Bubble Bobble PCB not booting

delroy666

Well-known member

Donor 2011
Joined
Jul 30, 2009
Messages
1,189
Reaction score
26
Location
Ottawa, Ontario, Canada
I've been trying to get this Bubble Bobble PCB working but I'm not having much luck. It's currently 'dead' - no sound, no video. The clock inputs at all 3 Z80 CPUs are good and the /RESET line is high. I have a sync output from the video board, but no RGB outputs and there is no activity on the address inputs of the graphics ROMs. Test mode does nothing different. Here's what I've done so far:

- the PROM socket at IC41 on video board was broken in half, replaced it
- pulled all ROMs and PROMs, dumped them, all verified in ROMIdent
- logic probed both boards, found a floating output at pin 19 of the "PS4" MCU. Swapped in another "PS4" chip, no change. Traced this output to a 74LS74 at IC30, replaced it, now that signal is high. All other ICs look okay (no floating signals anyway).
- hooked up the 9010a - all bus, ROM and RAM tests passed on all 3 CPUs, except for ROM 52 from the main CPU - looks like this is banked, with the PAL at IC43 decoding the ROM's /CE, A14 and A15 signals. Is there any way I can read this ROM with the Fluke?
- used the logic comparator on almost all the ICs, found nothing bad
- replaced all 3 Z80 CPU's
- continuity tested all signals in the ribbon cable, all good

I'm at a loss for what to try next. Anyone have any suggestions? I'd highly appreciate any help I could get on this.
 
Use the "Run UUT" option on the fluke, if the Fluke Z80 pod can run the board but the original CPU cant then you have a bad Z80, rare but not impossible.
 
Thanks for the reply. I did try 'Run UUT' but it does the same thing as having the CPU in there. I replaced all the CPUs with known-good Z80s too.
 
Thanks for the reply. I did try 'Run UUT' but it does the same thing as having the CPU in there. I replaced all the CPUs with known-good Z80s too.

Just curious what are the numbers on the z80's and if it is anything like the galaga board sets, the first z80 needs to be a z80a. Just sayin.
 
I've got Z80Bs (Zilog Z8400BPS) for CPUs 1 and 2, and a Z80A (Zilog Z8400APS) for the sound CPU.

check for signal on pin 6 of all z80's and probe pins 21 22 of the first z80. Should give you a clue as to what is/not happening.

Verify dc voltage to chips.
 
One HUGE thing on the Z80... Check the /RESET pin.

If it's not starting out low momentarily then going and staying high then the CPU is not being reset. Z80 CPUs will act anywhere from totally dead to having random crap strobing on the data and address lines if the reset signal is bad.
 
check for signal on pin 6 of all z80's and probe pins 21 22 of the first z80. Should give you a clue as to what is/not happening.

Verify dc voltage to chips.

CLK, /RD and /WR are toggling on all CPUs. Clock signal looks good on the scope. I have 5.04V at the ICs on the CPU board and 4.94V at the ICs on the video board.

Channelmanic - on power up, /RESET is low for about 1 second, then goes high and stays high. The watchdog circuit is getting updated by the CPU - the /TRES signal is toggling to clear the counter. If I pull the main CPU and power up, the watchdog does start toggling /RESET.
 
Last edited:
CLK, /RD and /WR are toggling on all CPUs. Clock signal looks good on the scope. I have 5.04V at the ICs on the CPU board and 4.94V at the ICs on the video board.

Channelmanic - on power up, /RESET is low for about 1 second, then goes high and stays high. The watchdog circuit is getting updated by the CPU - the /TRES signal is toggling to clear the counter. If I pull the main CPU and power up, the watchdog does start toggling /RESET.

Is this happening on all z80's or just the first one?
 
Is this happening on all z80's or just the first one?

It does this on all 3 CPUs and also on the MCU. That is, the signals going to the Reset pin of each of these devices (/RESET, /SBRES, /SRESET and /SEQRES) start off low, then after a second they go high and stay high. The sound CPU seems to get a short reset pulse immediately after it goes high, (and I hear a single 'tick' sound out of the speaker at that moment) but then it stays high after that. There is activity on the data and address pins of all CPUs and the MCU.
 
Last edited:
Can you check the RAM and ROMs on the sound section with your Fluke to make sure that subsystem is operating properly?
 
Can you check the RAM and ROMs on the sound section with your Fluke to make sure that subsystem is operating properly?

