Vector-gasm... Cosmic Chasm!

[bobbyb13]

I have been trying to assemble the parts to build a vector machine for a while now.

In the midst of some turmoil, there are actually a lot of cool things going on with vector action in the hobby recently as a result of a few guys who are obviously brilliant but also helpful and generous enough with their time to even help a jackass like myself.

In spite of that, it is remarkable that the written documentation about being able to do any of this stuff (and some fairly critical details of it all) absolutely SUCKS.

If I can succeed in actually getting this thing functional (and I feel like I am nearly there) it will be fun to be able to backtrack through my build and turn this into a fairly instructive thread in how to build yourself a legit vector machine without having ANY original components to begin with.

Kind of the pivotal thing to get this rolling (apart from actually learning some about XY monitors and amassing some parts) was ZapCon 8.2 when I got to play a REAL Cosmic Chasm.
I'll never forget the first time I played Tempest and getting to play CC (which I was entirely unfamiliar with except for knowing the name) gave me that same 'love this!' sensation that vector games gave me way back when.

Big props to Brian at classicarcadecabinets.com on this project.
I emailed him about the accuracy of the plans he has listed there (which consequently are almost identical to Dragon's Lair cabs too it appears) since I was considering having this one CNC cut to begin with and he offered to chat on the phone about the details.

His site is awesome and he is a another great community member generous with his time and knowledge.

I wound up just going cowboy on this one too as the place I asked about cutting it ghosted me.
Whatever.
I had Brian's plans and a protractor.



I have really refined the base building so that you wind up with a machine that is a solid build, easy to move, and still doesn't jump around when you are playing it.
Recently changed to using 2x2s to frame it so the side panels have something just sturdy enough to glue/screw to and make for a solid box.
2" hard rubber casters are key.  Fixed ones for the rear, locking swivels for the front, spaced inboard enough that the things spin freely and don't hit the side panels.



To make it so that sides aren't dragging on the floor I put 3/4" ply on the floor on either side and rest the side panels on it while I attach them to the base.

And if you make it so that the kick/coin door panel is off the floor enough (I set the bottom edge flush with the bottom of the base panel) then you have enough room to reach under and lock the front wheels without needing to pick up the whole cabinet.
No need for leg levelers either.

This has proven to be a fun build because the marquee area is pretty ridiculous.
When I had a good look at original cabinet pictures again it became obvious that there were some pretty weird angles to it all and to make it even more weird is that not only is the marquee on three different planes but the whole thing is tilted in a manner not parallel to much else on the cabinet.
Pays to have a good compound mitre saw for crap like this.



To skip through a bunch of boring stuff, I went through the cut list I made from Brian's handiwork with a DWG viewer and put the rest of the main body together.
The original designers made some weird calls with these cabinets but when you consider that they were trying to introduce games into a bloated marketplace it kind of make sense that they wanted the things to stick out like sore thumbs.
Odd angles to the whole thing, even if not really crazy, marquee lower, brighter and a really obnoxious shape, even the front of the control panel lit up (which I'm not doing round 1 at least.)



I altered the screen face angle a bit (so it isn't such a shoe-gazer) and also the control panel dimensions a tad.
The monitor carrying bit is very Ms. PacMan so easy, and I have stuck to the removable/back screwed kick panel so that I can add coin doors whenever I get around to it.
When I can stomach the cost of the damn things that is really.
Getting the bezel area built will be a ---smurfette---.

I couldn't wait to glass something and had some decent weather for painting (before it started raining incessantly a month ago) so I had at it.



I'm really liking the fiberglass over wood control panel thing.
If you rout the backside properly you have enough room for whatever you want for controls and the reinforcement offered by a few layers of glass makes it pretty tough feeling too.
If I ever feel like getting fancy with CP art then I can tape off and paint or resin up myself or anything you buy from somewhere will stick.

It was at this point in the project that I discovered that the marquee piece was actually a single piece of milky plexiglas that was bent to fit.
I had agonized about getting the angles of the parts carrying the marquee back(plexi)glas as true as possible because I knew that I was going to need to use the actual cabinet as the mold to get the freakin piece bent to fit properly.
I went back and forth with a piece of cardboard and a pencil and razor knife and got the pattern so that the plexi would be the right size and dimension.

Getting it bent to shape wasn't really that hard, just took some patience with a heat gun set on high (both temp and air flow) and getting it JUST soft enough to bend but not get gooey.
 Maybe I should go get some pics of that part in case anybody else is dumb enough to try this and wants some hints.

[javeryh]

bobby this is incredible.  What a cool looking cabinet.  Great game too - I had never played it before but I installed it on my cabaret because of the spinner and really like it.  Of course, the vectors on my CRT don’t look so great but it’s passable.

I’m really interested in your resin control panel - how are the results?  I’ve seen a few people attempt it but it seems like it would drip off the edge and make a mess plus I am not sure I could get it 100% level.

Can’t wait to see more updates and I’m glad I have another project to follow around here!

[bobbyb13]

I don't know about incredible but it has been fun up until now at least- so thank you!
Really is a fun shape cabinet.
Part of my scratch repro build mantra has been 'make something iconic' so this certainly fits.

It really is impressive how good you can get vector games to look on a raster monitor in mame really, but I liked the challenge of building a vector and therr really is no substitute..
It was really intimidating before I started but they are easier to deal with than rasters it turns out.
It's a simple system and between Fred, Barry, Jason, Mario, et al literally everything for a scratch build is a known entity and very do-able.

The resin thing can be tricky depending on how you go about it but this is the 8th one I've done in fiberglass over plywood and I won't make them any other way now.
Maybe the time has come that I do a detailed tutorial with good pictures?
The Neo Geo cabinet build will be a fun one for that, so maybe that is the one to document glass work.

Seperate faces require some steps be done twice but gravity makes it level for you if you do it properly!

I'll see if I can get some pictures to look right and add a few more.

[bobbyb13]

Getting better at picture tweaking!

So the marquee bit was not really so hard once I thought about it.
Get the frame (top and bottom panels) set solidly.
Make a cardboard template of what the plexi needs to be.
Cut plexi to shape and then heat along line to be bent and use the cabinet as the mold for how the bends happen, doing one of the creases at a time.

It actually worked out that way.

I did need to take two shots at making a cardboard template.
It hadn't hit me until I looked at my first attempt that not only were the top and speaker panels on different planes, that also the sides of the marquee panel would not have square corners- anywhere.



Thankfully I only had to do the bending part once- probably because I was agonizing over screwing it up.
I used two plywood scraps for that- one to brace the piece in place and then the other to press the other side to the right angle against the marquee frame.
I worked from the outside edge in and it worked like a charm.

