I don't detect a clock going to pin 16 on the DAC (D6376). When it's "playing" audio, it does create clocks to everything else, including the hex switcher. That signal varies depending on the actual MIDI being played, as well as the RAM it's pulling wavetable data from. I checked for continuity from the clock pin on the DAC to pin 60 on WaveFront ICS2115V, and it reads 0 ohm. Now to see if I can find out why the WaveFront ICS2115V isn't generating (or passing?) a clock.

(Edit: I tested wrong. There is a proper clock going to the DAC. So now, I'm back to trying to figure out why it seems to be processing, reading and writing RAM, and the two WaveFront chips are operating, but there's no actual audio going through the DAC. It's as though the chipset is going through the motion of playing MIDI, then throwing it in the trash.)

Can setwf143 see the card? You can also set the volume with that tool. What MOS file are you loading?

I just grabbed a copy and am trying it now. I've never used this tool before, so I'm going to spend some time researching and playing with it. Thank you! I've been loading OSMAUI.MOT. I was tempted to try the TBS-2001 tools and see if they produce different results.

"setwf -s" says

WaveFront Hardware Version: 1.64
Firmware Version: 1.14
Synthesizer Volume: 119
Number of Voice: 32
Synthesizer Tuning: 0
Virtual MIDI Mode: Disabled
Virtual MIDI Switch: Synth
MIDI-In to synth Mode: Enabled
Number of samples: 40
WaveFront Sample DRAM size: 262144 bytes.
WaveFront Sample ROM size: 1048576 bytes.
WaveFront Free Memory: 262144 byte(s)

I'm testing right now with the RAM removed, but it does report the additional RAM when I insert it. I'm curious whether the card is sending all MIDI out through the DB9 plug. (I'm also curious whether soldering in a WaveBlaster connector would allow it to run with a SB16.)

The output shows that volume has a correct value. Setting an illegal value can happen but isn't the case. I'm confused however by "Number of samples". I can't recall what I had with the Tropez/Maui but it seems an odd number. WaveFront does 512 samples max. and uses ~350 samples in ROM if I'm not mistaken. WaveFront Sample ROM size: 1048576 is odd as well. It states ~1 MB, you should have 2 MB of ROM.

I would remove all memory and inspect the solder joints around the ROM chips and run setwf again. Perhaps it's playing but missing instruments.

That would fit with its behavior, particularly if the ROM is only giving out zero's. Bank switching is happening during playback, too. Either a RAM chip flattening out all of the data or the ROM failing to push out any samples could do it.

I'll try that next.

I haven't sourced any 514256 chips I can use, but they are Toshiba chips (which are dubious for the period). *shrugs* I ordered some OKI chips, same speed. I did find some near shorts on the caps near the RAM, so chances are one of them is shorted out. (I did pull the caps and test them, just to be safe.)

I wish I had this level of understanding when my Tropez Plus died. That was my dream card...and I have not seen another for sale for any price.

What was the issue with the Tropez Plus? Those are rare but one of the greatest cards, at least on paper.

It simply stopped responding. I did a quick check for broken solder joints or shorts, but I didn't know enough to dive any deeper. I sold it on eBay for parts/repair, and even in non-working condition it sold for more than enough to buy this Maui. I know that wavetable sound is a very personal, subjective measurement, but I've played music samples for quite a few games and MIDI files for friends who were not around for the great wavetable battle of 1994, and they all agree that the Turtle Beach's samples and processing are hands-down the best. It was a real PITA to get working, but that's another story. 😁

I was hoping there would be a way to load at least the samples from the TBS-2001 into the Maui, but I had not gotten that far before this card stopped working.

Have you plugged in some speakers?

Yes I did, and I also probed the DAC to see if it was receiving or sending any signal. It has power and clock, but it's receiving zero's and sending silence.

Phooey. So far, I've confirmed that none of the data or address lines on the two TC514256AJ-70 DRAM chips next to the RAM slots had any activity. I replaced them with Siemens HYB514256BJ-70 chips. It identifies the 256K of RAM. (If I disable the chips, I can put 2M in the slots and it recognized just the 2M.) It now shows 3 loaded samples. Running MAUIDIAG used to pass, but now it results in ROM checksum errors. I reflowed the pins to the ROM and have the same result. Now, it *could* be that it doesn't like the Siemens chips. I haven't tried the OKI chips yet, and I *could* order the original PN Toshiba chips. It's also possible that there is a short elsewhere that's keeping the RAM at zero.

Edit: I'm stupid. I didn't flow the SampleROM. I flowed the boot ROM. I re-flowed the SampleROM and now it has 332 samples loaded.....but now it's zero bytes of DRAM! D= Gah, I can't win! Funny enough, DIAGMAUI doesn't bat an eye at having zero RAM.

I GOT THE *!@#%$!@$# WORKING!

This card must have had a fall or something. The RAM chip closest to the slot *still* doesn't have any data out. After reflowing the ROM, I got 332 samples loaded. Then, I added a matching 256K SIMM in the first slot, and it seems to be piggybacking the onboard RAM chips. (Removing the RAM chips and putting RAM in slot zero does not work. The card won't start.)

So, from what I can tell at this point, it's doing R/W to chip 0 onboard simultaneous with chip 0 on SIMM 0 while disabling chip 1 onboard and relying on chip 1 on the SIMM. 1M stick in slot 0 does not work.

I'm not sure if I want to try replacing the onboard RAM yet again or just being satisfied with a functional card.....

Edit: The OKI chips identify (incorrectly) as 1M chips. That makes me wonder if the board *would* support 1M chips onboard. I'm just too fatigued from one after another chip swap to risk any more changes. I'm ordering two NOS Toshiba chips of the exact same P/N.

Edit 2: I pulled the OKI chips back off, and pin 5 appears to be NC. 1M onboard chips won't work anyway. It seems like it should be possible to bodge that pin in (and theoretically go above the 8.25M maximum).

Glad to read you got it working. So, if I'm not mistaken, the on-board memory died and caused all these issues?

It was more the solder joints on the WaveROM, but the onboard RAM was shorted too. I feel like something must have happened to the card, like an impact or a surge.

I didn't mention the incident of failure. It was working fine until I decided to install Windows 95. It had 3.11 and was working fine, and it went bad sometime while installing 95. Turns out that was irrelevant, but the strange thing is that it the card didn't go anywhere between working and not working.

*sigh* Well...it's dead again. The WaveROM isn't recognized. I removed and resoldered it and checked all of the solder joints. At this point, I'm shelving it and searching (in vain) for another card.

On a hunch, I reflowed all of the joints on the two adjacent chips (the WaveFront IC and the Motorola CPU), and it's been working since then. Maybe the solder joints on the entire board are cracking. In any case, I'll keep my ears out for another TBS-2001, even in non-functional condition.