-3, and [Start] could be wired to a relay to start/stop the RTX. 

[Yvan256]

There's a tiny backlash problem with regular threaded rods and my friend and I are afraid the nylon nuts will wear out too fast. The machines would also be way too slow for relief carving, which is something my friend will probably be interesting in.

I'm now cutting new parts for both CNC64 machines so that they will be belt-driven. I solved the problem of the low-profile base by mounting the Y-axis motor upside-down. Simple enough but I didn't think about that solution the first time around. I lose the ability of letting the panels to cut go beyond the back of the unit, but it's a small trade-off so that I can keep using the currently cut, painted and assembled bases.

edit: the MDX-3 keeps failing, I'm now cutting trying to cut parts to replace both the Z-axis assembly and the spindle. I'm making the mounting holes compatible with the Black & Decker RTX mount, so I'm only three parts away from an upgrade (the parts in blue in the image below), I already received the 6mm bushings from China. About half of them were bad, so the seller is sending me another full lot of bushings as a replacement.

edit 2: I've attached the Black & Decker RTX mount to the MDX-3 Z-axis with a single screw. There is a small plastic nub on the Z-axis that sort-of prevents the mount from rotating, and so far all cuts seem to be clean. It's making a lot less noise than the old MDX-3 spindle used to. I'll take a photo of this Frankenstein of a contraption once it's done cutting the current part.

edit3: Photo of what I call the "Roland RTX-3". 

[Yvan256]

The new Z-axis assembly is in place and working, the MDX-3 now uses my Black & Decker RTX rotary tool instead of the original, badly-designed spindle! 

The part being cut is a plate for a NEMA23 stepper motor mount. In all the NEMA 17 stepper motors I currently have, none has a 5mm shaft and can be wired as bipolar, so my only option is to use my NEMA 23 steppers. I wanted to keep them for CNC86 or even my bigger machine that's going to use my Hitachi M12VC, but that can be changed later.

I already received my GT2 belts and 6.35mm ID pulleys, which were needed because those NEMA 23 steppers were made in the USA and the shafts are 0.25 inch.

The other good news is that the NEMA 23 steppers have almost exactly the same height as my NEMA 17 steppers with the S2M metal pulley, so I'll be able to use them on CNC86 if I ever decide to buy NEMA 17 steppers for CNC64.

Driver board: check
Stepper motors: check
Belt and pulleys: check

Progress!   

[Yvan256]

I made a serious mistake with the new Z-axis assembly. I guess I was tired when I did the mistake, because it's really dumb.


Here's how I designed the assembly in Sketchup:




But since I wanted to have flush bolt heads:




I had to countersink the bolt holes:




However, since the design required a slot for a part on the back:




I had to do the countersink on the other side:




But I had to do that countersink, manually, with a drill bit on the other side:




And what had to happen, happened. The thin wall of expanded PVC broke, the bolt could no longer hold the part on the other side, the RTX mount wasn't tied to the Z-axis assembly anymore, the RTX became randomly offset with the programmed pattern and the end mill bit just continued cutting anyway:




Ouch.

I learned two things from this mistake:

• the RTX is powerful enough to cut at least 1/4 of the thickness of the MDF at a time, so I'll program deeper cuts next time, maybe 1/16" just to be on the safe side. With 3/8" thick MDF, this means only 6 passes to cut through, compared to my usual... 70 or so.
• The RTX is powerful enough to actually burn the end mill bit, that means I don't have to worry about the power of my spindle anymore, which is a good thing for future cuts with my modified "Roland RTX-3", for CNC64 and for CNC86.

I'll patch the broken part with a big washer, offset the bolt head thickness with washers, cut new Z-axis assembly parts again (maybe in double, to have a backup), order more Dremel #650 end mill bits (with a few spares as backup this time) and continue forward. 

[Yvan256]

I had problems with the new Z-axis assembly, I think there was too much friction on the Z-axis rods because the cutting depth wasn't as constant as it should have been from hole to hole. So, I had install the original spindle back on. Again. 

I've made the motor pivot stronger, so it should last for quite a while now.