After I had the plexi bent properly I just needed to get some aluminum L bracket bent to match and hold it in place.
A single cut at the proper angle to one side of it and then I only needed to bend it into place and trace the line to cut the triangle out to get it look OK without needing to weld the silly thing.



Repeat for the other angle and then dupe for the top edge.
Spray with a little black primer and then top coat with matte black and there it is.
After that it was easy to pilot holes for trim screws (it is nice to use aluminim for this stuff) and it really fit together pretty nicely for something that I tossed together without having done before.



After I got the plexi to fit well with the brackets in place I needed to figure out how to backlight the thing- because for one, it is a huge cavern of space actually and two, I was concerned that getting light through that milky plexi might be tough.
I opted for ultra low tech in order to flood the space with light and it appears to be okay.
I bought a $3 old school style medium base bulb holder and screwed it to a piece of plywood.
Stuffed an LED floodlight bulb in it and I'm satisfied.



I also got the monitor going (that is a month long saga all its own) and have the USB-DVG working too, although the damn thing hangs here and never loads the VMMenu to actually be able to play anything.
Especially irritating since I have the spinner and buttons all wired up and ready to go too.
Rrrrrr....

I hate being Linux stupid still, but there it is.

[javeryh]

Great work!  I love the marquee shape.  Looks so much nicer than the flat ones.  I am so jealous how quickly you work. It snowed last night so I'm not getting into the shop any time soon (although I'm kicking around some ideas).

So that's a vector monitor in there?  I know less than nothing about them.  Are you converting a regular CRT somehow or are they completely different animals?  How do you source them?  Seems like there would be next to none in the wild...

 

[Mike A]

That USB DVG is quirky.

I know of at least one case where it nuked a G08.

And it never worked right with my G08.

[Xiaou2]

Great work!  I love the marquee shape.  Looks so much nicer than the flat ones.  I am so jealous how quickly you work. It snowed last night so I'm not getting into the shop any time soon (although I'm kicking around some ideas).

So that's a vector monitor in there?  I know less than nothing about them.  Are you converting a regular CRT somehow or are they completely different animals?  How do you source them?  Seems like there would be next to none in the wild...

 

 The main differences between a standard CRT and a Vector monitor:

 A standard TV draws the picture, like how you read a book:  From the top left to the top right...and then it moves down one line, and repeats this process, until it completely fills the screen (Zig-Zag).

 A Vector monitor is different, in that it can start drawing ANYWHERE on the screen..  and it will draw TO any point on the screen.   This means that it you want to draw a Triangle on the screen... you dont have to wait for the beam to draw the entire screen.  You only have to tell the beam to draw a line from one point to the next.

 How does this effect the Picture?

 Because you have 100% control over the Electron beam... you can achieve things that a standard CRT cant do.   For example... you can get Dots and Lines,  at a brightness that NO typically driven CRT can Achieve.  That brightness level, is only really Matched / Rivaled  by actual Lasers.   This is because... if the beam is left for too long in one spot (or is being told to re-draw that same area so often).. the amount of Energy that can excite the tubes Phosphors,  will be FAR higher than what is possible when the CRT is limited to a specific timing... when drawing the entire screen.

 In fact, you have to be careful about Burning the Tube itself.   Some Vector games whos electronics have gone bad... have caused the beam to stay perfectly in the center of the tube (not moving) ...and if you allow that beam to stay there for too long... it will cause permanent  "burn-in"  damage to that particular spot.

 The other feature about the Vectors... is that since you can control the brightness of the beams by how often you re-draw them... and or the voltage levels of the beam itself... you can create many different Intensity levels for your graphics.  For example.. in Asteroids Deluxe.. the players bullets are Extremely Bright, compared to anything else on the screen.  Also... when your ship breaks apart... you will notice that the pieces slowly fade from being mid-brightness... slowly fading in intensity, until the are invisble.


 I have theorized, that if you used something like an older Lamp based LCD.. but rather than using its standard backlight... you used 5 times the amount of lighting power... then with some Customized programming.. you could probably simulate a Vectors intensity, and effects.

 That said... Laser Projectors  are still one of the best ways to match a vectors brightness and dynamic brilliance.  (The type that they use for Laser-Light-Shows... not the modern version that are used to project images/films).

 There is a company that had made a relatively cheap Laser projector,  and people have made a custom set of vector games for it, based on vector arcade games.  The main problem with that particular projector... is that its not fast enough to be able to Draw all of the Vectors in our Emulated Arcade games.  As such, they have to make their own versions, which have reduced numbers of vector lines, and reduced complexity.

 Professional Laser light show projectors... that might be capable of keeping up with an Arcade vector games drawing times / number of vectors... cost several thousands of dollars.

[bobbyb13]

Now that I am actually learning about these things, it's remarkable how much less complicated the design of a vector monitor chassis system is compared to a raster one.

And I can see how a DVG board could roast a monitor Mike.
The relationship of supply voltage, flyback health and case transistor cooling have everything to do with how long any of this ---steaming pile of meadow muffin--- can survive actual use and the DVG is built to run stuff hard if it is set to.
Apparently the more you crank the input voltage to the deflection board the more the beam can deflect- within the abilities of the flyback at least, so not sending more voltage than the transistors can deal with is critical to longevity.
Larger screen can be filled with higher voltage of course then, but the transistors can only handle so much thermal stress.

I wonder if your setup (I figure you're talking about the Star Trek?) would benefit from a decent cooling fan or two.
Not sure how those deflection transistors are set up on a G-08 but I can tell you that the way they are on an original WG 6100 is just plain stupid in regard to keeping them cool.
A minor rearrangement of those case transistors and a single fan in the right spot could've altered the history of vector games.

And yes, javeryh, that is a consumer tube properly hijacked to think it is a vector monitor.
And the image it displays is beautiful !
Now if I can get the damn picture rotated properly for vertical games and get the RPi and iPAC to cooperate it will be fully functional.

It has taken me a few years to collect the parts and figure out how to set them up- through a lot of frustrating trial and error in some cases.
Probably spent $30 just on fuses out of paranoia.
 
It has made me crazy that you can't find this info in any one spot and I want others to benefit from my pain!
After all the drama I've endured it turns out it is a simple thing to pull off.

Once I actually get it to work properly for a few days at least I will document the whole vector monitor build here.

It REALLY is not that hard to do but knowing some details I have sorted out makes all the difference.

[Mike A]

That is from a G08.
I only used the USB DVG, borrowed from Arroyo, to narrow down where my problems were.
Then I heard it fried someone's G08.


It turns out that my problem was  bad -5v from the power supply.

Sent from my Pixel 6 using Tapatalk

[bobbyb13]

Man, if having a fan on there is stock it is a hell of a lot better design than the 6100 where the oh so crtitical transistors are just randomly scattered around on the monitor case itself- without a fan in sight.

