Joust crashes when too many enemies onscreen

Chupperson

Member

Donor 2016
Joined
Aug 25, 2016
Messages
204
Reaction score
10
Location
Tulsa, Oklahoma
This is an extension of my previous Joust thread, but I have a clearer mission now.

My Joust has recently become very consistent in crashing on wave 4 when there are plenty of enemies onscreen.
When you turn it on, it always gives a RAM 21 error. RAM 21 is verified good with a new socket. After the RAM test, the game boots as normal and the LED shows 0.
In wave 3, it will usually leave a copy of a buzzard (sans rider) on one side of the screen until something else moves into that space and paints over it.
When I get to wave 4, it usually keeps playing for a minute or two, but extra parts of enemies keep getting left on the screen. After that the screen flashes with several lines of garbage in the top third of the screen and the game's colors become inaccurate, and then the game resets back to the rug pattern.

So far, I have replaced the pin headers and connectors on the MPU and ROM boards along with all the sockets for the ROMs, decoder PROMs, and special chips, replaced the ROM ribbon cable, and have tried swapping out known good decoders, special chips, and PIAs. I have also tried my Robotron ROM board with my Joust MPU board and it boots up directly to 0. I am fairly confident there's a logic chip on the ROM board that is giving me problems but I don't know what to look for.
Either that, or maybe my power is bad, but my power supply board has been rebuilt and has good voltages.

The only other thing I can think of that I haven't done is validate the ROMs, but it doesn't seem like they would cause this problem.

If anyone has seen this happen before, please let me know what you think.
 
Last edited:
Reset is likely bad power.
>> voltage droop due to weak power supply or cabling
>> excessive AC ripple (load dependent)

Play and monitor the +5VDC on the CPU board (or ROM board).
(Might also be the 12V or -5V)
 
Based on your description it sounds like it's having issues writing to ram correctly. This will cause a reboot eventually when the stack or game state variables get corrupted. The fact that it happens when there's lots of enemies on the screen would make me think special chips. If you have already changed those then I would look at the logic/bus drivers around the special chips. You might need a scope for this as it's possibly a timing error or a device that's working but slow or not swinging the full 5V. Logic probe may not catch this. I could possibly help more if you could get some screen shots of the "left" over enemies when it's in this mode. You should also check power as VecrotColector suggested as well. It's possible one of the voltages is drooping when there's a lot of blitting happening (when lots of enemies are on screen)
 
Sorry! mad, typing skills.

LOL ... no dude, I like it ... many would agree I should change my KLOV login to that ... :)

And I agree ... might be a timing issue. Maybe OP has different speed RAM? Maybe other peripheral components too.

Classic case where another working boardset might help define which PCB is at issue.
 
Last edited:
I'm pretty sure it is the ROM board. I put the Joust ROM board in my Robotron and it crashed every time it tried to display the title screen.
I also put the Robotron ROM board in Joust and it booted up fine, but it also didn't display the fancy attract screen correctly. (See below.)

I forgot to mention, it has corrupted the SRAM a few times when restarting. That sounds like another indicator of bad power, but still doesn't explain why the ROM board didn't work in my fully working Robotron cab.

I'll get some video of the errors in a while and report back.
88fb07f695cf105d463f82fa05ee0612.jpg
 
Last edited:
I got a pretty good video of a typical play session. It lasted a little longer than usual this time and happened at a time when less enemies were onscreen.



Something else I forgot is when I put the Joust ROM board in Robotron, it booted up with an ALL SYSTEMS GO but still crashed at the title screen.
 
Last edited:
Sounds like multiple problems if your test is accurate. You put a known good Robotron ROM board in the Joust and had problems, and your Joust board is not working in your Robotron.

If those tests are reliable, you have multiple problems.

Check your 5V rail on the ROM board and MPU board. Anything below 4.6-4.7 and you could have problems. Ideally you want this at 5V.

