Weird sporadic collapse on Tempest UR

terrorinstinct

New member
Joined
Jun 17, 2025
Messages
27
Reaction score
12
I've become responsible for reviving a Tempest upright that burnt out (literally, as in a plume of smoke was rising from it) on location. After cleaning up the smoke mess and rebuilding the monitor, got the usual half-collapse on one vertical half of the screen. After completely rebuilding the monitor with an Alan-1 XY board and TUK, we fixed the half-collapse and started seeing a very weird "diagonal" collapse where both the top-left and bottom-right corners would kind of collapse towards the center in scenes with a lot of bright vectors, including the title sequence and triangular tube scene in the attract loop. I don't have video of that at the moment (working on getting it from my fellow tech) but some work that he did on the game board (again, working on details, but definitely including replacing the six X/Y pots) this was reduced to more of a "bounce" during the title sequence when the "TEMPEST" title is at its brightest.

Things that we've done so far:
  1. Totally rebuilt the monitor. HV has new flyback, new caps, new resistors (except for a couple oddball values we didn't have in stock and weren't roasted on the original) and everything else transferred to a fresh PCB from Arcade Jason. Deflection transistors and XY board totally replaced with upgrade kit from Alan-1. Confirmed correct behavior by hooking up the monitor harness to a known-good Star Wars UR. At this point I can say with confidence that whatever the issue is, it's not the monitor.
  2. Unidentified parts swaps in the analog output stage from another Tempest boardset. I didn't do this work and the other tech didn't leave a written worklog. I'm following up with him and when I have more details I'll edit this post with the scoop. EDIT: Most devices on the analog output section (with the exception of one DAC) were socketed already, and the XY op-amps were swapped (X->Y, Y->X) to no effect.
  3. What I know for a fact the other tech did: replaced all six XY vector adjustment pots on the vector board with new ones from APAR. After (2) and (3) the partial collapse stopped appearing in the gameplay parts of the attract loop and reduced in severity during the title sequence.
At this point I'm at a loss as to what the issue could be, and loath to deploy the parts shotgun for something as finicky (and expensive) as a Tempest. Any help from the vector gurus here would be greatly appreciated.
 

Attachments

  • DSCF3843.JPG
    DSCF3843.JPG
    629.9 KB · Views: 1
Last edited:
This is a perfect example of why there's no need to waste money on ArcadeJason and ALAN-1 parts. The original boards and frame transistors can all be restored on these. You transferred all of the deflection board parts to a new PCB, without troubleshooting what components were actually bad.

The issue is bad spot killer caps, causing the spot killer to kick in prematurely. See this thread, and the ones it references in post #5:


The reason you didn't see it when you connected the monitor to a Star Wars is because Star Wars draws the graphics more evenly across the screen. Tempest has a lot of instances where it's drawing vectors very unevenly, with most of them on one side of the tube (like when the scrolling TEMPEST logo happens in attract mode). This will cause the spot killer to start to kick in and bend the picture, when it shouldn't, when the spot killer caps get old.
 
This is a perfect example of why there's no need to waste money on ArcadeJason and ALAN-1 parts. The original boards and frame transistors can all be restored on these. You transferred all of the deflection board parts to a new PCB, without troubleshooting what components were actually bad.
Deflection? The HV power supply needed to be rebuilt. The reason it needed to be rebuilt is because it exploded and caught fire. There was literally a brown crater where one of the power resistors used to be. If you can "restore" that, please record it and send it to the Center for Inquiry, as there's a standing prize for documented miracles. Hence, the Arcade Jason PCB. Everything else on the HV was tested for obvious badness (ohmed out/diode test) before being installed. The Alan-1 boards were already on hand for another Star Wars project and were used in the name of expedience.

The issue is bad spot killer caps, causing the spot killer to kick in prematurely. See this thread, and the ones it references in post #5:


The reason you didn't see it when you connected the monitor to a Star Wars is because Star Wars draws the graphics more evenly across the screen. Tempest has a lot of instances where it's drawing vectors very unevenly, with most of them on one side of the tube (like when the scrolling TEMPEST logo happens in attract mode). This will cause the spot killer to start to kick in and bend the picture, when it shouldn't, when the spot killer caps get old.
Thank you. Poking my head back in the spot killer LED does flash for a brief moment during the title, at the very end when the TEMPEST is very bright and white. While it seems unlikely that the spot killer caps would be bad on a brand new Alan-1 you never know when tantalums are involved, I guess. I would still like to know what could be causing the sagging vectors other than "old" (less than ten years old) ceramic caps, though.
 