In the meantime, my friend ordered two more Dremel #650 end mills this week and received them two days later. Considering how remote our small town is, it's amazingly fast. Kudos to Amazon.ca for the delivery speed.

The MDX-3 is cutting a NEMA23 mounting plate for CNC64/CNC86.

Edit: plate is cut, perfect fit with the motor. I have two GT2 pulleys for the NEMA23 motors of the X and Y axis.

[Typefighter01]

Looks like you won the race Yvan    .

1st-I have a crush on amazon.ca.
2nd-Mounting plate looks perfect.

[Yvan256]

Looks like you won the race Yvan    .

1st-I have a crush on amazon.ca.
2nd-Mounting plate looks perfect.

The race is still on, CNC64/CNC86 are far from finished. By the time your machine is up and running, you'll be able to cut all the panels for dozens of my machines before my MDX-3 would have the time to cut a single large panel. 

[Yvan256]

All the parts for the NEMA 23 stepper motor assembly are ready, except for two coats of paint. All that's needed to upgrade the Y axis to belt drive is the plate for the other side where the idler pulley will be installed.

2014-05-13, update 1: I just ordered F686ZZ flanged bearings to make the belt idlers. That's another delay of two three weeks. 

2014-05-13, update 2: angelinvestment2010 has already shipped my flanged bearings. This guy is fast! 

2014-05-16, update 3: the tracking number for the package still hasn't shown up in China's postal service database. 

2014-05-19, update 4: "International item has left originating country and is en route to Canada"

[Yvan256]

If I'm going to wait for the F686ZZ flanged bearings to get here, I might as well cut panels in advance so that I'm ready when they arrive. I might even find something else to use temporarily in the meantime.

[Yvan256]

Since I need to cut new plates for the Y-axis for the belt drive, I've decided to try to fix the bushings alignment problem.

Lucky me, the new design for the part "Y1" fits four times on one plate.



The new Y-axis assembly design has seven distinct parts, with four plates having one bushing each. Two top plates will then join the front and back plates+bushings and the left and right side will be joined together with a plate with slotted holes. The end result is something that can be easily aligned after assembly.

Here's an exaggeration of the "sliding" design that should allow perfect alignment with the Y-axis rods. We're talking about plus or minus a fraction of a millimetre here. Alignment has to be perfect, as I learned the first time around. Those bushings are extremely precise.



And here's the four Y1 plates with their bronze bushings.



Belt attachment design and three more plates to go.

[Yvan256]

If I had started cutting panels this morning instead of this afternoon, I'd be almost done with this... 

Oh well, here's the almost-first-half of the new Y-axis assembly.

Half-enjoy!

P.S.: the next post should bump this thread to page 6, I'm probably not the only one who's tired of downloading all those old photos from months ago. ;-)

[Yvan256]

Here's the new Y-axis assembly for the belt drive.

As I mentioned before, the two panels are adjustable so that the bronze bushings align perfectly with the smooth rods.

The first photo shows the assembly sliding "closed", the second one shows the sliding "opened" and the third one shows the bottom with the belt anchor and the slotted holes that allow the alignment of the two sides.

[Yvan256]

The Y-axis assembly is finally in place and it's sliding very smoothly.   

I'm searching for something to use as a temporary idler pulley, installing the belt and then testing the whole thing with TurboCNC and the driver board. If all goes well, I should have a video online "soon".

[Typefighter01]

Looks like your adjustable "Y" panels are going to make aligning them with the rods nice and simple.

I should have a video online "soon".

Hurry up 

[Yvan256]

Looks like your adjustable "Y" panels are going to make aligning them with the rods nice and simple.

I should have a video online "soon".

Hurry up 

Still looking for a suitable "idler pulley" option with the parts I have right here.

[Yvan256]

Here's the underside of the machine, to see details of the temporary belt installation.

And the video of the first movements is .

2014-05-26: update on the F686ZZ flanged bearings, the package has arrived in Montreal, so I should have them tomorrow.
2014-05-27: I have received the F686ZZ flanged bearings.
2014-05-28: the new front plates for the flanged bearings are ready.
2014-05-28: I'm being stopped by a lack of 6mm bolts... I should be able to get them before the week-end.

