Atari ST OPL3 cart updates after feedback

In the meantime I’ve got some feedback from BadWolf and turned out that there was issue with software reset tied to A4 address line. I’ve completely forgot that same rules of OPL3 lines should be applied for reset signal (/ROM4 state wasn’t checked, which indicates that we are accessing cartridge space). Also BadWolf suggested replacement of two ICs (AND and NOT gates) with one NAND IC. There was no issue with former approach, but both ICs weren’t used in full (had unused inputs and outputs), with NAND approach we have only one IC and use all of it’s inputs and have the same results, which is nice.

Secondly design used polarised capacitors in some places, so I’ve had to make place for THT electrolitic ones on pcb, I dislike SMT ones for some reason.

I’ve also ordered all the parts for cartridge (enough for max. 5 cartridges, most of them arrived). Still waiting for quartz generators. As I couldn’t get standard DIP-4 ones and SMT versions were available in larger quantity, I’ve had to change PCB to accomodate SG-615P (by Seiko, officially out of production). It looks like getting 14,318MHz for 5V is getting complicated these days. Most of generators are either for 3,3V or frequency wasn’t suitable for OPL3 purposes (14,31MHz).

Anyway, here is pcb new look (rev.9) after feedback and refinements:

Hope to send gerbers to manufacturer soon, but still waiting for some more feedback and reviews.

[update 11.04.2024] I’ve received some more, great feedback from Willie (mouSTer / CT60e). Most of things were fixed besides one crucial and serious one. Turned out that I cannot use op-amps in current configuration. Thing is that current design was directly taken from existing ISA card designs. Main issue is that those op-amps originally worked with +12V/-12V and we only have +5V and ground at our disposal. So, I have to develop better DAC to analog output circuit from scratch using other parts.

The plan is to make prototype anyway, but without analogue part and expose crucial YAC-512-M pins, so I could test OPL3->ST communication and develop missing circuit on breadboard or protoboard. So, after that I will be able nail down the final design with proper digital / analog domain separation on pcb.