My Cabinet Build *Updated* Post Mortem

[Howard_Casto]

So I have a lot going on right now, but I'm still squeezing in a bit of fun when I get a chance.  I finished the 360 degree wheel.  I had decided upon an arbitrary pipe size to stuff the guts inside early on as it seemed as small as I could get and still be sturdy.  In hindsight I should have went one size larger.  It's 2" pipe and my design was based around those "slip" end caps that fit inside the pipe.... get some of them, cut a hole for some lipped bearings I found, bolt them tight and the whole thing is good to go.  Only I couldn't find 2" slip fittings, only traditional, over the outside, end caps.  So I tried some knockout plugs but they were too flimsy.  Eventually I cut up the 2x4 test rig I made in another thread to fit inside the pipe.  After entirely too much fiddling I managed to make it sturdy while looking fairly nice. 

The build is entirely too odd to explain, so I'll post some pics later. 

Next on the agenda is adding code to my teensy to make the spinner control the mouse.  I found some source and in theory it should be easy.  We all know how well turns out though. 

[Howard_Casto]

So I'm having trouble getting the 360 wheel to work on the teensy.  An incremental encoder isn't all that hard to setup... give the sensors power and read the two pins, but the pins are throwing back garbage, flickering on and off even when the wheel is at rest.  I hooked it up to a mouse hack and it works, so it isn't the encoder pcb.... perhaps it needs more power?

I might have to drop this one and do a mouse hack.... it's holding everything up. 

[BadMouth]

What game was the optical encoder originally out of?

The standard was "active low", but there were some oddballs that were "active high".
I have a badlands setup which IIRC is active high.

Since it's throwing garbage at rest then that might not be the issue, but it could play a role.

[SailorSat]

throwing garbage at rest? you might need external pullup or pulldowns as BadMouth suggested.

[Howard_Casto]

Well I tried it with a pullup for each pin, but then it just never changed, so I'm not sure what is going on.  I might try mapping each pin to a joystick button to see if I can see some sort of a pattern.... the serial monitor sucks for rapid debug. 

It was out of a wico trackball for the 2600, so it used standard wico/Atari pcbs. 

[BadMouth]

It was out of a wico trackball for the 2600

I didn't see that one coming.   

[Howard_Casto]

Eh the bearings wore out on it and since it uses arcade parts anyway.....

I'm beginning to wonder if I've got the wrong pinout, but I don't see how. 

A 5v Gnd B

That's pretty standard right?

I would check my mouse hack to confirm, but I did it 15+ years ago so I can't remember how I did the pinout. 

[Howard_Casto]

Well I'm at a loss on this one.  I still can't get any readable data.  If someone can find me an Arduino based example of a trackball or spinner maybe I can figure out what I'm doing wrong, otherwise it's time to move on, which is frustrating.

[BadMouth]

A 5v Gnd B

Those are the usual pins.  No idea about the order.
If "active low" it's easy enough to tell from the traces.
5v to all 4 thingamajigs.  Ground on a shared trace to the 2 emitters.
Then A and B each on a separate trace from the sensors.

You should be able to just test for voltage coming from one channel and it should change as the sensor is blocked.

I can see having a brain fart and reversing the 5v and ground because you're used to ground going to all the components, but in this case the 5v does.

[Howard_Casto]

Well they look correct to me. 

I hooked the outputs to analog pins and with them set as a pullup they don't move much at all.  Without the pullup they are all over the place.  So I'm thinking it's a power thing.  I might try hook them to an external 5v source and see what happens. 

[SailorSat]

No, you need pulldowns. (10k to gnd or something)
If the signal is active, the line is 5V/HIGH
If the signal goes inactive, the line will disconnect - hence the readings are all over the place.

a pullup will keep the line at 5V/HIGH once the opto disconnects.
a pulldown will pull the line down to GND/LOW.

[Howard_Casto]

I'm honestly not sure if I used pullups or pulldowns.... pretty sure I tried both and neither worked, but I'll try one last time tonight.  Thanks for the advice. 

[Howard_Casto]

I get what you mean now... I guess my brain was fried from trying this all day.  There's a resistor on the pcb that I assumed was a pulldown to ground, but looking closer it's actually going to the 5v, so that's most likely the issue. 

[Howard_Casto]

Well I tried a resistor with some power to keep the pin from disconnecting, but it's still jumping all over the place, so I'm not sure if it's a board issue or what.  Anyway, it's time to quit playing with this... it'll just be quicker to hack a mouse and I have one lone ball mouse remaining. 