I've seen these sets run down to 4.5V... but not ALL of them. That one kept me going for a while figuring the problem wasn't in my testing setup (as several other boards tested fine). I found the voltage at about 4.5. I fixed that, and then all the boards worked fine on the test station.

As these Williams sets are multiple boards, you need to try to get to a known state with each so your not trying to troubleshoot multiple problems at the same time.
 
if you switch rom boards and the MPU doesn't sense the change and force the factory settings restored, the crossing of games against the CMOS will make it crash. typically it knows and does this on its own, but sometimes doesn't. Joust boots direct to the high score screen, so it and Sinistar particularly blow up immediately.

you can either pull your batteries or hammer the Advance switch after the all systems go message to get into the settings mode and manually activate factory settings, clear bookkeeping, and clear high score table.

this is probably why your fancy attract mode didn't work in Robotron on Joust.

the Special Chip sockets could be bad, or you have bad Special Chips. my suggestion is to mark your Joust and Robotron ones and try swapping them, that way you'll determine if any of those are bad. or in other words, put your Joust/Joust and Robotron/Robotron (rom/MPU) back together and swap the Special Chips between rom boards to see if the problem carries.
 
My first thought was special chips as well. I swapped in the ones from Robotron and there was no change. I even put in new dual wipe sockets to be sure. I'm pretty sure I even tried the Joust special chips in Robotron to make sure, and there was no change there either.

I thought it could be a voltage issue which is why I just replaced all the pin headers. Voltage at ROM board power connector is a solid 5.01v. Should I check for AC ripple anywhere? The power supply has all new caps and bridge rectifiers (except maybe not the thin, rectangular one... really hard to remove that one).

It isn't the RAM either. I moved my 4164s from Robotron to the Joust MPU to check and it didn't change anything.

That's what's leading me to believe it's one of the logic chips on the ROM board.
To rule everything else out I guess I should swap ROM boards again and see if they boot after clearing the CMOS.
 
Last edited:
Help me fix my Joust ROM board

Been trying to keep fixing this off and on for over a year.
Things I have done so far:

Replaced ribbon cable with new one, no change
Swapped out known good special chips from Robotron, no change
Swapped out known good PIA from Robotron, no change*
Swapped out 8C, 9C, 8E (chips near special chips), no change
Swapped out 3E (7404) with 74LS04, no change
Swapped out ROM board from Robotron, running Joust ROMs - All systems go

*new intermittent error has started happening: sound line 6 missing from sound test; player sounds are wrong. I read this is caused by a bad PIA but it does it with either PIA and it actually fixed itself mid-game when I was testing just now. I'm sure it will happen again.

"Copied" graphics show up most in the middle of the screen, in three specific places: above the right side of the leftmost platform, above the right side of the center platform, and at the rightmost area of the screen. All of these are usually on the middle level.

Game most often crashes at the beginning of a stage when generating at least 4 or 5 enemies. The enemy spawn sound carries through the crash and subsequent rug pattern.

Once when I tested the game, it froze completely and stayed that way until I hit the advance button. Then it exited back out to the high score screen.

There is a worrisome gash in the board near the 74154. I replaced the 74154 because it looked like it had been hit with a screwdriver and all of the pins were slightly bent because the chip looks like it was moved. I don't know what this line is attached to but I wonder if it will work better if I jumper it.

I can try piggybacking some of the logic chips on the ROM board to see if they make a difference. I'll experiment with that tomorrow.

I am running out of ideas and I don't know enough about how the game works to know what to probe to check on the logic.
 

Attachments

  • 20191002_214157.jpg
    20191002_214157.jpg
    506 KB · Views: 19
  • 20191002_204446.jpg
    20191002_204446.jpg
    350.6 KB · Views: 15
Been trying to keep fixing this off and on for over a year.
Things I have done so far:

Replaced ribbon cable with new one, no change
Swapped out known good special chips from Robotron, no change
Swapped out known good PIA from Robotron, no change*
Swapped out 8C, 9C, 8E (chips near special chips), no change
Swapped out 3E (7404) with 74LS04, no change
Swapped out ROM board from Robotron, running Joust ROMs - All systems go

*new intermittent error has started happening: sound line 6 missing from sound test; player sounds are wrong. I read this is caused by a bad PIA but it does it with either PIA and it actually fixed itself mid-game when I was testing just now. I'm sure it will happen again.