[Yvan256]

I'm dropping this project and getting a Shapeoko 3 instead. It's much less complicated.

[Rick]

I'm dropping this project and getting a Shapeoko 3 instead. It's much less complicated.

Sorry to hear this! I like the idea behind the Shapeoko, but there is something to be said for the homebrew CNC machine. (I've been trying to determine if my own skills are up to creating a 4'x8' CNC or else, paying someone else for a full-on kit.)

Can we not sway you back?

EDIT: YOU SUCK. Post left for posterity. (Thanks for the PM.)

 

[Yvan256]

Sorry to hear this! ... Can we not sway you back?

I've almost posted a photo of a pile of CNC64 parts and wrote "too late it's already in pieces".

I've re-done and re-cut so many parts that I probably have close to enough parts to build another machine. :p

[Malenko]

can you use your Shapeoko 3 to make me a CNC64?

[Yvan256]

can you use your Shapeoko 3 to make me a CNC64?

I think I'll just attach my Black & Decker "Shapeoko 3" to my CNC64 instead. 

As for making other CNC64's, that's the medium-term plan.

[Nephasth]

 

You had me until the demonstration on various materials. Gotta love infinite axes!

[Yvan256]

I made my custom Web-based CNC software for my Roland MDX-3, so all the coordinates system is off by a factor of 40. I'm really not looking forward to fixing all that to support regular g-code.  Not only that, but every function has to be coded three times: one for the HTML canvas preview, one for the Roland RML-1 code and one for the g-code. I guess I should plan ahead of the g-code rewrite and think about splitting every little cut into a function so that I only code three times once.

In the meantime, I'm currently testing a new function to cut rectangle outlines with tabs (the tabs are in green in the image). I'm also testing new 1mm bits I got on eBay, and the noise reduction from a smaller cutting bit is just amazing. The stepper motors of the MDX-3 are the major source of noise now. 

I also got two bonuses: the spindle motor has a lighter cutting load and I'm now able to cut smaller holes and finer details.

I'm currently cutting extension panels for CNC64, to increase the Y-axis cutting length by 40mm. It's not much but it's the maximum length increase I can do for that axis so that CNC64 still fits inside my PVC box.

of my Roland MDX-3 cutting a rectangle with tabs in 3/8" MDF using a 1mm cutting bit, in order to test the RML-1 code output of the new "rectangle_outside_with_tabs" function of my custom Web-based CAD software.

edit: okay, so I messed up the code a little bit. The "tabs" being cut aren't tabs at all, they're uncut parts of the total material height (imagine the green lines not being there at all, or white). I'm fixing the code to cut the 2nd panel. 

edit 2: first photo below, plate with messed up tabs on the left, plate with proper tabs on the right. It's new to me, so it seems a bit weird to see the right plate "float" like that. 

edit 3: more pointless plates with ugly square corners. It reminds me of plastic models for some reason.

edit 4 (2014-06-15): as you might have guessed, I'm still busy re-writing major parts of my Web-based CNC CAD program. It's going well so far and will allow me to add tabs to any shapes and also output regular G-code in addition to Roland RML-1 code.

[Yvan256]

As I said in my update above, I'm still busy re-writing major parts of my Web-based CAD/CNC program. It was already a mess of code with almost-duplicate code all over the place depending on the shape to be cut. Adding tabs to all of them and adding G-code afterward would have been a maintenance nightmare.

So far the rewrite is going well and it will allow me to add tabs to any shapes and also output regular G-code in addition to Roland RML-1 code.

I've simplified the functions quite a bit and these functions call new "lines" functions that output HTML5 Canvas/RML-1/G-code, so any new function that I write from this point on is automatically compatible with all three output formats. In short, the function that outputs drawing/CNC code is the same for all shapes, no more duplicate code.

I'm planning to have custom functions in my CAD for all the common "2.5D" CNC design requirements . The basic ones being "hole", "pocket", "rectangle", etc. I'm still not sure how I'll design the complex arbitrary shapes because the path usually needs to be relative to the milling bit (inside cut, outside cut).

