Saturday, January 9, 2010

Polysix on the road to recovery

Don't underestimate the power of battery acid!

So, Christmas and New Year vacation is over. But, the day before I had to go back to work I made a last effort to get something done since I'd spent most of my vacation just playing computer games or doing nothing. I'd gotten the Poly 61 keys working and felt like the Polysix needed attention, so...

Battery and damaged chip has to go
I knew there was trouble with the chip with the horrible fix next to the battery. I also decided that the battery has to go, even if it looks new and clean, who knows when it will start leaking.
So, I wasn't expecting any actual improvements from this but I decided to go ahead with it anyway. The battery came off and I cut the damaged 74LS08 off and removed the pins by desoldering them afterwards (easier than trying to desolder the whole thing at once).
I put two wires for the battery so I can place it somewhere else in the synthesizer later. Where the 74LS08 had been I put a milled socket, in case something goes wrong. Luckily I happened to have some 74HC08 chips at home so one of those should be good replacements for the LS chip. I also tried to scrape off some of that crusty greenish stuff from the acid damage and cleaned the surrounding area with rubbing alcohol (isopropyl alcohol).
Here's what it looked like afterwards:

It actually sounds like a synthesizer now
To my surprise, after placing the chip in the socket and powering up the Polysix I actually got "tones" coming out of it, not just noisy constant pitch sounds as before. It sounded correct!
However, none of the controls for setting VCO parameters were responding :(
I mean octave selection, PWM, waveform selection and such things...

Long boring search for error
Wondering if it was something wrong with the panel, I removed the knobs, removed the board and checked if the selectors and everything worked, and it seemed to so I put it back. Not mechanical failure there at least.
I looked at the schematics (internet is good) and tried to see where the panel board is connected to. Turns out, naturally, it's scanned by the CPU and it's a bit difficult to see if it works but I assumed it did. Then I checked the connectors to the voice board for the signals carrying the information (like octave selection, waveform etc) that wasn't responding. No change there. So, the voices never get that information. I assume the voices are not the problem then. Where to the signals going to the voice board come from? They come from the CPU of course, but via a latch. Aha. I checked the schematics again and compared the in and output of the latch for the waveform selection line. The input to the latch was on the data bus so I couldn't tell what was happening there but at least I saw that things WERE happening. The output never changed state. Why not? Didn't it get the correct data from the CPU or was the latch fried? I checked the clock pin of the latch and saw a steady 0V level with some noise. No clock pulses reach the latch? Hm... Where are they supposed to come from? Schematics say pin 35 on the CPU. I checked conductivity between pin 35 and the clock pin on the latch... No connection! Aha! But why?
Trying to see on the PCB layout in the service manual and on the board itself where the trace on the PCB leads is tricky. After a while I found it, and.. yes... of course it runs just under where the damaged 74LS08 chip was. So, I attached a jumper wire on the back of the PCB instead, restoring the connection between CPU pin 35 and the latch. Success! I could now change octave, waveform and everything! JOY!

Let me just tell you, that using a USB oscilloscope is messy when you don't have enough room to put the synthesizer near the computers. So, I had to look at the service manual PDF, go to the synth, place the probes, gently turn around and look behind me at the computer screen without slipping and causing a short-circuit with the probes, only to realize I forgot to alt-tab back to the oscilloscope application. So, remove the probes, go to the computers, press alt-tab to bring up the oscilloscope, back to the synthesizer, place probes, turn around and check the screen.

I think I did this mistake of not bringing up the oscilloscope screen like 10 times. VERY annoying... Ah well...

What's left to do?
Of course the keyboard is in bad shape and most of the keys are not responding. Hopefully I'll be able to clean them and get them to work. A lot of the buttons on the panel are bad and don't always respond and a couple of them look like they've been hit by something hard.

Something that worries me is that the voices are a bit detuned towards eachother and I also think that the voices don't track perfectly across the entire keyboard. I also noticed that the second voice is a bit louder than the others. This means calibration. Brrr... I hope I can do that right. I'm a bit nervous that I'll just make it worse instead :)

MIDI on the horizon?
I was not planning on putting a MIDI interface in this or the Poly 61 since the only ones I've found are expensive. However, I recently found a czech company selling reasonably cheap kits that at least implements acceptable MIDI functionality.
They have a MIDI kit for the Polysix and they have a kit for the Poly 61 under development. Great! If/when I get my synthesizers in good working order I will most definitely order one of each once the Poly 61 kit is available as well.
Here's the link to their website:

No comments:

Post a Comment