Last edited:
Deflection? The HV power supply needed to be rebuilt. The reason it needed to be rebuilt is because it exploded and caught fire. There was literally a brown crater where one of the power resistors used to be. If you can "restore" that, please record it and send it to the Center for Inquiry, as there's a standing prize for documented miracles. Hence, the Arcade Jason PCB. Everything else on the HV was tested for obvious badness (ohmed out/diode test) before being installed. The Alan-1 boards were already on hand for another Star Wars project and were used in the name of expedience.


Thank you. Poking my head back in the spot killer LED does flash for a brief moment during the title, at the very end when the TEMPEST is very bright and white. While it seems unlikely that the spot killer caps would be bad on a brand new Alan-1 you never know when tantalums are involved, I guess.

Send me the old HV board. Yes, burn holes can be fixed, see below.

If it's a new ALAN-1 deflection board, I would suspect a wrong value component was installed somewhere in the spot killer circuit. However if you can get a pic or video of the screen, that will give more info. Ideally you would want to put the game board on a scope, and use that as the display, to verify that it isn't anything about the game board (which it could be, but a scope will tell you in short order).

IMG_3480.JPG
IMG_3495.JPG
IMG_3507.JPG

zz_animate1.gif
 
We did put the game board on a scope and saw identical results to the monitor, but that was before the pots were replaced. I'll be able to post some video of the deflector sag in a few hours.
 
We did put the game board on a scope and saw identical results to the monitor, but that was before the pots were replaced. I'll be able to post some video of the deflector sag in a few hours.
Does that mean you see the weird collapse on the oscilloscope?
 
Send me the old HV board. Yes, burn holes can be fixed, see below.

If it's a new ALAN-1 deflection board, I would suspect a wrong value component was installed somewhere in the spot killer circuit. However if you can get a pic or video of the screen, that will give more info. Ideally you would want to put the game board on a scope, and use that as the display, to verify that it isn't anything about the game board (which it could be, but a scope will tell you in short order).
Okay, have fun fixing this. Note the LV2000, this was already "bulletproofed".
IMG_20250326_083614.jpg

Paying the tax for a new PCB just made more sense. Here's the video of the spot-killer from the attract loop:
https://limewire.com/d/DOlr9#a5CyN3Qmd4
Does that mean you see the weird collapse on the oscilloscope?
Yes.
 
Ok.

Got a video of what it's looks like on the scope? I'd like to see some of the details, for comparison.

It looks like there might be more than one thing going on. The geometry looks off, from the way the whole picture is twisted, which isn't normal. That could be a matter of the BIP pot settings on the game board, or it could be that the actual yoke needs straightening. However if no amount of adjustment to the BIP pots straightens it, that usually means a bad DAC. (And it's probably the one that hasn't been socketed and replaced yet.)

However it also looks like the spot killer is starting to kick in during the scrolling TEMPEST logo. Maybe in theory if there's a bad DAC, it could be cutting out enough during that sequence to be tripping the spot killer, but that's just a guess. Or there could be a separate issue with the spot killer. But I'd start with the game board adjustments first.

The best way to adjust the BIP pots is to flip it into test mode. You'll get a screen that looks like the one below. (Ignore the brightness artifacts in the pic, that's just the camera.) Note the line in the middle, above where it says BONUS ADDER, that says X I I I. You want to adjust the XBIP and YBIP so those four X I I I characters are all lined up with each other vertically, and are all evenly spaced from each other horizontally. Just play with the XBIP and YBIP pots, and you'll get a feel for how to dial it in.

If you can get those letters straight, the rest of the picture in attract mode should be straight, if the rest of the video section of the board is ok. (And then you can adjust the X and Y SIZE and CENTER pots as needed, so the perimeter of the box is around 1.5" from the edge of the tube, on all sides.)

If you can't get the X I I I characters lined up straight with the XBIP and YBIP pots, replace the other DAC, then see where you're at.


Test mode X I I I.JPG
 
Ok.

Got a video of what it's looks like on the scope? I'd like to see some of the details, for comparison.

It looks like there might be more than one thing going on. The geometry looks off, from the way the whole picture is twisted, which isn't normal. That could be a matter of the BIP pot settings on the game board, or it could be that the actual yoke needs straightening. However if no amount of adjustment to the BIP pots straightens it, that usually means a bad DAC. (And it's probably the one that hasn't been socketed and replaced yet.)

