SOLVED: Asteroids vector generator issues

KPW66

Well-known member

Donor 2018, 2024
Joined
Jan 2, 2018
Messages
1,212
Reaction score
720
Location
Foster City, California
Working on an Asteroids PCB (-04 version) on my bench.

Got two vector rams replaced and now the game boots and plays fine but I have graphic issues, see video.

I think the issue is on the "X" side?? and I have two new DACs and two new 4016B's on order. I have switched the existing chips with each other but the graphic doesn't change (to the "Y" side) so those may not be the issue.

I did replace all four Op Amps and no change.

My 36vAC is supplied by an Atari brick and tests fine on the board but my rendered 22vDC is low, around 12v. That is obviously causing some voltage issues but not sure if a VR is bad or a downstream component is pulling it down. And not sure if the voltage issues is causing bad graphics, or the bad graphics (component) is causing bad voltages.

The +5v (7805) and -15 (7915) outputs both test fine, but +15 (7815) is running more around 10.7 (due to low 22v rail). In suit, the -8.2v is fine, but +8.2v is low (~6.7v). The +12v (7812) is also at 10.7 (due to low 22v rail).

So, I first focused on the VR's.

Pulling 7815 output leg out of circuit does not resolve the issue.

Pulling 7815 input leg out of circuit does not resolve the issue.

But then, pulling 7812 output leg out of circuit (with 7815 input and output leg still pulled out) and I get 22vDC on that input rail.

So, I figure it was the 7812 or downstream component on that circuit and I put 7815 input leg back in circuit (with 7812 output leg still out) and 22v rail drops to 18v. Then, I put the 7815 output leg back in circuit too (with 7812 output leg still pulled) and 22v rail goes back down to 7v. FML

So, it's clearly not the 7812… or just the 7812, and doesn't appear to be the 7815, or only the 7815. So, I changed out the 7812 but no change. I don't have a new 7815.

Then I wondered if somehow the current on the 36vAC was too low and with everything pulling on it that was causing the 22v to drop, and so I swapped in another Atari brick and same issue. But I will admit my jumper leads are not a large gauge.

So, when parts arrive I'll swap in the new 7815, both new 4016B's, and both new DAC's, but if that doesn't improve it what should I check next?

Hope that all made sense.

 

Attachments

  • IMG_4616.jpeg
    IMG_4616.jpeg
    373.9 KB · Views: 13
  • IMG_4622.jpeg
    IMG_4622.jpeg
    387 KB · Views: 17
  • IMG_4608.jpeg
    IMG_4608.jpeg
    367.1 KB · Views: 19
  • IMG_4618.jpeg
    IMG_4618.jpeg
    595.6 KB · Views: 18
Last edited:
whoa the man must be sleeping. 2 big checks on his bozo drive him crazy list and he hasn't responded. Calling @andrewb. that is not a dac or probably not an op amp problem. you are haveing a problem either at the counters or at the logic chips between the counters and the dac which is not getting proper data to the dac. I would probe around that area looking for a stuck pin. as for your voltage problem I would investigate the diodes a little closer pop them out of circuit maybe and check them.

his list is replaceing parts shot gun style
 
I'd be looking at your TL1081/1082s
 
You mentioned that you replaced all four opamps. Where did you source the replacements? A while back I got some but they must have been remarks of a different opamp, etc since they made the output worse and got some weird artifacts. Once I got some real ones the problem went away.

But from the images it looks like you're board has more issues than that.
 
You mentioned that you replaced all four opamps. Where did you source the replacements? A while back I got some but they must have been remarks of a different opamp, etc since they made the output worse and got some weird artifacts. Once I got some real ones the problem went away.

But from the images it looks like you're board has more issues than that.
Digikey, Mouser.
 
You mentioned that you replaced all four opamps. Where did you source the replacements? A while back I got some but they must have been remarks of a different opamp, etc since they made the output worse and got some weird artifacts. Once I got some real ones the problem went away.

