Circuitbenders Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Author Topic: Partially fried PSS-460  (Read 8718 times)

GDorn

  • Newbie
  • *
  • Karma: 1
  • Offline Offline
  • Posts: 2
Partially fried PSS-460
« on: June 27, 2008, 11:54:43 PM »

Been working on a Yamaha PSS-460.  It has a bunch of nice bends, but I think I took one too far and fried something.  In the interest of science, I tracked down the signal from the sound chip (a YM3812, same chip used in the Adlib sound card) to the DAC and wired up sound output directly from the DAC.

From the DAC, I get fairly normal sound, though I think without any interesting stereo effect.  I also seem to be unable to change instruments.
From the normal sound out, I get a lot of noise.  I'm running it into a mixer into a pair of speakers, and if I turn everything way up, I can still hear the keyboard, buried in a sea of noise.

Has anyone run into this sort of fried board before?  I'm thinking maybe the preamp got fried, or I partially fried either the CPU or the sound chip.

If it's the sound chip, I probably could just recover one from an old sound card.
Logged

Gordonjcp

  • This person is dangerously insane.
  • *******
  • Karma: 78
  • Offline Offline
  • Posts: 1005
    • http://www.nekosynth.co.uk
Re: Partially fried PSS-460
« Reply #1 on: June 28, 2008, 10:29:12 AM »

If you get normal-ish sound from the DAC, where does the signal go after that?  Into an effects section?  It's possible the effects chip is fried (although I'd have thought that would be done digitally - I don't know though).  It's also possible that one of the opamps from the output of the DAC is fried.
Logged
If at first you don't succeed, stick it through a fuzzbox.

GDorn

  • Newbie
  • *
  • Karma: 1
  • Offline Offline
  • Posts: 2
Re: Partially fried PSS-460
« Reply #2 on: July 03, 2008, 08:06:18 PM »

From further experimentation, I'm not actually getting normal sound from the DAC.  Of course, I'm not supposed to be getting normal sound from the DAC, as the output of the DAC is actually both channels merged into one.  This then goes into a buffer, a BBD, and a gate.  This is sent as the right channel, and the output from the gate also goes into an inverter and becomes the left channel.  These go into the power amp.

Somewhere between the DAC and the power amp, a whole bunch of brown noise is being added.  I swapped the buffer chips around (there's three of them in different places) with no change in results, so that leaves the BBD and the gate.

At the time I fried it, I was actually working on a bend that crossed the digital out with the dac sync signal.  Crazy overdrive-ish effect.  Only it was too loud, so I thought I might bleed off some of that and ended up shorting it to something I shouldn't have.

So it's still possible the 3812 is fried, but I suspect one of the later ICs.
Logged

Gordonjcp

  • This person is dangerously insane.
  • *******
  • Karma: 78
  • Offline Offline
  • Posts: 1005
    • http://www.nekosynth.co.uk
Re: Partially fried PSS-460
« Reply #3 on: July 04, 2008, 08:27:11 AM »

Stop and think about it for a minute - if the output of the DAC is good, then the output of the YM3812 must be good too.  It's a digital chip, which has fairly complex programming requirements, so if it was faulty it would most likely not work at all.

The YM3812 feeds into some sort of external DAC, and off to the signal processing side of things.  Now, I'm not sure what you mean about the channels being mixed together, because as far as I can tell the YM3812 is resolutely mono.  I could be utterly wrong, though.

Anyway, the DAC is usually fed into some sort of buffer, and then if there are more than one output from the same DAC they're fed into a demultiplexer.  Think of this as being a rotating switch, like the distributor in a car engine - it switches from one output to the next in quick succession.  Try feeding a low-level signal into the signal path starting at the power amp and working back, until you find where it gets distorted.

Incidentally, 4051 and 4052 multiplexers seem to fail more often than they work.

Gordon
Logged
If at first you don't succeed, stick it through a fuzzbox.