I've yet to complete the RML-1 and G-code outputs, but here's a screen capture of the new "hole" function which takes a parameter for the number of sides. Since the RML-1 doesn't allow for circles or arcs, my CAD software will also only use lines for everything including circles.

The green lines in the image are tabs. The ends of the tabs are even rounded according to the milling bit diameter, too! 

2014-06-16, edit 1: RML-1 code output already completed? At this rate, adding G-code output should take less than one hour, once I understand which codes to output.

2014-06-16, edit 2: Testing a cut with the pattern shown below, I should be able to add a photo by the end of the day. It's working well so far, but there's quite a number of circles to cut. Middle ones are done, still need to cut the last two big ones.

2014-06-16, edit 3: I had to stop the machine at the last circle, it's 22:30 here. I'll post a photo of the cleaned-up plate tomorrow morning.

2014-06-17: After cleaning up the front of the plate, which came out absolutely perfect, I turned it around to clean up the back. Nothing, no slots, totally flat. Turns out I forgot to select the correct material before outputting the RML-1 code for my MDX-3. But I did watch it cut the first few holes and the Z-axis did go up and down, so I know the ramping tabs are working. Next step, flat tabs. Attached are the cleaned plate (rotated 180 degrees on the MDX-3 so I can cut some more tests on the left part of the plate) and a close-up of my CAD software with my stupid mistake highlighted.

2014-06-18: since I'm now cutting with 1mm end mills, I need to rewrite a number of functions that were hard-coded for 1/8" end mills, such as the "pocket" functions (half-depth holes and cutouts). See the last two images for boring details.

[Yvan256]

I'm still busy with my CAD software rewrite. It's going very well, but a lot of CNC64 parts need to be "re-traced", so to speak, since my software is very basic and is only mill-diameter-aware for shapes like holes and rectangles.

One good thing from using 1mm end mills, apart from the much lower noise, is the fact that I can pack a lot more parts per plate. I'm slowly re-designing a "CNC64 Mark II" with the primary goal being to minimize board waste. My old "corners" were oddly-shaped, had two versions ("left" and "right") and could only fit four corners per board. My new corner is a single version and has been resized so that I can fit eight of them on a single board. My CAD also now has an "offset" function, which took under five minutes to implement. I guess my new code is getting easier to maintain. 

Included below is a screenshot of the CAD screen for the eight new corners. I will cut a plate of those corners tomorrow, I'll update with a photo after they're cut.

edit: the plate is being cut as I'm writing this. I'm also testing a new cutting depth which is three times deeper as usual. The noise increase is barely noticeable. It took about 45 minutes for the first corner, so there's about 5 hours of cutting left. Next time I'm going to increase the cutting depth even more. 

edit 2: the plate is cut, but... I messed up in two ways. First, cutting too deep totally ruined the milling because instead of having 0.5mm notches in the inside corners, I got totally rounded corners that don't even look straight to the naked eye. I guess the new cutting depth was too much. I'm lucky the end mill didn't break after cutting eight of those parts. Second mistake, I know this is a code problem because my 3.5mm slots for the bolts aren't even 3mm wide. All that cutting for nothing. I should have made more tests, taken more measurements and cut a single corner as a test before trying to cut eight of them. 

[Yvan256]

I added a poll, for those who would be interested in purchasing a CNC64 kit, if it becomes available.

Plywood should be a better environmental choice and be lighter for shipping, but would be more susceptible to weather changes (temperature, humidity) and could require varnish/painting (I'm not planning on offering a pre-painted option either).

Expanded PVC and HDPE would be the lesser choices for the environment, however they're less susceptible to weather changes and would come in color, which means that if enough people were to order a kit, I could offer multiple choices (ex: black/red, black/green, etc). HDPE would also probably cost more to ship because of its weight while expanded PVC is a middle ground between plywood and HDPE as far as weight goes.

You'll note that MDF and HDF are not available. 

edit: there's a lot of updates on the parts of the optimized version, but I can't post anything about it yet. 

[PL1]

Any ball-park (+/- $100) cost estimates for your kit and also for the stepper motors/control boards/other needed parts to make it functional?

Might make it easier for people to determine if a CNC build can fit into their budget.