All the Fluke bus, RAM and ROM tests passed on all 3 CPUs. (except, as I mentioned, that banked ROM 52 on the first CPU) The sound CPU data, address, read, write pins are all toggling, as are the inputs of the FM synth chips. There appears to be serial data coming out of them, but it could be just garbage since there is no waveform at the analog output of either DAC. Is there any way to test those FM synth chips directly using the Fluke? e.g. write to its registers to output a sine wave or something? I noticed in the mame source there is an address to send a 'command' to the sound CPU from the main CPU. Any idea what data to send to trigger a sound effect? It's possible the game is booting correctly, but the sound and video are both dead, so it would be nice to rule that out.
 
I've got Z80Bs (Zilog Z8400BPS) for CPUs 1 and 2, and a Z80A (Zilog Z8400APS) for the sound CPU.

I really hate the Zilog chips and replace them first thing! Sorry Womble, I know that is daf. I had a problem with a board and took me several hours only to find a Zilog z80 was marginal at best. Would start to work then crap out and now I rip em out like a bad tooth, each and every time. No board I send out will ever have a Zilog cpu chip in them, ever!

This is just me, but I really hate marginal or intermittent problems and they are the hardest for me to track down. Game works then doesn't. It is much easier for me to fix a non working board than one that works some then quits.
 
Just to follow up... After getting a working boardset to compare to, I was able to get this Bubble Bobble working. Swapping in a known-good video board, I found that the /V-RAM and /COLOR signals coming from the CPU board were stuck. These are generated by the PAL at IC49. All the inputs to this PAL were toggling except the one coming from /IORQ of the main CPU. The CPU's /INT signal was never active, so I traced that through an LS74 to one of the outputs of the custom "P4" chip. (a known-good chip) All the other signals were toggling nicely, as were the inputs and outputs of the LS245's and LS157's that connect to it, so I just started shotgunning these ICs that interface with the custom. As soon as I replaced the 74HC245 at IC59, the board came to life. :biggthumpup:
 
Just to follow up... After getting a working boardset to compare to, I was able to get this Bubble Bobble working. Swapping in a known-good video board, I found that the /V-RAM and /COLOR signals coming from the CPU board were stuck. These are generated by the PAL at IC49. All the inputs to this PAL were toggling except the one coming from /IORQ of the main CPU. The CPU's /INT signal was never active, so I traced that through an LS74 to one of the outputs of the custom "P4" chip. (a known-good chip) All the other signals were toggling nicely, as were the inputs and outputs of the LS245's and LS157's that connect to it, so I just started shotgunning these ICs that interface with the custom. As soon as I replaced the 74HC245 at IC59, the board came to life. :biggthumpup:

Sorry for bumping this interesting thread after so long, but you previously mentioned that you had gone over all the ICs on the PCB with a logic comparitor. Did you previously test the 74HC245 at IC59?

Its a shame the comparitor cannot test these buffers.
 
Just to follow up... After getting a working boardset to compare to, I was able to get this Bubble Bobble working. Swapping in a known-good video board, I found that the /V-RAM and /COLOR signals coming from the CPU board were stuck. These are generated by the PAL at IC49. All the inputs to this PAL were toggling except the one coming from /IORQ of the main CPU. The CPU's /INT signal was never active, so I traced that through an LS74 to one of the outputs of the custom "P4" chip. (a known-good chip) All the other signals were toggling nicely, as were the inputs and outputs of the LS245's and LS157's that connect to it, so I just started shotgunning these ICs that interface with the custom. As soon as I replaced the 74HC245 at IC59, the board came to life. :biggthumpup:

great job and thanks for sharing.
 
Sorry for bumping this interesting thread after so long, but you previously mentioned that you had gone over all the ICs on the PCB with a logic comparitor. Did you previously test the 74HC245 at IC59?

Its a shame the comparitor cannot test these buffers.

For any chip that has too many pins for the comparator to fit, I piggybacked a known-good chip on top. Obviously that doesn't always work though. I agree, I wish the comparator could be used on these 20-pin ICs.
 
For any chip that has too many pins for the comparator to fit, I piggybacked a known-good chip on top. Obviously that doesn't always work though. I agree, I wish the comparator could be used on these 20-pin ICs.

Thanks for that, I'm in a similar situation with a konami frogger at the moment, watchdog keeps barking and screen fills with zeros. I have gone through all the TTLs, ROMs and RAM but still cannot find the problem, I'm at a loss and no working PCB to compare against.
 
Back
Top Bottom