But from the images it looks like you're board has more issues than that.
I got the 081's from Jameco, and the 082's I've had for a long time so don't recall the source. Considering nothing changed when I swapped them I don't think that's it.

At this point I have socketed the chips and replaced the 4 op amps, but can swap the old chips back in so not really a shotgun situation.

Not too worried about replacing the 7812 as it's a quality part from Peter.

Andy actually told me to check and swap out the 7815 in a text message but I didn't have one in stock.

I can start check counters and logic chips
 
check the diodes and see if they work out of circuit. they might be only half wave rectifing the ac
 
Last edited:
Sorry to say this but the fake opamp chips that I ran into the issue with came from Jameco. They gave me a refund on them. It was really obvious since instead of perfectly straight lines it had an odd curve at the end. For the most part I have gotten all good parts from them but I wouldn't buy any of the TL opamps from them. Only Mouser or Digikey for those parts or hunt down old stock parts.
 
Sorry to say this but the fake opamp chips that I ran into the issue with came from Jameco…
Now you tell me!! lol

I can swap back in the originals but the issue has not bettered, or worsened with the chip replacement, so I don't think that's the issue.

Jameco is literally 3 miles from my house so an easy source for me, and to date I've never had any issues with their parts.
 
I didn't want to knock them since they have been overall a good supplier and that was the only part I had an issue with. Other than a few flakey remarked 6802 chips. The issue was the TL082 opamps. I had them in a socket and all of the ones I got from them had an odd artifact that disappeared when I put the original ones back in or ones from another vendor. They probably have it cleared up by now but at least a few questionable parts found their way into the supply chain.
 
Removed the 7812 and 7815 from the equation altogether and fed the board solid voltages on those rails and same issue persists so I believe I can rule out a power issue causing the problem. So, we're back to probing around but so far nothing is jumping out, no stuck pins thus far
 

Attachments

  • IMG_4666.jpeg
    IMG_4666.jpeg
    721.7 KB · Views: 8
  • IMG_4667.jpeg
    IMG_4667.jpeg
    395.2 KB · Views: 8
Actually, was still having power issues so I removed all four VRs and sent good power to each respective circuit and it's playing pretty well. As I've already swapped the caps and diodes for those VRs I'm going to install four new VRs and see if I can get stable board-made power into the vector circuits. That'll be next week some time.
 

Attachments

  • IMG_4669.jpeg
    IMG_4669.jpeg
    807.5 KB · Views: 10
Ok, everything is working perfectly with the bench power so I'm confident the new VRs will solve the problem.

When the board ran for a bit it all of a sudden glitched and it looked like the image was kinda doubled… so I probed around the X and Y position counter circuit and found pin 7 on the 157 at D/E10 was stuck low. Swapped it out and back to all good.
 

Attachments

  • IMG_4693.jpeg
    IMG_4693.jpeg
    179.6 KB · Views: 10
Update: all new VRs are installed and back to a voltage issue, but have determined it's on the +12v line as when I pull the 7812 output leg out of circuit I get a strong vector video image on the scope and good voltages in the vector circuit.

I also have a good +12v on the lifted 7812 output leg so will continue to probe around trying to figure out which audio component is driving this big draw. Annoying
 
Ok, figured the last issue (vector voltages) out…

The short answer is Operator Error (meaning I screwed up).

When I pulled the 36vAC from the Atari brick (pins 6&7) I didn't pull a GND as I wasn't using an AR board, and was assuming for AC I didn't need it. But I failed to notice that the center tap for that 36v winding goes to the GND on the brick.

When I swapped in a known-working Asteroids board I was still getting the same bad vector voltages so I figured I needed to run a GND from the brick to the PCB. I ran a
GND from pin 4 (or 5) from the brick and voilà! All good, perfect voltages and vector.

Spent some decent time on this mistake, but guess that's the cost for an education. Thanks to those that helped along the way.
 
Back
Top Bottom