Scott

[Yvan256]

Any ball-park (+/- $100) cost estimates for your kit and also for the stepper motors/control boards/other needed parts to make it functional?

Might make it easier for people to determine if a CNC build can fit into their budget.


Scott

I haven't really priced all the parts for any given volume yet, but my target would be around 200 to 250$USD for a complete kit (CNC64, stepper motors, belt+pulleys/ACME rods+nuts, control board and stepper drivers and all the various hardware such as nuts and bolts), you would only need to supply two things: your own rotary tool (I would probably offer two or three choices for the mount) and your own power supply. I'm trying to use a regular ATX power supply for the setup, since those are easy to get everywhere, even if it's from old PCs.

Because of the popularity of RepRaps, the price of the hardware needed for a small desktop CNC has gone down quite a lot for the last few years, even more since the time I started my CNC project thread.

Even so, I'm guessing volume pricing on the parts needed would probably allow me to go even lower than the prices I mentioned above.

edit: revised the price range.

[Yvan256]

I'm still busy writing new parts for my CAD256 program.

The latest addition is a function to cut alphanumeric text. I've based the font on what is possible with a 16-segment display and retouched some characters a bit.

[Yvan256]

Took quite some time, but they're ready. It's nice to see eight plates on the same board. 

[Yvan256]

Well, adding support for G-code was easy enough. It's not totally optimized yet, but it should work.

I even tested it with SimpleGCoder HTML5 G-code Editor, and it looks the same as the output as CAD256, apart from my program displaying tabs and not showing the above-the-workpiece movements.

CAD256




SimpleGCoder HTML5 G-code Editor




So... since I haven't had time to complete CNC64 yet, is anyone stupid brave enough to volunteer and test the G-code output of CAD256 on his/her CNC machine*?

* 1/8" end mill (Dremel #650) required and limit switches on all ends of all axes is highly recommended. 

[Typefighter01]

Send it to me, I am brave supid enough 

[Yvan256]

This file should cut four J22B and four J22N plates, using a 1/8" end mill, in 9.7mm thick material (cut through 9.950mm, requires a spoilboard underneath your material, I use this setting for 3/8" MDF) with 1mm passes, with plunging cuts (I need to re-add ramping cuts in the new code).

There's only four kinds of G-codes used in my files (G21 to set system to metric, G00 for above-material rapid moves, G01 for all cuts and G28 at the end to send the cutting head home) because my CAD needs to stay compatible with RML-1 which is more limited. All curves are converted into lines. No speed codes are used so anything you set in Mach3 or other should stay set for the whole cutting operations.

Home is assumed to be to the left/front of the material, with the zero Z being the surface of the material.

The breakaway tabs are "ramping cuts" so I'm not sure they're long enough in the current file, be careful.

Please air-cut first, make sure you're wearing eyes protection and please make a video of it. 

[Typefighter01]

Hey Yvan,

Downloaded and ran your file. I dragged it to the desktop and tried opening it inside Mach, but it didn't see it, so I renamed it as a ".TAP" file, and it loaded fine. The aircut was a success and it behaved as expected.

I will pick up a 1/8" endmill this week and cut some material (yes, I will make a video).

It took 24min to cut a plate at normal speed and cranking the feed rate upto 300% it only took 8min (makes sense).

The program actually produces a catchy tune when the steppers are cutting the holes and turnung up the feed rate during this portion of the code was pretty cool.

Good work so far buddy...I will report later on in the week once cuttin bit is purchased  .

[Yvan256]

Hey Yvan,

Downloaded and ran your file. I dragged it to the desktop and tried opening it inside Mach, but it didn't see it, so I renamed it as a ".TAP" file, and it loaded fine. The aircut was a success and it behaved as expected.

I will pick up a 1/8" endmill this week and cut some material (yes, I will make a video).

It took 24min to cut a plate at normal speed and cranking the feed rate upto 300% it only took 8min (makes sense).

The program actually produces a catchy tune when the steppers are cutting the holes and turnung up the feed rate during this portion of the code was pretty cool.

Good work so far buddy...I will report later on in the week once cuttin bit is purchased  .

Thank you for the feedback.

