Gunbird from HELL

[Crafty+Mech]

I picked up a cheap Gunbird pcb (Japan version) that turned out to have some sprite glitches.

The board was slightly warped, so I started pressing on the board in various places and the graphics would clear up. The source was hard to pin down though, and after reflowing a couple smaller SMD chips, I finally found the culprits near the large SMD chip with the Psikyo logo.

Somewhere in the mist of reflowing and flexing the pcb, the sound cut out. The big 2200uf cap turned out to be quite loose, so I re-soldered it and prepared to celebrate my victory. Denied! After reattaching the cap I'm getting different results every time I cycle the power. Sometimes digital only, or FM only, occasionally both, or just dead silence.

Self test on boot passes ram checks, etc.

I'm started to think a demon of infernal silence has possessed the board and is just screwing with me now. Anyone ran into an issue like this with a Psikyo board, or Gunbird in particular?

[system11]

Suggest checking everything you reflowed.. Also reminds me of sound CPU ram or reset cap faults.

[Crafty+Mech]

Thanks for the tips, I did some more work on the board this morning. First I reflowed the sound RAM/ROM chips that I could trace back to the Yamaha sound chip. No change, but it did give me an idea to look at the Yamaha reset line.

Sure enough if I momentarily pulled the reset line low, the chip reset and the sound came back. Some of the digital samples are mismatched (wrong sample for character), and the coin up sound is wrong, but at least now I have a way to consistently restore the FM audio & sampled sounds.

Next up I'll take a look at the two chips that are connected to the Yamaha reset line of the sound chip. One connection is the reset pin of a flip flop, and the other is a small Sharp SMD adjacent to the Yamaha. Hopefully the issue is not with the SMD chip, I can't find any datasheet or info beyond it being based on the Zilog Z80 series.

[Crafty+Mech]

Unfortunately it looks like the reset signal comes from the Sharp surface mount chip (also found on Strikers 1945). Google turned up no datasheet or pinout for that chip.

I think my best bet will be to find the reset line on the Sharp and tie that to the reset line of the Yamaha. At least then the board will boot with FM audio, and most of the digital samples. 90% audio is better than none for the time being.

[system11]

Is the sound CPU showing activity on the address lines before you reset it? If not, and you see brief activity before it stalls again when resetting, it's fairly likely the sound RAM is bad.

[Crafty+Mech]

Spent a little time with the board this weekend, and as suggested watched the address lines on the Yamaha sound chip before/after a manual reset. There is no activity on the lines before a reset, but the address lines spring to life once the reset occurs. I also found that the reset pin on the sound chip is pulled low briefly during the boot process (and the address lines subsequently become active) but the data lines don't show any activity until I manually reset the chip.

I also double checked if I missed any smd chips that were connected with the Yamaha in some way, and I found one. Reflowed that one, and now when I do a manual Yahama reset after boot I get 100% working sound. So the board is 100% if I provide the manual reset, but otherwise will boot to no sound at all.

So close...

[system11]

Trace back on the reset circuit, look for caps used to create the reset delay.

[Crafty+Mech]

FIXED !

The reset line was connected directly to the Sharp custom, with no caps in the area (outside of small filter caps for each IC connected through the inner ground/power planes). So stumped in that direction, I traced backwards from the Sharp custom until I found a watchdog chip @ U20. So on a hunch I tied RESET on U20 to RESET on the Yahama, and now the board boots with perfect sound every time!

I'd like to know what was actually happening during the boot process to cause the sound to drop, but with the hours I've already put in on this board, I'm good with just calling it fixed.

$40 for Gunbird seemed like a such good deal at the time

[matrigs]

FIXED !

The reset line was connected directly to the Sharp custom, with no caps in the area (outside of small filter caps for each IC connected through the inner ground/power planes). So stumped in that direction, I traced backwards from the Sharp custom until I found a watchdog chip @ U20. So on a hunch I tied RESET on U20 to RESET on the Yahama, and now the board boots with perfect sound every time!

I'd like to know what was actually happening during the boot process to cause the sound to drop, but with the hours I've already put in on this board, I'm good with just calling it fixed.

$40 for Gunbird seemed like a such good deal at the time

Impressive! Congratulations!

I would really want to be able to FIX something as well from time to time.