Vintage Synth Repair

I'm working on a SH-101 that has tuning problems, and I can't seem to
get the modulation to stop. I am not very familiar with operation of
this, so some may be due to operator error. I have the service manual.

At one time there was a C battery leak, but it apparently was lying
flat the whole time, the corrosion was contained to the metal bottom
lid.

The tuning is way off, the entire keyboard doesn't quite cover one
octave. I went through the D/A tune/width/linearity adjustments with a
voltmeter, the D/A width was off a few millivolts, but did not change
the problem. The next steps of VCO adjustment require a scope. I
recently acquired an old Tektronix 465, but have no probes (is 100mhz
passive good for this?). I am a scope newbie, and am reluctant to
trust my limited knowledge on this.

The service manual states:

"Connect the oscilliscope to the SH-101 OUTPUT. Supply the reference F
note (based on A = 442Hz) to the scope EXT. Input for the Lissajous
figure."

Does this mean connect the SH-101 to channel 1 or 2 of the scope, then
supply a reference F from another keyboard to the EXT input? I assume
a sine wave would be good, for the least amount of overtones. The
further adjustments are talking about adjusting until the figure is
motionless. If the scope is simply being used to make sure the SH-101
is in unison with an external synth, I have a good enough ear to
detect beating.

Can I stumble through parts of this just with an instrument tuner?
What kind of probes should I get for an antique scope like this? Is
this a low-tech enough use that I could perhaps hook up straight wires
from a BNC connector, or will the impedance be too far off? I live in
a very rural area, and would like to work on this before the amount of
time it will take to order probes.

The modulation seems to be always on and affecting the sound. There
are Mod sliders for VCA and VCF, both seem to work throughout their
range. I have read that all the sliders are the same, so swapping some
could be an option, but it looks moderately hard to get to (not really
hard, just further disassembly). I am always leery of 20 year old
plastic connectors and wiring harnesses being brittle.

Still further, the CV input jack seems to have something stuck in it.
From my limited knowledge of CV, I think the keyboard is disconnected
if it is being operated by CV (not the case on this one). I don't know
if that could contribute to some of these symptoms, I will remove it
to rule that out.

Thx in advance,
Scott S.

It looks like VR6 may be my friend in the tuning issue, I'll check it
tonight.    http://www.guerillasofsoul.com/chad/sh101/101calibrate.txt
There is also a 300dpi scan of the service manual that went up a few
months ago. It's much better than the very old one floating around the
web. http://matrixsynth.blogspot.com/2009/03/new-scan-of-roland-sh-101-service.html

Also, perhaps a 1M resister in series may make a better simple scope
probe than a straight wire?

thx,
Scott S.

--- In vintagesynthrepair@yahoogroups.com, "Scott S." <picoshark@...> wrote:
>
> I'm working on a SH-101 that has tuning problems, and I can't seem to
> get the modulation to stop. I am not very familiar with operation of
> this, so some may be due to operator error. I have the service manual.
> 
> At one time there was a C battery leak, but it apparently was lying
> flat the whole time, the corrosion was contained to the metal bottom
> lid.
> 
> The tuning is way off, the entire keyboard doesn't quite cover one
> octave. I went through the D/A tune/width/linearity adjustments with a
> voltmeter, the D/A width was off a few millivolts, but did not change
> the problem. The next steps of VCO adjustment require a scope. I
> recently acquired an old Tektronix 465, but have no probes (is 100mhz
> passive good for this?). I am a scope newbie, and am reluctant to
> trust my limited knowledge on this.
> 
> The service manual states:
> 
> "Connect the oscilliscope to the SH-101 OUTPUT. Supply the reference F
> note (based on A = 442Hz) to the scope EXT. Input for the Lissajous
> figure."
> 
> Does this mean connect the SH-101 to channel 1 or 2 of the scope, then
> supply a reference F from another keyboard to the EXT input? I assume
> a sine wave would be good, for the least amount of overtones. The
> further adjustments are talking about adjusting until the figure is
> motionless. If the scope is simply being used to make sure the SH-101
> is in unison with an external synth, I have a good enough ear to
> detect beating.
> 
> Can I stumble through parts of this just with an instrument tuner?
> What kind of probes should I get for an antique scope like this? Is
> this a low-tech enough use that I could perhaps hook up straight wires
> from a BNC connector, or will the impedance be too far off? I live in
> a very rural area, and would like to work on this before the amount of
> time it will take to order probes.
> 
> The modulation seems to be always on and affecting the sound. There
> are Mod sliders for VCA and VCF, both seem to work throughout their
> range. I have read that all the sliders are the same, so swapping some
> could be an option, but it looks moderately hard to get to (not really
> hard, just further disassembly). I am always leery of 20 year old
> plastic connectors and wiring harnesses being brittle.
> 
> Still further, the CV input jack seems to have something stuck in it.
> From my limited knowledge of CV, I think the keyboard is disconnected
> if it is being operated by CV (not the case on this one). I don't know
> if that could contribute to some of these symptoms, I will remove it
> to rule that out.
> 
> Thx in advance,
> Scott S.
>



