jaZz_KCS wrote:

Please put me on the list

Ok.

And about my question above?
LPT DAC (0x378/0x278), Covox Sound Master (0x2X2/0x2XF), Covox Voice Master (0x22F)... Anything else? No more DACs were known?
(no-DMA variants)

Rainbow Arts Soundman also listen to 0x300 maybe Voice Master too?

I look forward to completing the work on the FMonster, the SID chips purchased and tested according to the technique of the respected Fagear are already waiting for their finest hour:). Please also include me in the list.

matze79 wrote:

Rainbow Arts Soundman also listen to 0x300 maybe Voice Master too?

I've tried to decode and simulate address decoder of Rainbow Arts Soundman from photos.
That's what I've came up with.

This card puts OPL2 on base addresses: 0x380, 0x388, 0x38E.
And 8-bit Non-DMA DAC on base addresses: 0x220, 0x221, 0x228, 0x229, 0x22E, 0x22F, 0x300, 0x301, 0x308, 0x309, 0x30E, 0x30F
And all of those at the same time! 😲
That's very odd choise of ports. And weird schematic of decoder.
It uses two 74LS138 ICs, two elements from 74LS10 and two elements from 74LS04.

Crang wrote:

Please also include me in the list.

Ok.

I tested the soundman with Tiido a while back and the DAC only produced sound on ports 220 and 300 if that helps

Two more months passed... I've fixed some errors in some blocks and rearranged those on the board.
New order (left to right): SN PSG, 2xSAA (CMS), 2xOPL2 (2xAdLib, SB Pro), AY (CSM), 2xSID (2xSSI-2001, DuoSID), OPL3.

I've finally got to test the "3-voice PSG" board (SN76489)!
And it works just fine.
Of course there were some minor bugs as:
- I've forgot to connect power to one logic IC that provided clock for SN76489
- Some capacitors and resistors needed to be added to reduce noise level
- I've connected LED to the wrong rail and it didn't light up on access

Everything else is fine.
Address decoder works, data bus buffer works, "garbage-tone-blocker" logic works (SN76489 doesn't have a reset circuit and starts to produce random tones as soon as power is applied), power regulator does not overheat.
After noise-reducing tweaks I've managed to get less than -60 dB of noise on the output.
Games that should work with "Tandy 1000 sound" work fine. TSRs from TNDY and Lotech projects help a lot.
I've finished filming a video about debugging and testing of the board, but I have not put it together yet, so no audio samples yet.

What did you use to make the heat map? One of those cheap heat map guns on ebay?

Hi,
I don't understand why you use a Buffer and @decoding for each chip.
Only some buffers and a CPLD for @decoding cad do everything and it will save so many space.

He does not want to use any programmable logic or device at all, thats why.

i neither have logic analyzers or thermal camera at all, nice equipment 😀

If you don't have a thermal camera, what did you use to make the heat map?

i use finger sensor.

awgamer wrote on 2020-01-12, 10:10:

One of those cheap heat map guns on ebay?

Exact opposite. Flir E4. Bought it recently.

FreddyV wrote on 2020-01-12, 10:23:

I don't understand why you use a Buffer and @decoding for each chip.
Only some buffers and a CPLD for @decoding cad do everything and it will save so many space.

I do not want any programmable device on the FMonster, even ROM.
And data bus buffer is more or less required for each chip that outputs analog audio (that's every chip except OPL2 and OPL3).
CPLD or FPGA may be used in USB-variant to save space and cost. But ISA variant will use only hard logic ICs.

I've tried to playback VGMs from Sega Master System and that's what I've got: https://www.youtube.com/watch?v=5B3-YxNLJCE
Software used: SBVGM v1.28

Fagear wrote on 2020-01-12, 17:48:

I've tried to playback VGMs from Sega Master System and that's what I've got: https://www.youtube.com/watch?v=5B3-YxNLJCE
Software used: SBVGM v1.28

Wow 1k views on first day! 😳
Thanks for sharing, btw! Though I'm not familiar with all these tunes, I can say that Green Hill Zone sounded perfectly fine to me. Good work! 😎

Edit: Typos Fixed.

Fagear wrote on 2020-01-12, 14:04:

I do not want any programmable device on the FMonster, even ROM.
And data bus buffer is more or less required for each chip that outputs analog audio (that's every chip except OPL2 and OPL3).
CPLD or FPGA may be used in USB-variant to save space and cost. But ISA variant will use only hard logic ICs.

I believe that I can understand/share your feelings in that case.
Personally, I've used to have got no problems with, say, NES clone consoles (based on NOACs) until emulator boxes arrived (NES Classic, FIrecore etc)
Perhaps it's just me, but I love to work with hardware that's real ("alive") even thought that may cause more work or imperfections (though I also tinker with emulation, which is a neat invention).
The knowing that there's something being processed in real-time deeply within the circuits, simply brings joy to heart.
Even if his the results in some funny quirks or glitches in rare times (imperfect audio or visuals; ie, missing/flickering sprites/tiles).
An analogy to this might be MOD/MIDI (dynamic, but with varrying results) vs. MP3 or Wave/CD-Audio (high quality but fixed/pre-recorded).

Finally...
The video about first fully tested debug module - SN76489A/SN PSG/3-voice PSG/Tandy sound.
Video is in Russian, as usual. But automatically translated subtitles can help with that.
https://youtu.be/-4YIZy-07sI

And additional video, test of this module in 8 DOS games: https://youtu.be/DoE0Qtb7F4M

Dude, that's some cool stuff. Watching you adjust wiring connections while debugging the Police Quest audio was impressive. I can barely see the back of my computers to connect an LPT device sometimes, and here you are with video capture of you adjusting an ISA card while it's still in the slot. Totally nuts.

So what's the next board that's receiving the debug treatment?

MrSmiley381 wrote on 2020-01-19, 16:02:

Dude, that's some cool stuff.

Thanks!

MrSmiley381 wrote on 2020-01-19, 16:02:

So what's the next board that's receiving the debug treatment?

Probably SID (SSI-2001/DuoSID), but that should be easy.
Next... 2xSAA1099 (CMS) I guess.

Yep, Russian, translation at 7:42 says the CPU is a drunk 386:) ..and 8:18 turning on the machine is prophetic that it's all a woman so check the power supply:) Mixing girls, booze and sound cards there's bound to be fireworks.

yeah russian to german is also worse.
it barely makes any sense 😉