The pin layout of a teensy is rather odd, so I'm going to wire everything up to a breakout strip and that way I can re-visit the issue later easily. 

[Howard_Casto]

I hate to keep posting about lack of progress, but the whole reason I do these logs is to show people what kind of work they will have to do and what problems they may run across when building their own rig. 

I started mapping out the pins to the teensy, and I'm going to run out anyway, so keeping the 360 degree wheel attached to separate hardware is probably for the best anyway. 
I've got the led controlling part of the firmware written and it works quite well.  I'll do the inputs next.  I've been reading up on how to save settings in a program directly to the avr, so the hope is I'll be able to re-program the device on the fly like the pros do.  I'm actually thinking that changing variables in memory might be better though... it doesn't write to the device, so it would be less wear and tear. 

I'm doing some creative things with the programming and I will release my source... I think it might be especially helpfull for those doing conversions. 

[Howard_Casto]

I'm chugging right along with the avr software.  I got inputs working along with a user-definable alt-mode key and alt modes for all the inputs.  All keys can be re-defined via serial communication.  Right now everything is a keyboard key, but I'll add the ability to switch to joystick mode.  Tomorrow I mess with my analog shifter idea and see if it's going to pan out. 

I need to order some parts for the tacho/speedo so I'll re-assemble my rig after my shifter experiments as they will take a while to come in.

[Howard_Casto]

So I've confirmed that some of the pins on this particular teensy is a little wonky... it's been abused somewhat though, so I can't complain.  Anyway, I had to do some rather creative wiring to get a clean signal for the analog joystick I intend to use for my shifter.  Totally worth it though.  I'll code that tonight hopefully. 

[Howard_Casto]

So the virtual gear switches from the analog joystick is actually working pretty well, so I think I'll go ahead and build a simple h-shifter. 

[Howard_Casto]

My little ps2 to usb adaptor came in today, so I just simplified my life and hacked one of my few remaining ball mice.  Man, those traces seem to have magically shrunk over the years, I can barely make them out.  Getting old sucks you guys. 

Anyway, the hack is in place.... I played a few games of Offroad.... all is right with the world.

I might try to build my h-shifter box tomorrow.

[Howard_Casto]

So originally I thought I wouldn't put a coin and escape button on my rig.... I just don't like a lot of button clutter, but upon further testing I figure at least an escape button is necessary.  So I hunted up some buttons I bought for another project.... they are black and only 3/8 inch across.  I can put those on and they will be fairly non-descript. 

The part of the puzzle I'm trying to figure out now is the shifter box.....  I did the math.... I need a 5.5"x5.5" box.  The thing is I'm not sure where to mount it.  Ideally beside the seat would be best, but It's have to be 10 inches tall to be high enough and a 5.5x5.5x10 box is going to look really weird. I kind of wanted to stick to my 2' x 2' footprint as well and that would decidedly stick out the side.  Also I'm not sure about the wiring.  As you all know my rig is collapsible, which makes anything mounted to the seat tricky to wire up to the control panel, which is mounted to the sliding platform.  I'm open to any good ideas. 

[Howard_Casto]

Well I'm still working on the code for the avr.  My problem is I really like to program.... it's distracting sometimes.    I've decided to abort my plan to steal one of the pedals for an axis on my star wars yoke.  I've got pins to spare on the avr and there aren't that many games in which having the axis from two different joysticks will cause an issue.  Borrowing the axis with the pots I want to use would require a lot of trial and error in regards to adding resistors to get the right values that I just don't want to bother with. 

So I've got some cosmetic stuff to do and I need to drill the holes for the coin and esc buttons, but I should have everything back together and playable very soon.... just in time to tear it all apart again for stain/paint.  So yeah, before I do that I'll be sure to take some more pics. 

[Howard_Casto]

I guess a monthly update is all I'm doing now. 

So a bit of a setback.  Apparently while the airflow in the bottom of the cabinet is more than adequate for the pc, adding in the transformers for the wheel and amp made it too hot.  I'll have to add additional ventilation and possibly some fans.  Also the cheap amp I mentioned in another thread went poof on me.  It technically still works, but the board is getting crazy hot and I've got a hum that I can't get rid of via normal ground loop isolation.  The speakers are still fine though, so I'll just pull another amp. 

I picked up some stain and I've also cut the box for the gear shift.  Things are moving right along.

[Howard_Casto]

Man I really need to post some pics don't I?  I'll try to pull it out of the corner and take some snaps tonight/tomorrow. 