Too much voltage will overwhelm any transistors no matter what the cooling scenario is I think, so who knows what paricular thing roasted that other G-08.

The Alan-1 guys built a fan cooled box like that G-08 one as an upgrade piece for the 6100 and it looks like it works perfectly.
I've got a last generation 6100 deflection and neck board, new repro HV board, and one of those Alan-1 transistor blocks too.

I think the magic is finding a tube that has a really low resistance/impedence inner yoke winding so that you don't need to mess around with purity and convergence later by using mismatched tube components.

By the time I am building the horizontal vector machine I should have that 6100 assembly running.

Still battling with my damn advmame settings on this thing.

Can't get vertical game image rotation.
Can't get into 'in game' adjustment because the Tab key doesn't do squat.
Can't get any controls at all to work in game- even though the spinner and buttons work in the menus.

Totally stupid since I can do this on a PC likety split.
Maybe I need to ditch the RPi for this build.

[Xiaou2]

Just a heads up...

 There is a guy that Ive seen on Youtube,  that I believe has made new boards for Vector equipment.. and much more.

 His Channel is called  " Arcade Jason "

[bobbyb13]

I've been keeping an eye on that guy but hadn't seen that video yet!
Appreciate the tip Xiaou2.

Jason has actually been helpful to me in getting this machine working.
I have been really interested in his Masteroids project, especially now that it appears that there is color capability added to that project.

One way ArcadeJason was massivley helpful was his video on yoke winding.
He speaks the truth- it is NOT really that hard to do.
I screwed up by winding my ferrites in opposite directions the first time around.
Man, did that ever make the HV chatter more like a shriek.

Second time around was much better.
He confirmed for me that the much lower value I wound up with for resistance/inductance was still ok, even though it was 30-40% lower than what I had read- from what I could find anyway.

And this time around I came up with an easy way to manage the spool of magnet wire while I was at it.



Clamp a huge screwdriver to your bench with a bit of tilt to it to keep the spool from falling off.

I still have no idea why Jason, Barry, Mario, Chad, et al aren't regulars in posting their work over here (where it would be appreciated!) instead of taking heat from all the grouchiest people in the klov crowd or having all of it reside in such a disheveled fashion over on fb.

The hobby as a whole would benefit greatly from a larger dissemination of their efforts- especially outside of social media crap like facefart.
This machine has taken so much more time to build just from me not being able to find important details without weeks of searching and finally getting lucky.
And it still isn't working properly.

So a fellow arcade nerd out here came over today to help me try and get this cabinet to behave and made a huge leap by getting my in game screen orientation fixed- by changing a setting that I had glossed over as having no bearing on the situation.
Thanks Chad!
Now everythying at least faces the way it supposed to.


Probably even more frustrating than that even is that although I have had spinner functionality straight through (I have a SpinTrak plugged directly into the Pi through USB) the device does NOT show up at all during game play.
The iPAC and the buttons I have connected to the Pi through USB do somewhat, but the spinner- no dice.

Extra frustrating since if this was a PC and GroovyMame I would have had it fixed immediately, but I have had no clue how to get this sorted in AdvanceMame as the configuration file setup is almost entirely different (and in Linux of course, in which I am pretty much entirely illiterate.)

Well, I discovered something accidentally today as regards in game adjustments on the Pi for AdvanceMame at least.
The whole time I have thought that the Tab key on my keyboards (I have tried 3) must be broken because I would hit Tab and get no pop-up menu for adjustments.
What I didn't know until today was that it was happening the whole time except that it ONLY shows up on the screen I have connected to the Pi via the HDMI 0 port and NOT on the vector display!
Yeesh.
That was somewhat helpful as I could map the control panel buttons the way I want them, but there is still no capacity to register the spinner there as the in game menu for AdvanceMame feels abbreviated in comparison to the one for GroovyMame that I have been working with on the other machines.

I DID find something about forcing the Pi to look for mouse inputs in AdvanceMame but even after trying to insert those lines still I got nothing.
The way that entering data lines goes in Linux gives me flashbacks of mainframe torture in college using Fortran77.

I still suck at it all.

[javeryh]

The vectors look so cool.  What software are you running on the pi?  Any reason why you aren’t using a PC since you have the space in there?  As you know I’ve been looking for a dead simple SBC solution for a while now and the pi4 worked out OK for my last 2 cabinets but if I was going to build one for myself I don’t think it would cut it (yet) because of the severe limitation on the games it ran at full speed.

Anyway, seeing this come together has inspired me to build… something.  Not sure what yet but I cut 2 side panels this afternoon (58 degrees!) and routed the t-molding slots.  I have no idea what I want to play yet but I’ll figure it out…

[bobbyb13]

I like the adventure of just cutting some panels without a complete plan yet!

If you can find a consumer tube with a yoke that has 2.4 ohm or less resistance on the inner windings (red-blue wires) then I suggest you build yourself a vector machine.

I have been wanting to try a Pi build for a while now actually and making a Cosmic Chasm seemed like a good place to give it a shot.

One of the guys in the group making all this cool vector stuff built a Pi image that is pretty much a plugNplay thing to use with a vector monitor and a USB-DVG board to be able to run AdvanceMame and just play vector games.
Here's a helpful link for anybody who might want to give this a shot.
https://www.vectorheadarcade.com/post/usb-dvg-installation-guide
If you roam arou dthe site there is a pile of awesome stuff those guys are working on.

I need to open up this 24" consumer set I have and see what the yoke values are.  For a few years now I have been saving it to build a horizontal vector machine but I really don't even know if the yoke is a good candidate for rewinding yet.

Maybe I should start on that tonight and document the entire thing here- start to finish.

[bobbyb13]

As threatened, I am going to document hijacking an old school consumer CRT set for a vector monitor build.

Just like Arcade Jason has said (and he was the guy who inadvertantly caused me to sack up and try it) it is NOT that hard!

So, the magic really is finding yourself the right set to begin with.
I have been hoarding these things for a while because I refuse to play all these old games I love on LCD and there are so few monitors available out here (currently this reads as-none) that I knew I would be building these things from parts I can buy from the mainland

So I am ahead of the game here in that I have a stack to crack open and try.

My experience thus far has been the crappier the set, the better.
Sony sets (except maybe unless you can find a really early one that has a single focus flyback- and even that might not work) are out.

The thing I have been saving for this mayhem (and I have only seen one on the island this size) is this:



Like I said, the crappier the better, but this is actually what I would think was a good generic TV back in 1994.
And American built!



Straight outta Indiana!