The Tek 465 is one of the best scopes ever made. 100Mhz passive probe would be fine. Make sure you cal the scope against your new probe when you get one. There's a Calibration loop on the front of the scope

There's also a TEKscopes yahoo page for seasoned people and newbies

On Tue, Jul 7, 2009 at 7:46 AM, Scott S.<picoshark@...> wrote:
> It looks like VR6 may be my friend in the tuning issue,

<snip>
No luck on VR6, the entire keyboard covers only a fourth, no matter
where it is set. From the 'tuning without a scope" URL:

_13. Alternate playing the lowest key (F3) and highest key (F5) while
_looking at the tuner/counter.
_
_14. Adjust VR-6 until both notes are exactly 2 octaves apart. VR-6
_adjusts the frequency spacing (width) between the two notes. It should
_be exactly 2 octaves, however VR-6 can make it greater or less than this
_amount.

I didn't have a tuner/counter connected, but I can hear a fourth.  The
D/A tune/width/linearity voltages are still within spec. Suggestions?

TIA,
Scott S.

At 21:09 Uhr -0500 06.07.2009, Scott S. wrote:

> Still further, the CV input jack seems to have something stuck in it.
> From my limited knowledge of CV, I think the keyboard is disconnected
> if it is being operated by CV (not the case on this one). I don't know
> if that could contribute to some of these symptoms, I will remove it
> to rule that out.

Did you remove it, whatever it was?

Next step: Measuring the output of the CV out. We need to narrow it down to
the section that is reponsible for the issue. Perhaps it is the DAC,
perhaps it is the oscillator itself.

The voltage change at the CV out should be 1 volt for each octave. How do
things look at this point?

-- 
Malte Rogacki gacki@...
----------------------------------------------------------------------------
"Don't forget to TURN ON THE SYNTHESIZER. Often this is the reason why you
 get no sound out of it." (ARP 2600 Owner's Manual)
----------------------------------------------------------------------------

The CV is outputting more or less one volt per octave. I say more or
less, because some are off by as much as a hundredth of a volt, but
since the whole thing is not yet calibrated, that is approximately
correct. They remain constant across all three L/M/H positions of the
transpose switch. The lowest possible is displaying .4 volts, the
highest possible is 5.01 volts. This is on a quick powerup, without
letting it warm up.

I will remove the CV input jack tonight. It looks like it is a
normally closed jack, with it's center connector normally connected to
the center of the CV out jack. I will jumper these two connections to
simulate the jack being in place, so that should work for
troubleshooting. If it's being held open, that is not the normal
operating condition. Thank you for your help!

thx,
Scott S.

On Wed, Jul 8, 2009 at 9:55 AM, Malte Rogacki<gacki@...> wrote:
>
>
> At 21:09 Uhr -0500 06.07.2009, Scott S. wrote:
>
>> Still further, the CV input jack seems to have something stuck in it.
>> From my limited knowledge of CV, I think the keyboard is disconnected
>> if it is being operated by CV (not the case on this one). I don't know
>> if that could contribute to some of these symptoms, I will remove it
>> to rule that out.
>
> Did you remove it, whatever it was?
>
> Next step: Measuring the output of the CV out. We need to narrow it down to
> the section that is reponsible for the issue. Perhaps it is the DAC,
> perhaps it is the oscillator itself.
>
> The voltage change at the CV out should be 1 volt for each octave. How do
> things look at this point?
>
> --
> Malte Rogacki gacki@...
> ----------------------------------------------------------
> "Don't forget to TURN ON THE SYNTHESIZER. Often this is the reason why you
> get no sound out of it." (ARP 2600 Owner's Manual)
> ---------------------------------------------------------

OK, that shows that the DAC is working properly.

Next test: Put the filter into self-oscillation (resonance max, cutoff so
that you hear some sound) and set keyboard tracking to max as well. Play
some notes of check if the oscillating filter itself (not the oscillator)
roughly tracks properly or if the behaviour is identical to the oscillator.

BTW, those small offsets of the CV are nothing to worry about at this
point. The synth should work almost prefect even with those small offsets.

You're right, the CV in jack is normalled; when thre is no CV in connected
the keyboard CV is used automatically (same for the Gate). The CV then
passes through the Portamento circuit to the oscillator and the filter.

Another question: Does the "Range" switch change to tuning of the
oscillator (it will not change the filter, though)?

-- 
Malte Rogacki gacki@...
----------------------------------------------------------------------------
"Don't forget to TURN ON THE SYNTHESIZER. Often this is the reason why you
 get no sound out of it." (ARP 2600 Owner's Manual)
----------------------------------------------------------------------------