When you say 24 minutes per plate, do you mean it would take around 3h12m to cut all eight plates included in the file? It seems incredibly long, considering my MDX-3 didn't even take three five times that long.

I'll be adding a few things, like add a dead-zone when dropping the cutting head above the material, ramping cuts, etc. They could be ready before you purchase your end mill, so once the optimizations are done I'll make a new post with an updated file.

[Typefighter01]

Hey Yvan,

Downloaded and ran your file. I dragged it to the desktop and tried opening it inside Mach, but it didn't see it, so I renamed it as a ".TAP" file, and it loaded fine. The aircut was a success and it behaved as expected.

I will pick up a 1/8" endmill this week and cut some material (yes, I will make a video).

It took 24min to cut a plate at normal speed and cranking the feed rate upto 300% it only took 8min (makes sense). It seems incredibly long, considering my MDX-3 didn't even take three times that long.

The program actually produces a catchy tune when the steppers are cutting the holes and turnung up the feed rate during this portion of the code was pretty cool.

Good work so far buddy...I will report later on in the week once cuttin bit is purchased  .

Thank you for the feedback.

When you say 24 minutes per plate, do you mean it would take around 3h12m to cut all eight plates included in the file?

I'll be adding a few things, like add a dead-zone when dropping the cutting head above the material, ramping cuts, etc. They could be ready before you purchase your end mill, so once the optimizations are done I'll make a new post with an updated file.

Yes Yvan, if no changes to your file are made, it would take 3h12m to cut all 8 plates. 300% increase to feed rate is as far as Mach goes, but from my own experience with routers and MDF, I would say, I could drop cutting time from 8min by another 2 or 3 minutes and still safely cut all 8 plates. That puts a total cutting time of around 40min. How fast do you want to see them cut?

Question...where (when) is cutting speed generated? I know in Artcam express, when I select different cutting tools, before I generate code, each tool has a feed rate listed beside them (suggested rates anyway as they are user defineable).

I will wait for your optimized code and aircut a potion before cutting real wood.

[Yvan256]

Yes Yvan, if no changes to your file are made, it would take 3h12m to cut all 8 plates. 300% increase to feed rate is as far as Mach goes, but from my own experience with routers and MDF, I would say, I could drop cutting time from 8min by another 2 or 3 minutes and still safely cut all 8 plates. That puts a total cutting time of around 40min. How fast do you want to see them cut?

Just cut them at the speed you are comfortable with. With my MDX-3 and the 1mm end mill, I'm used to something around 8 to 10 hours to cut all eight plates. Taking only 40 minutes to cut all eight plates seems like lightspeed to me.  Just be careful if you do buy the Dremel #650 end mill, I've read that it burns easily.

Question...where (when) is cutting speed generated? I know in Artcam express, when I select different cutting tools, before I generate code, each tool has a feed rate listed beside them (suggested rates anyway as they are user defineable).

My program doesn't generate any specific cutting speeds at the moment for the G-Code output, only paths. I do use codes G00 and G01, so there's "travel speed" and "cutting speed" parameters in the file, but the actual speeds for those codes should be defined by your software. FYI, the travel speed only occurs above the workpiece.

[Yvan256]

Typefighter01, did you have time to pick up a square end 1/8" end mill yet?

Hold on before cutting the plates, I'll make you cut something else instead. 

[Typefighter01]

Yes I did 

Grabbed a 1/8" for your testing and a 1/4" for me.

I just need to come up with a hold down system. I do have double sided tape that I will probably use for now.

Send me some g-code when your ready.

EDIT: These are Up-Cut Spiral bits, not end mills (thanks macattack)...

[Yvan256]

Yes I did 

Grabbed a 1/8" for your testing and a 1/4" for me.

I just need to come up with a hold down system. I do have double sided tape that I will probably use for now.

Send me some g-code when your ready.

EDIT: These are Up-Cut Spiral bits, not end mills (thanks macattack)...

Double-sided tape? It really didn't work on MDF last time I tried. Doesn't your router table have T-slots? Let's switch this conversation to your thread.