This happens to be a steep angle tube like an original WG 6100 but this is actually irrelevant for our purposes.
The key is that you find something that has as LOW a resistance/impedance value on the horizontal windings as possible.
Original Wells Gardner 6100 yokes have values somewhere in the neighborhood of 0.6 for vertical windings (yellow and green wires) and 1.2 for horizontal (red-blue wires.)
This thing happens to look like nothing I have seen before- and I have had dozens of sets of various sizes open to look.
Typically the yoke wires are soldered on and the board side has connectors- but no.


For whatever reason this thing is reversed- but whatever.

And like I said ~about~ those values for the original ones.
I don't have a meter to read Henrys, so there you go.

The components that Barry Shilmover is now making EDIT for accuracy! (the reproduction Amplifone deflection board and HV units) are built better than the original parts and have enough adjustment on them that there is a broader range of yoke values that are acceptable and won't blow up your boards.

Anyway, the key to a yoke that will work is finding one that has as LOW a value on the horizontal/inner windings as possible, because the internal windings are wound saddle style and let me tell you, unless you are Buckaroo Banzai, you are NOT rewinding those things.

Find a TV with a yoke with a value of 2.4 or less and you will succeed.
I have rewound ones from horizontal resistance values of 2.4 to 2.0 and they have all worked.

When you rewind the ferrites for the outer/vertical part of the yoke you will probably wind up with a value somewhere between 0.6 and 1.0 for resistance- and that is where you want it to be.
Here's a link if you haven't seen this awesomeness from Arcade Jason yet.

https://www.google.com/search?client=firefox-b-1-d&q=arcade+jason+yoke#fpstate=ive&vld=cid:47e66c5c,vid:Ci9qiGVMF7s

His youtube channel is a rabbit hole for me because his stuff is both informative and funny.
Set aside some time if you like a little comedy in your hobby.

The yoke in this thing is looking good thus far as regards resistance value.



But this is one weird yoke.  I haven't seen one like this yet, and I'm not sure I can even get it off the neck because I am not seeing the standard screw collar here either.
yeesh.

Either I'm stoked or I'm boned.  I'll need to dig into this thing a bit.
I have heard people rewinding the ferrite pieces without taking the yoke off the tube so that they don't need to mess with alignment/convergence but we'll see if I can do that here.
On the other one I had success just by marking how the yoke and rings were sitting to begin with and just lining them back up.

I'll post again after I have learned something!

[Mike A]

Nice work.

Vector games should be played on vector monitors.

[Arroyo]

Damn Bobby you don’t sit still for a minute!

Looks like you beat me to posting this stuff.  Nice job.

The boards from Barry are Amplifone clones.  Nice part about Amps are they have pots that allow you to dial in the height and width as they vary the current flow through the yoke.

In an ideal world your yoke winding inductance and resistance would be almost the same (they are very close on 6100’s and Amp’s).  This is because a vector draws free form in X & Y and therefore you want the chassis to be able to draw just as quickly in on direction as the other (a little more inductance for X due to the longer length) .  You are able to have variance in those two as Jason pointed out because of the pots on the deflection board.  With a 6100 you don’t have pots so that is where having the values close to the original yoke becomes important.  If it’s any help Amp yoke readings are here:

Vertical - 0.6 ohms
Horizontal - 1.2 ohms

Vertical - 0.625 mH
Horizontal - 0.796 mH

I’ve come across a yoke like that one before.  They just put a connector to the terminals rather than soldering them.  It looks like your ferrite cores have a cover at the base.  There should be a way of removing that to more easily get to the ferrites. 

As for taking it off or winding it in place.  Yeah ideally you remove the ferrites in place without moving it, but as you mentioned I think it’s just as easy to mark it and remove.  The one argument for keeping it in place is convergence strips.  If you have any either on the tube or embedded in the yoke then you will want to take great care in making sure they don’t move.  Getting them back if you didn’t know where they went is the real trouble.

That tube is valuable if it is indeed a 100deg.  Many mainland folks would be happy to swap you for a standard 90 degree 😁.

Keep up the good work.

[Arroyo]

Oh and here’s my setup that I’ll be posting about at some point:

[Mike A]

Nice.

This is good stuff.

[abispac]

Keeping my eye on this one, it would be so freaking cool to build a vector monitor and play vectrex games on it. I still trying to figure out, so im reading over and over, please guys keep it coming. Thanks
OMG plaring starwars on such monitor.... 

[bobbyb13]

Arroyo!
Good to see you come up for air.
I appreciate you chiming in on the mess I'm making. It is obvious I have a little editing to do and I'm happy to hear what I am doing wrong at any point.
Of course if I knew that you were at this same endeavor I wouldn't have bothered posting my hack work!
 
I did actually have some time to do a little harness building today for the next one. I am curious to know what the wiring situation for those heatsink cooling fans is since the set I have of course didn't come with any instructions.

I'll take a shot at posting some pictures here once my work day is finally over.

[Zebidee]

Nice work Bobby! Was hoping to see a vector build post. I've been lurking on Barry's FB group and noting your participation.

[bobbyb13]

Sadly at this point my participation is mostly begging for help to get AdvanceMame to recognize my spinner!
Thanks Andrew.
I've been wanting to get at this for a while now.

Of course this may be an exercise in futility because this otherwise choice tube has a bonded yoke and no adjustment rings.
Not the best thing to be working with to try to instill confidence in anyone really since every other set I have cracked into was far easier to deal with.

But anyway...
Last night I stayed up far too late (can't help myself) and rewound the ferrites for this yoke.
I had no choice here, but leaving the yoke on is a great idea I think.
This thing had some spacers that both hold the ferrites off the yoke body and also direct how the windings can group.
I did not rewind this exactly as it was done originally, but opted to space the wire out as evenly as I could.
It may not work right like this but I will try it this way first.



After I had it all back together the resistance value of 0.6 I got was what I have seen work on my other build so I figure maybe it will be ok here too.

I spent a little time building harnesses and doing layout while I was stuck on the phone today.
Hopefully the vga connectors I ordered get here soon, as the earlier version of the board I have here for this build doesn't have a vga plug, so I need to solder one up.

The rest of these connectors are really straight forward molex kk style ones and I believe that the kits come with all of them required.
Like I said, both of the ones I have here I got second hand  (impatient waiting on the one I got in line for a while back!) so I can't be sure of what they were originally shipped with.

If you include the power supply, there are only 4 parts to this whole thing apart from the tube and yoke.



The degauss posistor board, which I have opted to run line voltage from my power distribution block.
It should be fused on the line side of course.  I have used a 5 amp fast blow and had no issues.



The high voltage unit gets its power from the deflection board through a 4 pin molex connector and has the neckboard (which I suppose I should have taken a picture of) already attached here.
There is also a 3 wire connector out to feed the deflection board fans but I am not sure if the third wire is a signal wire or what.
Help Scott! 