"Copied" graphics show up most in the middle of the screen, in three specific places: above the right side of the leftmost platform, above the right side of the center platform, and at the rightmost area of the screen. All of these are usually on the middle level.

Game most often crashes at the beginning of a stage when generating at least 4 or 5 enemies. The enemy spawn sound carries through the crash and subsequent rug pattern.

Once when I tested the game, it froze completely and stayed that way until I hit the advance button. Then it exited back out to the high score screen.

There is a worrisome gash in the board near the 74154. I replaced the 74154 because it looked like it had been hit with a screwdriver and all of the pins were slightly bent because the chip looks like it was moved. I don't know what this line is attached to but I wonder if it will work better if I jumper it.

I can try piggybacking some of the logic chips on the ROM board to see if they make a difference. I'll experiment with that tomorrow.

I am running out of ideas and I don't know enough about how the game works to know what to probe to check on the logic.


Reflow the solder on all 3 of the connectors of the ROM board (not the ribbon cable). Just get that out of the way, then troubleshoot further.

.
 
I've heard/read that some CPU boards have weird issues with 4164 Ram. Not any specific rev, just random boards.

Did the issue appear after going from 4116 to the 4164?



Did you read the Joust Roms and Romident them to make sure the data is good?
 
I'm running 4116 RAM currently. I tried 4164 as well but they both behaved the same.
I don't think it's a ROM issue since the game loaded fine with the same ROMs on a different board.
 
There isn't a whole lot left on the ROM board, there might be a cracked trace that's just getting intermittent contact/ contact that changes with temp.

Your initial issue really did sound like RAM or Ram access issue.

But if you've narrowed it down to 100% an issue with the ROM board then it may be worth it swap out those few TTL chips. I'd put money on the 7474 (74LS74) being flaky.

But for sure check al the traces, and double check all the sockets. Especially if it looks like someone else worked on the board.
 
Replaced 7404 with 74LS04? You have to be careful about switching families on the Williams boards. Depending on fanout, an LS can only drive a few standard TTL inputs. The fact that the CPU board won't work properly in either game indicates that there is an issue with the board. I had a Bubbles that drove me crazy; it would run the RAM test for days without issue however once I put it into attract mode, it would lock up. I ended up replacing the data buffers for each bank. I checked the 7474s which control the CAS/RAS lines on the DRAMs and they compared fine. I replaced those as well. It was strange because none of the chips (couldn't test the data buffers because they are tri-state) tested bad yet once replaced the board worked. I had another Bubbles that acted the same way and it ended up being a bad Special Chip.

Especially when boards are mixed together, strange things happen. Some of the early ROM boards had a small aux board with the 74LS244 and pullup resistor, I've seen CPU boards with a small aux board with the 74LS367 and a pullup resistor.

Try swapping the ROM chips (after verifying they are 2732s or 2532s) to make sure that the game will run with the good ROM, CPU and Joust program. If the good Robotron ROM board doesn't work with the Joust CPU, concentrate on fixing the CPU board. Chasing multiple problems on multiple boards is difficult and frustrating. I ended up replacing every chip on the Bubble ROM board because I was sure it had to be there.
 
Taking Charles' advice, the next thing I swapped out was the two 7474s on the ROM board. The graphics copying problem is vastly improved. It only started leaving copies of enemies onscreen around wave 8 or 9 and it didn't crash, so I was able to actually finish a game.

I had to swap them out with 74LS74s because it's what I had available. Do you think it would be better to replace them with 7474s instead? I socketed them in case thos might happen again.

I still get RAM error 21 when I boot the game with this ROM board, so maybe there's another problem chip to replace?
 
Back
Top Bottom