[Yvan256]

Small text-only update for whoever is still following this thread: I'm having problems with the 1mm cutting bits. I think they're flexing a little while cutting and it's messing up all my measurements on the parts by 0.2mm, which doesn't sound like much until you try assembling parts together.

I've added 200 milliseconds of dwell time between all lines to see if it helps. It does slow down cutting holes quite a bit though, I'll have to add an on/off parameter for the dwelling time. 

edit: nope, cuts are still messed up. I'll switch back to my 1/8" end mill to see if the problem is with the 1mm bits or with the MDX-3 itself.

2014-08-10: tested cutting a 10x10mm part with the Dremel #650, 1/8" straight router bit, there's still a 0.2mm difference on the cut part. The weird thing is, it's 0.2mm smaller so it's definitely not a spindle runout. The MDX-3 is getting old, maybe there's play in the frame since one side is broken, maybe the belts are under-tensioned, etc.

[Yvan256]

Coming soon to a thread near you: CNC-2116.

And a huge thank you to Typefighter01 for helping me out! 

[Yvan256]

Thanks to my "sandwich" shipping design concept and Typefighter01's packaging skills, the six plywood panels seem to have arrived intact. 

Also, thanks to Canada Post for delivering said package directly to my apartment instead of making me go get it at the post office.

[Yvan256]

Since the parts where shipping in panels, they had to be held in place on the panels by bridges which I then needed to cut once I received them.

A few hours with a Dremel cutting wheel and a router flush trim bit later, I had a pile of parts. I still need to cut out about eight parts but in the meantime I started to assemble everything. It's beginning to look like a CNC.

And this time it's not made of %$#@ MDF.   

Refer to the filename of each photo for a description.

[Typefighter01]

And this time it's not made of %$#@ MDF.   

The cabinet grade birch was really nice to work with and looks good, even without any sort of stain or surface treatment.

It was fun testing and prepping for your final design and it is cool seeing something I cut on my CNC being assembled 100's of kilometers away 

Now hurry up and get this thing together so you can help me design bartops 

[Yvan256]

Now hurry up and get this thing together so you can help me design bartops 

We don't need my CNC to be up and running for your bartops, we can work on them before that. 

[Yvan256]

The idler wheel for the X-axis is in place. The one for the Y-axis is another story, since I made a 1mm error in my plan. 

I'm now hunting for 6mm washers that are 0.5mm thinner. 

[Typefighter01]

I'm now hunting for 6mm washers that are 0.5mm thinner. 

You got a bench mounted belt sander? I have resized washer thickness many times on one. Literally just hold it against the belt for a few seconds should do it, or, relieve the area where washers go with a Forstner bit (you will want to do this on your drill press with wood clamped).

[Yvan256]

You got a bench mounted belt sander? I have resized washer thickness many times on one. Literally just hold it against the belt for a few seconds should do it, or, relieve the area where washers go with a Forstner bit (you will want to do this on your drill press with wood clamped).

I'm just going to trust gravity on this one. I have the same setup on the Y-axis of CNC64 and moving it manually doesn't make the idler bearings go up vertically, so I've put a washer at the bottom only for now.

[Yvan256]

Sorry for the lack of updates on my CNC, but I'm working on Dainamikade with Typefighter01, along with other personal projects.

The CNC is finally up and running, but there's still a lot of things to complete. I didn't notice when they were on CNC64, but the rods for my Z-axis aren't even 8mm (they're 5/16") so there's a tiny bit of play and my cuts aren't 100% precise. I'll have to cut one of the Y-axis rod of CNC64 to complete CNC2116. 

In the meantime, here's a few photos of the machine for those who are still following this old thread! 

The first photo is the "completed" machine without any tool attached.

The second photo is with the Black & Decker RTX rotary tool and a plastic spacer board (to clear the bolts heads of the bed itself, as seen in the first photo) installed on the bed along with a hardboard spoiler board.

The third and last photo is something the machine cut this afternoon (I really need to upgrade the code in my CNC program) which is half the spacer board. Tomorrow I'll be cutting the right half and I'll be able to replace the plastic spacer which I'm afraid is too slippery and could add movement to the part to cut.

[Typefighter01]

Looking good buddy.