Sounds like the problem needs some real debugging. Maybe some other outputs also need to be in open-collector mode instead of push-pull output?
Does anyone know if original Creative chips work with problematic CT1336A cards? And happen to have access to oscilloscopes and logic analyzers?
I only have non-A CT1336 card with DSP 2.01 (IIRC), so I'd need a CT1336A card to see the differences. I don't have extra CMS chips or PAL/GAL chips (even though I have one old EEPROM programmer that could work).
In addition I have CMS chips soldered tight on SB 1.5, so I know how they work.
It seems some kind of test suite for SB/OPL2/CMS chips is needed to test their functionality and to assist in debugging with scopes and analyzers.
Any input how you think this kind of test suite should be made?
I can use C/Pascal/assembly to write this but I want to use freely available tools so everyone can modify/compile it. I will appreciate input what tools to use. Turbo C from Borland Museum might work?
At least the test suite should detect SB card presence (at 220h or perhaps at some other address as well?), DSP version, OPL2 chip precence at both 2x0h and 388h addresses.
Then it could output some OPL2 tones and CMS tones.
Note that without CMS chips SB 1.5 should hang when trying to access CMS chips. Same thing could happen on SB2 cards when PAL is inserted but CMS chips not.