Ok a ton of progress, but it's the fiddly stuff that drives you mad.  I've used a poly-stain on the pull-out platform... antique walnut.  I've also taken an old microwave shell and used the metal case to skin the sides.  I did this because it already had these nice metal louvers on it which will help with the ventilation issue.  I sanded all the paint off of the metal and polished it.  The idea is to give you the feel of a 1930's/1940's sports car.  Why?  Well because in that era airplanes looked virtually the same as race cars, and I also intend to do flight games on it.  On that note I made a star wars yoke with a walnut and brass finish which turned out fantastic.  I still need to paint the handles and wire up the buttons, but it's good to go. 

Right now I'm struggling with the shifter.  I'm trying to make a design that can be removed as it adds so much to the width of the cab.  Nothing I've tried thus far seemed right.  I'll try to fix that later this week and if all goes well I can stain the box the seat sits on, skin some remaining cosmetic details and put the avr innards in to finish the programming. 

I also need to make a wooden wheel, but from what I've researched that is a project in and of itself.

[baritonomarchetto]

Pics or nothing is happening

[Howard_Casto]

Ok so I'm bad at updating hardware projects... I'm trying to fix that.  Anyway, the racing cab is officially dead, I'm rebuilding it.  Before I start a new thread let's got over the finished project so others can learn from it. 

First let's look at a pic of the finished project.....  Excuse the dangling wires as I've already started taking it apart to cannibalize it for my new rig.  Also the seat cover is new.

[Howard_Casto]

As you can see the finish is pretty rough, but it does collapse and it did have most of The features I wanted.  Here are a few more pics before I talk about the results:

[Howard_Casto]

Let's just break it down.....

What Worked:

The detachable wheel system I came up with really worked well.  I just cut the wheel off from the base and attached pvc pipe fittings to each end.  The pieces already stayed in place rather well via friction, but I drilled a hole through the assembled shaft and put a cotter pin in there to keep the two halves connected.  This allowed me to remove the analog wheel when the optical wheel is in use or install my homemade star wars yoke.  The yoke also worked quite well btw. 

My miniature view buttons also worked well.  The smaller cluster makes for an easier mount in a smaller dash and I think I might actually prefer them to the original. 

I think the smoked plexi and old coffee cup rings for the dials worked quite well.  It's a shame I never finished the software to light them up. 

Finally the rig did collapse quite well, which was the main goal at the time. 

What Sort of Worked:

The optical wheel, which inserts in the center dial worked pretty good but it had one issue... I didn't have the sensor pcb mounted well and it'd occasionally scrape the encoder wheel.  It's currently off the tube atm... I now have a 3d printer so I'll design a new mount for the next rig. 

The lighting and controls hooked up to a teensy worked well, but I never got around to finishing the software or fully programming all the scripts in mamehooker.  It worked for Offroad thunder, Outrun fxt and a few other games I bothered to setup.   The main reason I didn't finish is the fact that the teensy and attached buttons/controls weren't easy to remove... so  I had to program on the racing rig's pc... which was rather awkward.   

The powered seat was also a partial success.  It was fine for sitting, but rather heavy and ugly.  When I started this project my grandma didn't need a lot of help physically so it was fine.  Now that I'm lifting on her everyday it's just a pain in the butt.  For now I'll keep it but I'm actively looking for a replacement. 

The analog shifter, made from a pc raider joystick and some L brackets worked surprisingly well for what it was.  The only problem is it shredded the rubber washers I used for stops, the return spring wasn't strong enough and the software I wrote for it had a slight bug regarding 6th gear and I never bothered to fix it because I play 90% of the racing games in automatic mode.  I'm going to do a brand new shifter build based upon the th8a shifter in the future because I want a shifter for daytona and those older hi/low shifter games. 

What Failed:

The overall end look wasn't at all what I had intended to do.  the plan was to cover the whole thing in polished metal with rivets ect. similar to a WWII era plane/race car.  The problem was, as the years rolled on my free time lessened, and I just didn't want to work on that sort of thing in what little time I have to myself.  In addition I totally screwed up the stain, which each piece of wood ending up a different color. 

The led array dials I created for the smoked plexi inserts are also a failure because I never installed them.  I soldered them up, put a test program on there to make sure they worked and just never got around to it.  This is mainly because I wanted to use them more like a secondary monitor, which again required a lot of arduino programming that I just never got around to.   


Overall I just feel that I tried to do to much.  The new rig will be scaled back and I'll start a new build thread soon.