It is also the only one of the components that needs to be grounded to the tube.
There is a through hole on the front corner to solder to.

The deflection board has the most activity on it.
I don't believe the original Amplifone boards had any active cooling system so this beats that already, and I know the current versions of this kit include an even more massive heat sink but no included fans.
Not sure what temperature those transistors are happy to run at for a long time but I have read 175 F on the other newer heatsink right next to the transistors so that could probably benefit from a fan too, even if it isn't attached directly to the heatsink.

The other parts of these boards appear the same no matter which version you might have.



Everything is very well labeled so this is rather redundant, but...

Connector for HV unit is the one upper left.

The one lower left goes to the neckboard.

It's important to make sure you follow the printing on the board mask because I know on one of my kits the wires cross in that harness.

The connector in the middle of the board is for the yoke, one pair left, the other right.

Video signal is the one on the bottom middle.  I need to dig up a diagram I found of how this should be soldered to that vga connector to mate with the DVG properly.
When my connectors show up I will post that too.

Finally is the power header on the lower right side.
This was the part that scared me the most because I didn't want to blow anything up- and I immediately thought I had.

It requires +24VDC and -24VDC.

The board literally uses one leg to send the beam one direction, and the other leg to send it the opposite way.
I believe it is the positive sends the beam right on the X axis and up on the Y axis and the negative sends it left on X and down on Y.
If you think about it, that makes troubleshooting issues easier as things will fail in halves or quadrants.

Cool stuff.

So really all you need is two 24 volt DC power supplies.
The +24 one you wire normally and you swap the positive and negative wires from the other supply to feed the -24 side.
That is it.
You don't need any crazy toroidal 50VDC center tapped transformer to get this to run.

That said, still the first problem I had was outsmarting myself and mixing the wires on the harness and sending +24 V to both sides of the board.
Remarkably (and thankfully) it didn't kill anything.
It wouldn't, it just sends voltage that the board can't properly amplify the signal of to get full deflection.
If you sent a lot more than 24VDC then you might blow stuff up of course (I recall one reply from Fred saying that you could probably actually send the board 35 VDC and it could handle it.)
More voltage, more ability to amplify it and make for a wider image.  No idea what the threshold for that might be for screen size, but I do know that Barry said he intentionally was sizing things on the later board kits to mimic the ability of a 25" Amplifone.
That is the dream of course.

And all this explains why the first image I saw was just this cute little 1/4" wide winking ball in the middle of the screen.
The board was unable to deflect anything one direction because I had given it the wrong polarity of voltage to do so.

As far as the other axis being squashed, I hadn't realized yet that I had wound the ferrites in opposite directions either.
Hence the little ball of no deflection.

And even after I had corrected my first obvious failure and finally sent proper voltage to the deflection board, all the cross wound yoke did was squash the picture into this 3/8" horizontal line in the middle of the screen.
It also made the flyback screech to raise the hair on your neck so I shut it off fast and tried another yoke, which I accidentally wound properly the first time!

Now I know.

As soon as my other pair of power supplies and vga connector get here I will finish the bench build of this thing and fire it up and see what we get.

Should be interesting given this yoke situation.
 

[Mike A]

Of course if I knew that you were at this same endeavor I wouldn't have bothered posting my hack work!

You are doing good work.

I love to see this kind of stuff on BYOAC.

[PL1]

There is also a 3 wire connector out to feed the deflection board fans but I am not sure if the third wire is a signal wire or what.
Help Scott! 

You guessed right.  It's the tachometer sensor wire.



A protective shutdown circuit plus a 3-pin fan is a good design choice for applications like this where overheating due to not enough air flow can cause severe damage.


Scott

[Zebidee]

If the signal wire is disconnected or not used, the fan simply runs at full speed.

[bobbyb13]

It is comforting to know that the people who have replied here are seeing this fiasco unfold.
If anyone can keep me in line it will be you gents- so thank you all.
 

Now if we can just find somebody who knows Linux well enough to help me fix AdvanceMame to see my spinner...
 

I had suspected that with regard to the fan wiring but wasn't sure- so thank you both.
At least I know what to do with that now.

It appears that the other parts I need for this won't get here until the weekend.
That's an adequate amount of time for me to get REALLY antsy about it.

I guess any free time I get this week might be spent putting bulbs in the Simpsons pinball-
 

[bobbyb13]

Well, been so damn busy that no bulbs went into anything.

I DID however make time to try AGAIN to get the spinner for this beast sorted out- and I learned something.

My iPAC 2 board is knocking out the spinner- in game- somehow.

Not sure why, but when I started with a fresh SD image for the Pi, added the few lines of code I still don't fully understand to the advmame.rc and advmenu.rc files, and then fired it up with only a plain old USB keyboard, and the SpinTrak via USB-

It worked.
The freakin spinner worked.
Not just in Chad's awesome vector game menu the way it has all along, but even in a game of Tempest.

I remember seeing in my otherwise useless meanderings around the interwebs that the iPAC allocates for a joystick (not just a secondary keyboard)- which could be breaking my spinner in AdvanceMame somehow.

When I sort this out (i.e. actually learn something more about how this crap works) I will post again.

Oh, and my power supplies showed up, so maybe I can get the other vector build fired up before the weekend is out.
 

[Zebidee]

My iPAC 2 board is knocking out the spinner- in game- somehow.

You might be having some device ID conflicts?

Someone (@drventure) created a controller remap utility to help analyse and sort out these issues. I haven't used it myself, but looks like it may do the job:
http://forum.arcadecontrols.com/index.php/topic,108767.0.html

[bobbyb13]

Biggest issue is that I still understand almost nothing about Linux.
I have already suffered this pain with Windows enough times that I can sort these things out for a groovymame build.

I have been back and forth with burning a fresh image to this sd card so many times that I sometimea forget where I am.

The fact that I don't know how to meander through files on the Pi makes finding and altering things I need to or correcting things I break frustrating and difficult/impossible.

It just occurred to me that now I finally got those two elements ov I-PAC and Spintrak to play nicely with the Pi that maybe at this point I can alter what else I need to through the in-game tab key pop up.

I reflashed the I-PAC with new firmware and I forgot that I need to register those button strokes with AdvanceMame again (as I have two keyboards connected now as far asthe software is concerned.)

The joys of custom builds.
This part I need to document here step by step also when I finally get it all working- which hopefully is today!

[Zebidee]

Biggest issue is that I still understand almost nothing about Linux.
I have already suffered this pain with Windows enough times that I can sort these things out for a groovymame build.

I have been back and forth with burning a fresh image to this sd card so many times that I sometimea forget where I am.

Ooops yes, you are slicing Pi. I probably know even less about Linux than you 

If you document your efforts, then at least you (and the rest of us) will know how to do it next time 

[bobbyb13]

And that's where I'm at!
Man, this has been super frustrating but I just pulled it off.
The machine actually works now- 100%.

The last time I refreshed the sd image and fired it up with the I-PAC connected and the spinner plugged into it, this appeared on the HDMI monitor (which if you have a prayer of making this happen you MUST have during setup at least.



I was encouraged because it meant that the spinner was actually registering as a device where I needed it to finally- even if it had a monstrous deadzone?!
Holy crap-

And sure enough, it was working, even if it acted like I drank a gallon of espresso 15 minutes ago.

Mind you, I wasn't getting control button function anymore, but since I had it before I was confident I could get it back.
And since I was on a roll anyway, and the ultimate goal for this machine is that it COULD play Cosmic Chasm, Tempest, and with a control panel swap Quantum also, I figured why not push my luck and see if I can get a trackball recognized here too while I was at it.
At least I would know if I had all of my configuration stuff sorted and can get on to finishing the cabinet itself.

So I built the harness I knew I needed to attach directly to the I-PAC.


What had occurred to me this morning when I woke up ridiculously early and was perseverating on the deadzone issue (which I haven't been able to fix in the \root\boot\cmdline.txt file yet) is that I might be able to get the deadzone fixed in the in-game menu (via the magic 'Tab" button.)
It then immediately occurred to me also that my lack of button response wasn't device conflict, crappy harness connection or screwed up I-PAC configuration, but that since last time I started over with a fresh sd image for the Pi I had NOT gone back in to map my buttons again.
Yeesh.

With this revelation and my maxed control panel setup I decided I should give the whole enchilada a go before my rather mudane tasks the rest of the day.
I got the trackball temporarily plugged in for testing and cleaned my button contacts just for good measure and plugged it all in.



I fired it up and nobody could be as surprised (pleased!) as I am that it actually did what I wanted it to.
Hallelujah
Finally.

How many weeks of Pi torture has this been?!

So, now refreshed, I will go get through my chores for the day and then sit down this evening to outline the ENTIRE process of going from CanaKit to functional vector machine and list the parts I used and steps in order.

And then HOPEfully somebody else can do this in one straight shot without all the tears, loss of hair and desire to throw things.

[bobbyb13]

Once again, as threatened, I am posting some more.
Warning:

Not only is this next bit pic and link heavy, if you are already well versed in the art of Pi you will find it extraordinarily boring- so I apologize now.

The other caveat is that I am so Pi stupid that I am using a PC to get all this crap sorted.
Working on the Pi itself to do some of this stuff makes my hair hurt and you can do everything except the finer points of controller details on a PC via network connection before you even fire the Pi up.

If you are like me and can hardly tell your ass from a hole in the ground as regards the Pi do your self a HUGE favor and just buy a complete CanaKit.



It has everything you need to do this properly and easily.
Follow (only!) steps #1-3 in the QuickStart guide directions to put it all together and get to where you need to be.

If you actually know what you are doing you can probably just follow the install method found here (which you should have at hand anyway even if you are counting on me to not destroy things.)
EDIT!
For very understandable reasons, Barry has nuked association with some portions of the USB-DVG project and the link previously here leads to a 404.
Here is the still functional google drive files uploaded by Mario M to be able to accomplish this feat of vector wonderfulness!

https://drive.google.com/drive/folders/1KNNcKDJn_mh87HGbJHj9IvhJUWHlr3CX

There is a pdf in there detailing things I go over here also, but I have added quite a bit throughout this post actually.

Also, if you missed it before, I can't remember if I linked to this one yet, but you REALLY need to watch Jason Kopp's YT bit here also.



If for some reason you think sticking with me is a smart move then read on.

I started fresh here with a new Pi for visuals (and this one will actually be what I use for my horizontal build  ) and it is remarkable how quickly you can get all this up and running.

The reason you don't even hit step #4 is that we are going to nuke the image that the included SD card comes with and write one that was written specifically for using a USB-DVG wo play vector games.

Also, download the Raspberry Pi imager so you can do this easily and follow along step by step here.
I probably have an old version here since I have been torturing myself with this for so long but I imagine it works the same way.

https://www.raspberrypi.com/news/raspberry-pi-imager-imaging-utility/

Get your SD card into a USB reader and plug it in and then fire that Pi imager up and you should see this thing.



Be prepared to have a few Windows Explorer windows open for file management because it gets a little nutty if you don't want to just copy and paste this stuff from what I will post here.

N.B.
At ANY point in this process you might/will see all kinds of warning pop ups randomly appear because Windows is pissed off that you are using something else, and the PC will warn you that it can't understand what is going on.
It is important to IGNORE these alerts and just close them without any action otherwise- or you can possibly confuse the Pi imager and break things.



I have seen it at numerous times during this process and had it occur with different types of warnings.
Hit the magic 'X' and disregard them all.

Ok, so when you click on "Choose System" on the imager app you will see a list of possibilities in another window and you need to scroll all the way down to see this:



You want to pick "Use Custom" of course and then you will get another window to pick what you want to write to your card.
Dig around until you find the image you need to use, which currently is this:



You already read the REAL installation guide I linked above before and downloaded the image you need to use, yeah?!
If somehow you are lazier than me (unlikely) and nothing has been moved than you can probably still find the latest version in the above link (and again right here below as before) to all of Mario's google drive files with regard to the Pi.

NEW STUFF!
Mario has also produced a new image for use the new Pi 5 so make sure you are using the right disk image for your build.

https://drive.google.com/drive/folders/1KNNcKDJn_mh87HGbJHj9IvhJUWHlr3CX

Once you have found where Windows tried to help you hide your Pi image from yourself and chosen it, you will need to choose the target for the image.
Mine looked like this and I imagine yours will look similar even if the drive designation is different:



If you have followed my directions and somehow this has all still worked you will see it first writing and then verifying the image with progress of a cute little green bar.
Be patient, it does actually take a little bit for it to start and then get through the whole thing.

Patience there should net you this:



At that point you can close this all up.
Close all windows that could have anything to do with the SD card (including any more stupid Windows alerts that may have popped up) and when you are finished you can just pull the card reader out.
Windows can't see the SD card as it is (it won't show up as existing in the far rigth side of the tool bar anyway) so you won't break anything without ejecting it first.

Enough for this round.
I need to upload more pics anyway.

Oh, and this all works much better if you are listening to The Cramps and/or The Chats while you are at it.
 

[bobbyb13]

While I'm waiting for the risotto to simmer it's way to greatness I figured I should have some more wine and keep going!

In the interest of using up more of the space of that which Saint hath built, I will post more crap.

