Resurrection of the Dead Synth

It’s been a while since I’ve posted anything technical here. Time to change that!

So, I’ve had this Yamaha TG33 for a while (bought it in 1999 or 2000)- I used it pretty regularly until some night in the 2000s when it suddenly stopped playing during the middle of a radio broadcast. Technically, it didn’t stop, it just suddenly started playing really quietly, even though the volume was cranked. I got as far as taking it apart to determine if it was the volume knob, but it appeared not to be. I was then stumped, and decided to shelve it until I could decide what to do with it later. At some point in 2008 or 2009, I removed the joystick to make a little X-Y CV controller for my MFOS Sound Lab. I later bought a Korg Wavestation SR to get my vector on.

Fast-forward to a couple of nights ago. There was an SY22 (sort of the keyboard version, but with a few changes) that was very tempting on eBay (61 keys, velocity/aftertouch, voice architecture I like, etc.). At some point I was doing some due diligence and reading reviews and whatnot, when I found a statement about how SY22s have a known issue that causes the audio to suddenly cut out and get really quiet, and that the fix is relatively simple- replace a transistor that controls a relay on the outputs.

Armed with this information, I thought that perhaps the TG33 might share this same problem. I pulled it from where it had been sitting and took a look around. I didn’t see anything like a relay anywhere on the PCBs. Hunting for the Service Manual proved mostly fruitless (there are some sources that may have a manual for sale, but none seem to be available online). However, the SY22 Service Manual is readily available online, so I grabbed that to see if that offered any assistance or enlightenment. There was no major enlightenment, but there was some wisdom to be gleaned from it.

From tracing back from the outputs, I can say that in place of the relay, the TG33 seems to use a bank of four 2SC2878 NPN transistors, which were specially marketed for “muting and switching applications”, and are designed with “reverse Hfe” in mind. There was also what appeared to be electrolytic capacitor leakage in the general vicinity of those capacitors (though I don’t think they’re damaged as one is outside of the “leak zone”), and some more near another transistor that may be related (a 2SA or 2SC1015), which may be the one at fault. I replaced both of the caps, as I had replacements on-hand.

Since replacing that transistor may not fix the problem, I decided to do some more sleuthing around the boards to see what else is possible. The SY22 service manual does have a little information about the VCA that the TG33 uses, a Mitsubishi (now Renesas, I guess?) M51132L dual VCA (two, actually- one for each pair of outputs). I made up a quick audio probe, and started poking around the pinout that I had- attaching to the input pin did give me a nice loud signal!

I think my next steps (when I get some time) are:

1. Poke around the VCAs some more- see if the output signal is as loud as the input signal.
2. See what the voltages on the VCAs are, and if they change when the volume control is turned.
3. Poke around that lone transistor that may have had some cap leakage, see if it should be replaced.
4. If it turns out that it’s both VCAs that have failed or something I can’t outright determine or fix, I can wire up jacks directly to the VCA inputs and just worry about volume on an external mixer.
5. Oh yeah, and I’ll have to remove the joystick from its current home and get it back into the TG33. Could be a challenge…

So, I’ve got at least one hack solution if I can’t determine the actual cause. If it’s not a full victory, it’s certainly an okay one. And no, I didn’t win the auction on the SY22. If you’ve got one of those or an SY35 you want to part with for cheap, let me know 🙂