However it also looks like the spot killer is starting to kick in during the scrolling TEMPEST logo. Maybe in theory if there's a bad DAC, it could be cutting out enough during that sequence to be tripping the spot killer, but that's just a guess. Or there could be a separate issue with the spot killer. But I'd start with the game board adjustments first.
No, I don't have video of it on scope. I know the geometry is off (and the vectors are rather jittery) but we were planning on fixing the collapse issue before twiddling with the geometry pots. I suppose it's worth a try, but you still haven't told me what else could be wrong with the boardset that doesn't involve an NIB product with hundreds of satisfied customers having an assembly error nobody's ever reported before.
That damage is fixable. Clean the area up, replace the traces with wire.
Will that reinforce the PCB, which after being on fire has the structural integrity of a graham cracker, and was frankly marginal when it was new? Will you somehow secure the rest of the PCB (including the through holes) so the differential thermal expansion of your conformal coating (or nail polish, or peanut brittle) doesn't help pop further solder joints adjacent to it, causing other components to catch on fire? And most importantly, will you do this for free and send me your known-good XY board so I'm not throwing good money after bad?

I know there's a lot of pride in fixing vintage parts in the hobby, especially when you can talk other people into paying you big bucks to do it, but there is such a thing as the time value of money and being reasonably confident that you'll never have to touch a notoriously unreliable assembly again.
 
Last edited:
No, I don't have video of it on scope. I know the geometry is off (and the vectors are rather jittery) but we were planning on fixing the collapse issue before twiddling with the geometry pots. I suppose it's worth a try, but you still haven't told me what else could be wrong with the boardset that doesn't involve an NIB product with hundreds of satisfied customers having an assembly error nobody's ever reported before.

I gave you a game plan.

- Try to fix the geometry with the pots. If it can't be fixed, that points to there being some other issue in the video section, and video section issues could be triggering the spot killer. So we want to start there.

- If you can't adjust the geometry straight, that suggests the DAC that hasn't been replaced is bad. This is actually a common issue on these, see another example here. So replace that next, then see where the geometry and collapse issues are at.

- If you do replace the DAC and still get the collapse, the next step would be to inspect the deflection board, specifically the component values in the spot killer circuit. Assembly issues are always possible, especially if they did not use a Tempest boardset to test the deflection board, and used some other game to test that would not trigger this issue, because this issue is especially exacerbated by Tempest. But that's where I'd look next if this was my game.


You could also have some other issue with the game board, especially if a lot of work has been done to it, as shotgunning parts on these can result in unknowingly causing new issues. So to fix it right, you have to troubleshoot with a scope and find the problems, not guess and replace things hoping you eventually guess right.

Another option would be to send me the game board, ALAN-1 deflection board, and frame transistor block, and I can test them all in my own cab, and repair whatever is wrong. That would probably be the cheapest option, as it's a lot easier to troubleshoot these boards in person.

If you don't want to do that, post a video of the scope to give us more data, which is the next best thing to having the boards in front of us.


Will that reinforce the PCB, which after being on fire has the structural integrity of a graham cracker, and was frankly marginal when it was new? Will you somehow secure the rest of the PCB (including the through holes) so the differential thermal expansion of your conformal coating (or nail polish, or peanut brittle) doesn't help pop further solder joints adjacent to it, causing other components to catch on fire? And most importantly, will you do this for free and send me your known-good XY board so I'm not throwing good money after bad?

I know there's a lot of pride in fixing vintage parts in the hobby, especially when you can talk other people into paying you big bucks to do it, but there is such a thing as the time value of money and being reasonably confident that you'll never have to touch a notoriously unreliable assembly again.


Send that board in and I will restore it for what I normally charge for a non-burned board. Yes, everything you mentioned can be fixed. I've done it, see pics above. You could sell the ALAN-1 parts and recoup most of the cost.

As it is, you've already thrown a bunch of money at this, hoping to 'never need to touch it again'. And yet here you are. Vectors are finicky, and are not the kinds of things you can just 'set and forget', regardless of what ALAN-1 tells you. If you're going to operate one on location, you have to understand them.

I've refurbished more 6100 boardsets than anyone over the last 10 years here, and have well over 150 original sets running in other people's monitors. I warranty every one of them for a year, and provide support.