Alright, so you have the USB-DVG Pi image written to your SD card already so it is time to put that thing in your Pi and get it tied into the network that I hope you have available to you at your house.

We are blessed to live in the middle of frackin nowhere so the only reason I can torture any of you with this stuff is that we get radioed some internet from a few miles away- hence the weird router cube thing.
Pardon my ramble, but plug your Pi into an available port on your router, plug your transformer in and click that little button to turn your Pi on.



Give it a minute to understand that it is alive before you try to log into it through Windows.

If you open an Explorer window on the PC you should be able to type this in and actually get somewhere.



My system here thinks about it for a minute, but it gets there eventually.

This is the folder into which you want to copy your taboo elements.
Don't ask, don't tell.

For the rest of our network enabled adjustments the rest is important.

The following bit is the code that someone FAR smarter than me came up with to add somewhere in the Pi- and it actually works.
If the person who wrote this is reading here please message me and I'll send you something mea culpa style for listing your work unacknowledged!

Do the copy and paste thing with the following mouse additions in the manner I set forth in the following if you want to not agonize over this for (at least) weeks like I did.

device_mouse raw
device_raw_mousedev[0] /dev/input/mouse0
device_raw_mousedev[1] /dev/input/mouse1
device_raw_mousedev[2] /dev/input/mouse2
device_raw_mousedev[3] /dev/input/mouse3
device_raw_mousetype[0] ps2
device_raw_mousetype[1] ps2
device_raw_mousetype[2] ps2
device_raw_mousetype[3] ps2
input_map[p1_dialx] mouse[0,x] mouse[1,x] mouse[2,x] mouse[3,x]
input_map[p1_dialy] mouse[0,y] mouse[1,y]mouse[2,y] mouse[3,y]
input_map[p1_trackballx] mouse[0,x] mouse[1,x] mouse[2,x] mouse[3x]
input_map[p1_trackbally] mouse[0,y] mouse[1,y] mouse[2,y] mouse[3y]

Easy, ain't it?!
Copy and paste EXACTLY this to a notepad file that you can pull up and copy and paste all/portions of (as detailed here shortly) into the proper places in your Pi image files.

And whoever wrote this is brilliant, in that they accomodated FOUR different ways that both you and your Pi could screw this up.
Bear in mind, I am also setting up to use a trackball on a separate panel also and this allows for that.
If you aren't going to use one these lines additions will not break your build as far as I have seen.

Ok, first addition (that I believe helps but at least doesn't break anything?!) is to advmame.rc


Click the little 'one directory up' button while in the rom directory to get to where you need to make the changes that make all this stuff actually work.



You could always type the location of the advancemame.rc file into the nav bar too, but that is for industrious people who know what the hell they are doing here.

Now here is the part that MIGHT be redundant, but since it doesn't break anything apparently and I don't know for sure you can NOT do this and have function without drama.
I have been back and forth with this so many times that I don't have the emotional fortitude for another round as a 'test.'
Anyone who knows better (id est, anyone who knows a Pi and AdvanceMame better than me) please chime in so that we can do away with this step of adding to advmame.rc if it is ok to do so.

Here is what I had up for explorer windows to make this happen.



Open that advmame.rc file, and if you haven't toyed with this on your PC previously then you will probably need to tell it to open these files from your Pi using Notepad.
I literally copied and pasted the entire spinner settings file I show to the end of the advmame.rc file.

Dig in and get that addition made.

Save your changes and close the window.
Done.

Next is adding to advmenu.rc

Navigate into that and roll down until you find the first mouse device line.



The line that shows mouse_device none needs to be

mouse_device raw

The eight lines following the raw_keyhack line in the Pi on the left also need to be updated as shown with what I have highlighted in my spinner settings from the window on the right:



The last four lines of these additions needs to be what is shown in my spinner settings on the right also, so for the sake of clarity scroll down until you hit the i's in your code and then add/overwrite so that what you see looks like this if it isn't already what you need:



As far as things to alter in your code via PC/network connection, that is it.

Don't forget to save, close all the windows you had open in regard to this surgery and then you can shutdown your Pi and unplug it all in preparation to install in your cabinet.

Things to note as you continue:

*As far as I can tell the Pi actually remembers where you had things plugged into it via USB (unlike Windoze, which loves to rearrange crap for you randomly) so make your life easier and be mindful of EXACTLY where you plug things in for future reference.

*I have had ZERO success using a wireless keyboard/mouse combination device in configuring the Pi once in the cabinet.
Do yourself a favor here and just use a plain old corded keyboard for controller configuration.

*I have NO idea if AdvanceMame fails to launch without having an audio device connected (the way MAME has for me in the past) so do yourself a favor here also and just have something plugged into the Pi that will make some noise.  The Pi has a 3.5mm stereo jack so make it happen.
It also helps because you can hear Chad's menu wake up when it starts and hear keystroke and spinner/joystick indications as you navigate and exit too.

*You MUST have a display (in addition to the vector CRT you have hopefully already acquired or built and is connected to your USB-DVG) plugged into HDMI port 0 on the Pi while finishing up or you simply can't do it.
Unlike any other MAME build I/you may have attempted, NOTHING will show up on the CRT except what the vectormenu or game are doing.
When it comes to getting AdvanceMame to register your control panel inputs (via the 'Tab' key while in a game- or to adjust things later on) the secondary display is imperative.

Ok-

Now that you have your Pi all set up the next thing is to take care of your controls.

You can be cheap if you want to and try to use something else (and a properly configured Arduino board will suffice here of course- but I am not that smart yet PL1 !) however the following detail is written in reference to an I-PAC and a SpinTrak because I can attest that the following just simply works.

Andy at Ultimarc offers top notch customer service and quickly replied when I asked about what my build was gagging on so that I could fix it.
I've been buying stuff from him for years and he always helps.

I have also bought really nice parts from Randy at groovygamegear (whose support is equal to Andy's) and I am sure that he has a board and Twist that will do exactly the same, although the configuration process will probably differ slightly.
If I can get the time to buy stuff from Randy also I'll test that out too and add notes here if anyone would like.

So, before you fire up the machine with the Pi connected and have fits like I did, let's get your control interface set up properly.

One of the cool things about the I-PAC is that there is are a number of utilities to get your board to act the way you need it to.
If you head to Andy's website you will see a few downloads available for I-PAC boards.

https://www.ultimarc.com/control-interfaces/i-pacs/i-pac2/

I decided to flash the one in the Cosmic Chasm cabinet with the 'Mixed Mode' firmware as it referenced Retropie and I had tried other iterations with no success.
This MAY be the only way it works, but that may not be the case for you.
Again, I am nervous to break what I have that works so I am not starting over to test something else without a pretty good bribe to do so.
 

