Joust, Missing Colors [SOLVED]

you have to manually do Restore Factory Settings, Clear Bookkeeping Totals, and High Score Table Reset and see if the blank 4000 point scores go away. if they keep coming back then the switcher is probably messing it up.

what was your procedure for replacing the shift register? how did you decipher that particular one was bad?
Used a logic probe, found 3I had a pin that was stuck low, so that's the one I swapped out. The blank 4k high scores were at least staying... Before they were a mess and changed with every boot. I've reset everything, and it looks normal now. We'll see if they revert to chaos before Brad's adapter shows up tomorrow.

1718947883079.png

Another thing, however... once again, if I leave the back off the glitchy graphics aren't happening. Anyway, progress was made today. Thanks again for the help!
 
Used a logic probe, found 3I had a pin that was stuck low, so that's the one I swapped out. The blank 4k high scores were at least staying... Before they were a mess and changed with every boot. I've reset everything, and it looks normal now. We'll see if they revert to chaos before Brad's adapter shows up tomorrow.

View attachment 750509

Another thing, however... once again, if I leave the back off the glitchy graphics aren't happening. Anyway, progress was made today. Thanks again for the help!
which pin was stuck low? the data outs from the rams are what tunnel into the shift registers.

captain obvious moment here, but I always wondered how 3 banks of ram translate into 4 shift registers. they only use 6 bits, CDEFGH, AB are grounded and unused (4x6=24) LOL consulting datasheet though it's a little more complicated than that. shift register failures don't usually result in what you're experiencing. https://microcontrollerslab.com/74l...out-datasheet-examples-applications-features/

hopefully you notated somewhere which pin was stuck low, if I know what function it is I'll be able to deduce more
 
which pin was stuck low? the data outs from the rams are what tunnel into the shift registers.

captain obvious moment here, but I always wondered how 3 banks of ram translate into 4 shift registers. they only use 6 bits, CDEFGH, AB are grounded and unused (4x6=24) LOL consulting datasheet though it's a little more complicated than that. shift register failures don't usually result in what you're experiencing. https://microcontrollerslab.com/74l...out-datasheet-examples-applications-features/

hopefully you notated somewhere which pin was stuck low, if I know what function it is I'll be able to deduce more
It was pin 13, see post #6. When I replaced 3I, it did fix the initial problem that started all of this - red was missing. Which is to say, it doesn't feel like a shift register problem anymore maybe. Perhaps that one failing was a symptom of something else though.
 
it doesn't happen often, but the 2H 74257 could be on the fritz.* that takes the 8 bits from the 8 shift registers (this is where the Screen Control cocktail flip is made possible) and turns that into 4 serial bits that pass through the next 2C 74257 before getting output to the 7489 color rams, and then one last 74374 buffer before going to the RGB output transistors.

chances are the normal display half of 2H is what's gonked if you're able to force cocktail flip and it exhibits no problems with the graphics scrambling. and would explain why it doesn't appear to be a traditional shift register failure.

74257.jpg
 
it doesn't happen often, but the 2H 74257 could be on the fritz.* that takes the 8 bits from the 8 shift registers (this is where the Screen Control cocktail flip is made possible) and turns that into 4 serial bits that pass through the next 2C 74257 before getting output to the 7489 color rams, and then one last 74374 buffer before going to the RGB output transistors.

chances are the normal display half of 2H is what's gonked if you're able to force cocktail flip and it exhibits no problems with the graphics scrambling. and would explain why it doesn't appear to be a traditional shift register failure.

View attachment 750611
So I was on the right track then. I think what I'll do is force the flip and let it run closed up to see if it has the same issues.

I do have some 74257s - maybe I'll piggyback that too and see if that makes any difference.
 
The 166's and 257's are all after the RAM/Compute and only used for rendering the image (you could remove and game would still boot with 'Zero' on the ROM board). They have no DIRECT effect on the CMOS/RAM errors etc (It is possible due to short to interact with RAM on the 166's). I think the original problem with the color(s) missing was solved. Your CMOS getting wiped is something else. Is the watchdog pad cut? Usually corruption occurs if the shutdown happens when the +5v gets too low before the CPU is put into Reset. I sent a 4.7v Zener (ZR2) with the PS adapter.

~Brad
 
The 166's and 257's are all after the RAM/Compute and only used for rendering the image (you could remove and game would still boot with 'Zero' on the ROM board). They have no DIRECT effect on the CMOS/RAM errors etc (It is possible due to short to interact with RAM on the 166's). I think the original problem with the color(s) missing was solved. Your CMOS getting wiped is something else. Is the watchdog pad cut? Usually corruption occurs if the shutdown happens when the +5v gets too low before the CPU is put into Reset. I sent a 4.7v Zener (ZR2) with the PS adapter.

~Brad

Agreed, the topic of the thread as been solved. I put the diode in last night with the adapter. Funny enough, zero graphical errors since I installed your adapter. In this moment, it's working really well. I'll start a new thread if something else comes up.

Recap, I was missing red, monitor was fine. Turned out to be the shift register at 3I.
Bonus fixes: Learned that the CMOS was actively buying the farm, replaced that and installed a lithium coin battery to solve the high score saves.
 
Back
Top Bottom