There's no such thing as throwing good money after bad, when you send stuff to repair people who know what they're doing, and can stand behind it. Is ALAN-1 offering you any help or support in this case?
 
for those in the future who might need a new wg6100 hv pcb

unless i missed it i don't think you said you used an alan-1 transistor block. i am not a fan of this transistor block due to the fact that you can simply upgrade the transistors directly on your monitor frame for less money and get the same results.

the alan-1 deflection pcb does not have any voltage regulation and because of this they recommend you pair their deflection board with their transistor block due to the added stress that would occur with the original transistors.

adding a Alan-1 deflection board adds approximately 10% higher voltage if used with the original transistors by bumping the normally 28v+- to 33v+-

as for the alan-1 deflection boards i have found them to sometimes have a ring condition and input clamp issues i think you might be describing one of these with your collapse statement.
input clamp issues will look like a vertical collapse and in severe conditions will show no picture at all. the ring condition looks like jagged lines that are crunched up in the center of the screen.
both of these issues happen randomly and are both triggered by tempest, omega race and others. my point is your deflection board may look great on star wars but have issues with tempest. this is my experience with this board i dont think anyone in the world has tested an alan-1 defection board in as many possible situations as i have though lol
 
for those in the future who might need a new wg6100 hv pcb

unless i missed it i don't think you said you used an alan-1 transistor block. i am not a fan of this transistor block due to the fact that you can simply upgrade the transistors directly on your monitor frame for less money and get the same results.

the alan-1 deflection pcb does not have any voltage regulation and because of this they recommend you pair their deflection board with their transistor block due to the added stress that would occur with the original transistors.

adding a Alan-1 deflection board adds approximately 10% higher voltage if used with the original transistors by bumping the normally 28v+- to 33v+-

as for the alan-1 deflection boards i have found them to sometimes have a ring condition and input clamp issues i think you might be describing one of these with your collapse statement.
input clamp issues will look like a vertical collapse and in severe conditions will show no picture at all. the ring condition looks like jagged lines that are crunched up in the center of the screen.
both of these issues happen randomly and are both triggered by tempest, omega race and others. my point is your deflection board may look great on star wars but have issues with tempest. this is my experience with this board i dont think anyone in the world has tested an alan-1 defection board in as many possible situations as i have though lol
I am using the TUK along with the deflection board from Alan-1. Again, it was a matter of expedience: we already had the parts in stock for another vector game that had been put on hold and while we could have "done it right", that would have involved waiting another week or two for shipping from Digikey and APAR. What kind of "upgraded" transistors can you get that fit on the monitor frame? I wouldn't think you could still get modern monitor transistors in the same bottlecap package.

After reading about the input clamp issue we tried turning down the size pots a hair, and that "fixed" it. The spotkiller LED on the deflection board still flickers on during the title, but the weird "sucking" kind of collapse we were seeing is gone. Still more vector jitter than I would like and I'm now hearing "thinking" noises over the speaker, but it's at least playable now. Might take up andrewb's offer on repairing the deflection board, depends on the cost. After all it has a big smoking crater in the middle that took out multiple connectors.
 
I am using the TUK along with the deflection board from Alan-1. Again, it was a matter of expedience: we already had the parts in stock for another vector game that had been put on hold and while we could have "done it right", that would have involved waiting another week or two for shipping from Digikey and APAR. What kind of "upgraded" transistors can you get that fit on the monitor frame? I wouldn't think you could still get modern monitor transistors in the same bottlecap package.

After reading about the input clamp issue we tried turning down the size pots a hair, and that "fixed" it. The spotkiller LED on the deflection board still flickers on during the title, but the weird "sucking" kind of collapse we were seeing is gone. Still more vector jitter than I would like and I'm now hearing "thinking" noises over the speaker, but it's at least playable now. Might take up andrewb's offer on repairing the deflection board, depends on the cost. After all it has a big smoking crater in the middle that took out multiple connectors.
ah speaking of this someone just posted the same Alan-1 deflection issue on face book. are you the same guy?
big holes can indeed be fixed although most people do a terrible job at it. a pcb with a repaired burn spot in most cases is just cosmetic and will work just fine. some have different standards some are fine with sloppy pcb patch jobs. some search for a clean original and others want a perfect new pcb. whatever camp you choose is your decision.
yes the mj21195-mj21196 or the mj21194-mj21193 and others can be used in place of the original transistors.
as for the spot killer alan-1 copied the original spot killer circuit but did not compensate for the LEDs difference in current compared to the original. to "fix" this you can replace r808 and r809 (which should be 1/2w instead of the 1/4w ) with 470ohm.
 
Back
Top Bottom