appiah4 wrote:

Interesting.

@root42 did you ever do an SBCHECK on the clone DSP firmware for the Snark Barker? Do you know what it reports?

No, never tried that. But will do that eventually.

Much, much delayed, but (technically) finished at last:

Not going to win any beauty contests, but only messed up one resistor (soldered the 5% tolerance version where I should have done 1%, found out when I tried to figure out where the 1% had to go...) and after sticking it into a PC, it all works perfectly 😀

dionb wrote:

Much, much delayed, but (technically) finished at last:
<snip>

Not going to win any beauty contests, but only messed up one resistor (soldered the 5% tolerance version where I should have done 1%, found out when I tried to figure out where the 1% had to go...) and after sticking it into a PC, it all works perfectly 😀

Looks fine! And the sound is more important anyway 😀

Cyrix200+ wrote:

dionb wrote:

Much, much delayed, but (technically) finished at last:
<snip>

Not going to win any beauty contests, but only messed up one resistor (soldered the 5% tolerance version where I should have done 1%, found out when I tried to figure out where the 1% had to go...) and after sticking it into a PC, it all works perfectly 😀

Looks fine! And the sound is more important anyway 😀

I'm impressed - not only with the fact I didn't mess up, but also with the result. It's as primitive as expected with SB1.x, but there's a lot less noise than on some SB16s I have (CT1750MCD 😦), OPL2 FM sounds like well... the way OPL2 should, digital audio sounds good so long as you don't throw too many sounds at it at once (OMF2097 did not go well, as expected) and CMS - all I can compare it to is a Sinclair 128k +2, and it sounds sort of familiar 😵

dionb wrote:

Cyrix200+ wrote:

dionb wrote:

Much, much delayed, but (technically) finished at last:
<snip>

Not going to win any beauty contests, but only messed up one resistor (soldered the 5% tolerance version where I should have done 1%, found out when I tried to figure out where the 1% had to go...) and after sticking it into a PC, it all works perfectly 😀

Looks fine! And the sound is more important anyway 😀

I'm impressed - not only with the fact I didn't mess up, but also with the result. It's as primitive as expected with SB1.x, but there's a lot less noise than on some SB16s I have (CT1750MCD 😦), OPL2 FM sounds like well... the way OPL2 should, digital audio sounds good so long as you don't throw too many sounds at it at once (OMF2097 did not go well, as expected) and CMS - all I can compare it to is a Sinclair 128k +2, and it sounds sort of familiar 😵

It does sound cleaner than it should, doesn't it? I think we also noticed that when we tested mine.

I also really like the CMS sound. Monkey Island (1) theme is also great in CMS. Prince of Persia also (but buggy on a faster PC (needs XT speeds).

C/MS is sometimes better in cases where the Adlib score received little love. And I bet some skilled demo coders and musicians could do something awesome with it.

root42 wrote:

C/MS is sometimes better in cases where the Adlib score received little love.

Any obvious examples?

If one used metal film resistors for the entire analog section there would probably be a significant reduction in hiss.

LABS wrote:

root42 wrote:

C/MS is sometimes better in cases where the Adlib score received little love.

Any obvious examples?

I think there might be very few. Maybe some of the Sierra games. I will think about that and come back to this. Definitely not many examples out there. I think we might need to make an SAA1099 tracker and get some musicians produce stuff. Maybe even a combined SAA1099/OPL2 tracker.

I finally tracked down what was breaking the OPL2 audio on my card - the LM393 is bad, which is odd, because for some reason I did socket that part and thought it was unnecessary to do so. After a whole lot of troubleshooting, trying a different Op-Amp, replacing the Y3014B DAC chip - it turned out that the +5VA line on the card, which powers the Op-Amp, was being pulled down to 1 volt instead of the 5 volts it's meant to sit at.
Removing the LM393, which is part of the microphone circuitry, puts it all back to normal. Though for some reason, I'm seeing some slightly inconsistent wave playback, like some sounds in games just arent playing, but that's potentially this weird 386 board I'm testing right now.
So anyone else coming across audio issues, check that the +5v and -5v going into the op-amp are both working correctly. It's cheaper than buying a new OPL2 + DAC.

My card is working properly now, the problem with missing audio may have been caused by my PC's setup.

I tested out the card in a case and it fits nicely 😀 - here's the backplate if anyone would like to print it out. Print it at 0.15 or 0.2 layer height, with supports in the orientation it should load as. That should print out on its side and the supports break away very easily.

Thermalwrong wrote:

My card is working properly now, the problem with missing audio may have been caused by my PC's setup.

I tested out the card in a case and it fits nicely 😀 - here's the backplate if anyone would like to print it out. Print it at 0.15 or 0.2 layer height, with supports in the orientation it should load as. That should print out on its side and the supports break away very easily.

soundblaster-bracket-printing.png

Amazing. Gettinh one printed out for myself.

Wow, nice printout. There are a bunch of chinese companies offering CNC and laser cutting sheet metal that require a 3D model as input. I might ask some of them for a quote on a very small run of these. If a single bracket comes out to 5 EUR max I might order some. But the 3D printed one will have to do for the time. Thanks for the design!

Replacing the LM393P chip on the card has resulted in the card being fully working again 😀 I'll have to try recording something (or play with the talking parrot) to see if that's working correctly, but swapping out the chip has got the OPL2 audio and everything else working. What a relief, I was worried I had mixed up a capacitor or resistor on the board.

How's the bracket fitting? I designed it in fusion 360 using CAD models of the parts it uses, so it should be accurate. It has a couple of optimisations to help it print reliably on an FDM 3d printer, but I can remove those if you want to use it as a general model?

So my placeholders are here! Next up I’ll call in some quotes from China for metal brackets.

And it fits. I had to widen the top hole a bit, since my headphone jack seems to be larger (it has a plastic sleeve), but other than that it's perfect. Thanks for the design!

root42 wrote:

So my placeholders are here! Next up I’ll call in some quotes from China for metal brackets.

E54D87B7-AB25-4E1B-8266-5252CCBFFA39.jpeg

If you find a good quote, I'm in - have messed around a bit, but no results remotely like this one.

Finally I've completed my Snark Barker, with "line" out wired before the volume pot. It sounds better with less noise, and the volume level is ok. Looks bit messy though.
The SAA1099 chips are pretty hot - is this normal?

^ I noticed the same thing with my 1099 chips installed on CT1350B , it it normal I guess

I love these things, and have been wanting once since I first saw this project but I cannot solder to save my life...

Is there anywhere that I can buy a pre-made snark barker? Or is anyone up for making one for me on request? I would compensate you of course 😁

Edit: Found one on ebay...I pay for others expertise in soldering 🤣. Oh well can't wait to give it a shot!