Andy's app and directions are very straightforward so I won't bore you with reiterating.

And after I had flashed the board I mounted it to the panel and wired in my controls appropriately.
There is a header on the I-PAC to plug both the SpinTrak AND trackball (I have a regular old Happ here) into at the SAME time- and they both work.

This is absolutley awesome from my seat since it allows me to play Cosmic Chasm, Tempest, Tac/Scan with this panel and also Quantum with a panel swap.

Love this.

For the sake of being thorough, this particular cabinet was built solely as a vertical game unit.
X Y monitors don't care about orientation really as they are deflecting from center so you can always be smarter than me and just build a wider cabinet with a 25" tube horizontally oriented and be able to play EVERYthing in one cabinet reasonably well as long as you get per-game configuration proper.

I don't do much reasonably it appears.

I need to revisit the vertical vector game possibilities here.

[bobbyb13]

I realize that I am in a timezone inconvenient for those who aren't insomniacs or early risers but I sometimes feel like I am in an empty auditorium-
Even if there have been over 1,100 views more than comments here!

Suckahs-

So the saga continues.

I fired up my build after the LAST time I burned a new image here and although I can ACTUALLY play Tempest now...

NO Cosmic Chasm again!

The rom files are there but there is no entry in vmmenu- again.
Not sure why this doesn't show up when there is a rom for it present (two versions of it in fact) so back to configuration we go.
I suspect that Chad's menu config writing script is only set to check the first time it is run.

I have however seen in Chad's quite rad vector game menu that you can enable/disable games there once the menu is running on your vector display.
I need to check it again.
Can't remember how to get to it at the moment, and when I am doing this stuff I am going back and forth between my laptop and network in the house and the machine- which is in the workshop 100 yds from here.

After I have had a bottle and a half of wine it begins to get a little fuzzy sometimes.

I dragged the Pi back down here anyway, as I know that I can add a line to the vmmenu.cfg file (for some reason it isn't an '.rc' file in this particular instance) and force the Pi to realize it is there.
If I knew more about how Linux works this wouldn't be so confusing and cryptic- sorry- but at least we know what I am doing works.
 

I may go try it again before I go to sleep because I can't help myself.

No, of course I will try it again before I go to sleep.

[Mike A]

I am watching, but I have zero input to offer on working with a Pi.

I am just here for the vectory goodness.

[bobbyb13]

Appreciate the moral support Mike!

The Pi is like anything else really I guess, but anyone who says it is easy to work with has been working with them for a while obviously.

And as I thought, the lines referencing Comic Chasm were missing from the total game list in vmmenu.cfg
My only guess is that when Chad was writing the code initially he just missed it.
When you include all the different versions of various games and then add to that all the Vectrex ones that are in there also it winds up being a LOT of titles- so it is easy to see something as obscure as CC being missed.

I added a line for CC following the convention I found in that file and when I fired it up again, it worked.

EDIT!!  Here are those lines.

Cinematronics|Cosmic Chasm|cchasm|cchasm
Cinematronics|Cosmic Chasm|cchasm|cchasm1

Got to play some CC, a little Tempest, even a few games of Major Havoc - just because I could.

However, the oddness of the Pi continues in that the in game configuration doesn't always do what I expect it to and the spinner wasn't registering in Tempest Tubes for some reason.

More weirdness to run down.

[Zebidee]

I'm keenly following your progress Bobby.

Although I'm more interested in the vector side of your adventures, your trials and tribulations with the Pi are also interesting to follow. I can see why you went with advancemame and Pi, it is really all you need. For now, I just read and let it wash over me, rather than trying to understand it all. 

One of these days, when I've worked down my wishlist of projects, I'd like to similarly do a vector cab, possibly with a Pi. So this thread will be a useful resource.

[bobbyb13]

The intent was to make this so that someone could sit down with the parts, scroll through this thread, build the thing, and have it just work- minus the months of agony!

As always, the devil is in the details.
I'll be stoked to hear from even one person at some point who followed this and found success with it.

The only bit left is particulars of setting up power for the monitor at this point really.
I have a lot going on this week so it may be a few days before I can add that here.

[bobbyb13]

20+ hours of rain gave me a reprieve from some crap and a little time to mess around in the workshop.

So the Cosmic Chasm machine is going, and also plays Tempest and Tac/Scan well!
No blowing fuses any longer, boots without issue, controls are working well, BUT-

The image on screen still isn't what it needs to be.
I have been able to play with the yoke a bit and get purity to be OK and in the center of the screen convergence is fine but as you head to the perimeter it goes to hell.



I messed around with the position and orientation of the neck rings, position and tilt of the yoke, done a dance and pleaded with the vector gods and I still can't get it good.
I'll play it mind you, but at some points in some games it REALLY does not have the vector sharpness that the initiated know and love because the images shadow in various colors sometimes.

Now THIS part is making me nuts.

Tomorrow hopefully I can go buy a stack of refrigerator magnets (the thin flexy kind of bits you can cut into strips- it is what I have seen on tubes anyway) and begin experimenting.

It's comforting to know that at this point, with all new components laid out just so, and everything getting the voltage it wants, that I can play around with how the beam is defelecting at this point and not be afraid that I am going to blow something up.

What's wrong with just warping the trajectory of an electron gun beam a little, right?!

Seems like really the only things to be concerned about with Barry's Amplifone repro kit is to-

1. Make sure your deflection board is getting proper clean voltage.
2. The thing has room to breath (because it DOES get hot still- even with that massive heat sink on it.)
3. The yoke you are using doesn't require that you be at the maximum of any of the adjustment pots on both the USB-DVG and the deflection board, as I imagine that this would indicate that you are stressing the thing and that will most certainly affect longevity of compopnents.

My only real concern with the thing at this point is the question of HOW hot is TOO hot for the deflection board heat sink?



It appears to hover around 150-175 F most of the time but when I was working on yoke and ring adjustments and had a test pattern up for a few minutes the thing got HOT.
REALLY hot.
Like over 200 F

Right after a game of Tempest it is frequently at 185 F right over just one of the transistors on that heat sink but the temp drops quickly, and even then most of the heatsink never gets over 165.
The operating temperature of these things was quite a shock.  No wonder so many vector monitors blew up way back when.

Not ONE of them really had adequate design/layout to accomodate cooling this stuff and I think the Electrohome G-08 was the worst because it was even more at the mercy of flaky voltage from the game boards than even anything Atari had made.

No matter what I may hear from anyone about component heat tolerance I will be putting a fan on this heat sink also.
Only question is should it be mounted directly to it or not and should it be pushing or pulling in relation to air movement across the heatsink.

Also looks like I will be putting some good size vent holes in this cabinet.