Totally off topic...mostly. Busted SH09 synth.

[sean k]

Friend of mine has entrusted his broken Roland SH09 analog synth to me to have a go at fixing and time is of the essence as he needs it for a gig soon. I did some testing of it and used an audio probe on some of the test point which are all at the outputs of opamps. The VCA had no output, the VCF had no output but the VCO square wave test point kinda blatts out a bunch of square waves for a half second or so on start up then goes dead except for a glitchy type single square wave intermittently every 10-15 seconds after that.

Colin at experimenters anonymous suggested it might be the power supply and I've yet to check that over but I checked the VC outs and the bender and keyboard check for Voltage at the proper intervals, or at least for what looks like the proper intervals.


You can see the square wave test point down in the right hand bottom corner at IC6a.

So I've figured out the voltage divider stuff up in the top left corner and the oscillator around IC6b and IC7a but can someone remind me of the function of the matched transistors that are IC5, and is incidentally a metal can device with lots of legs.

I'm kinda hoping that its either the caps or the fet in the oscillator, simply because they'll be easy to fix, but it's just wishful thinking really.

So my quandry is this. I have somewhat limited time myself at this time of year and I either have to fix this baby quickly or hand it on to someone who can... if I can even find someone. I'd love to have a bit more time with this as this is the area, analog synths, that really interests me these days and I'm pleased that one has fallen into my lap but its a bummer the time factor is such that I haven't got time to break it down and probe it with a scope, and learn how to use that as well.

Long post eh? any help would be greatly appreciated.

[frequencycentral]

That dual tranny is the throbbing heart of the VCO I think. Its a match pair in one package for the purpose of temperature stability.

The SH09 is a great little synth, I had one around 1984!

I would post over at the Electro Music forum for some more meaningful help.

[snufkin]

i would offer to give you some working sh09 voltages as i have one here but my meter is off at home and I'm here at my parents House for christmas

i have no idea on the problem

as stated above Electro Music forum diy section would be a good place to start

[Sir H C]

Some things to look for:

1. The power switch is 2 switches, I have had several of these where one half the switch didn't work so you get lights and some other stuff, but no synth.

2. Mechanical things break most often.

3. Measure voltages on the op-amps, are they right for the supplies?  Most synths use + and - rails, so one might be down. 

4. On most synths the VCO is always running, so if that is not happening, check voltages and report back that schematic bit with voltages on it. 

[sean k]

Thanks Chaps, I've posted over at Electro-music. Took me a while to figure out even where to post as it seems to be more for players as opposed to fixers and modifiyers... though I would quantify that with the idea that when those guys get into building they go really deep. So I've posted and I'll wait to see if the middle ground is picked up.

I was looking for the temperature compensation resistor around IC5 and didn't see one so maybe that IC has an internal Tempco... which is why they used it. I'll whip the lid off the synth again and get my magnifying glass out and get the moniker on that IC to check whether they are available and where else they are used.

[sean k]

Okay Sir, I'll do that... measure the voltages on the pins of the IC's around the VCO. Gotta go to work now though.

[Sir H C]

I have fixed an SH5, SH3, system 100 and an SH1000, so I have dug inside these things.  They are pretty well built, and not much goes wrong in there.

[sean k]

I've done a bunch of measurements and I'll scan the sheet and post it in an hour or two.

[Sir H C]

Cool, was wondering what was up.

[sean k]

So there we go, excuse the confusion around the pinouts on IC5. The left hand row is the actual pinout whereas the right hand row is how I counted them from the top. All the IC's in the area had 14.8VDC + and - on the power pins.
Pin 6 on IC4 may be the problem area as its at 0V and the bender and keyboard voltages are working fine up to R32 and R33. The other problem may be the collector on Q1 of IC5 which has almost +10V on R37 but no voltage on the collector.

Does this help you?

[Sir H C]

Those voltages all look pretty good.  IC4 off the emitters of IC5 keeps the collector voltage of Q1 at 0 volts.  In general, op-amps want the two inputs to be at the same voltage, and will force the output to try to get this condition to happen.  Looking at all the opamps, the input voltages all seem to be the same on the + and - inputs.  So the common point of R32 and R33 should be at 0 volts as the other input of that op-amp is at ground.  All said, this portion of the circuit seems to be working right. 

It would be good to look around IC6 and 7, as this is the sawtooth generation it looks like.

[sean k]

Okay, I'll go into IC6 and IC7 and see whats going on in there. That is the actual oscillator isn't it?

So the DC input to get the whole thing oscillating is pin two on IC7?

[Sir H C]

Yes, I believe so, I have to look closer to be sure, but the 1000pf looks like the oscillator cap, with the FET resetting it.