PCB Repair: Typhoon

Problem

Incorrect PCM music and sound playback.

Diagnosis

Typhoon is the European version of A-Jax. It has a different level order and thus  manages to be even more frustrating than A-Jax. At least the music is great (when it works!).

Typhoon_PCB.jpg

PCM sounds are produced by two Konami 007232 ASICs, which produce two channels of sound each. H19 is used for sound effects and has 4Mb of ROM, while E19 is used for music (in conjunction with the YM-2151) and has 2Mb of ROM.

The sounds produced by E19 were barely audible and the sounds produced by H19 were sometimes incorrectly selected.

Probing around with the oscilloscope, I found that the NE output of E19 was stuck low:

E19.png

The NE signal is used to time multiplex the sound ROM bank signals for both channels of the two 007232’s:

NE.png

I had a non-working Bottom of the Ninth PCB that I’d already harvested customs from, so I desoldered one of the 007232’s from that. With this replacement 007232 installed, sound had improved but there were still issues. Sounds from E19 sounded truncated (the intro drum roll in this video sounds particularly odd):

For amusement, I captured some of the digital output of E19 with the logic analyzer:

Probing with the oscilloscope revealed that the signal on address line SA11 was rather weak. Damn! Had I just replaced one bad 007232 with another? I took another 007232 from the donor PCB and tested it. The result was the same, suggesting that it was actually the mask ROM that was bad and was attenuating the output signal of the 007232:

ajax.png
The low level is hovering around the indeterminate TTL input region and the high level is a bit low.

I’m not aware of a direct equivalent for the 2Mx16 mask ROM used, so I used a 4Mx16 27C400 EPROM with duplicated halves. With this replacement, everything now sounded correct.

Fix

Replace 007232 at E19.

Replace 2Mx16 mask ROM at F22.

One thought on “PCB Repair: Typhoon”

Leave a Reply

%d bloggers like this: