In for the long haul…

There is a certain joy in getting a repair in, diagnosing it quickly, obtaining the spare parts locally, and then getting it finished the same day. If there was a Service Guy’s dream this would probably be it.

Alas this very rarely turns out to be the case, the variety of equipment in the workshop and in storage at any one time results in a very convoluted chain of spare component acquisition, often with several streams of parts arriving in short spaces of time from all corners of the world.

Also for thorough investigation and knowledge of the equipment under repair at the time, both the service and owners manuals have to be read. This leads to an endlessly rotating intimate knowledge of hundreds of pieces of music equipment at any one time, it is endlessly rotating because if a particular machine doesn’t come in for a while, it drops below the current info radar and has to be relearned again the next time one ariives in.

Most pieces of equipment do get to leave us in a reasonable time, but some are just so broken, or difficult to obtain parts for that they are in for the long haul.

The Juno 60 mentioned back in the April blog is actually coming along nicely, its only remaining electronic faults are in two of the filters, the resonance control is non existent. This could be due to the Resonance VCA chip, but it could also be due to faulty coupling capacitors, obviously the latter will be cheaper to repair.

The Polysix also from April also got another going over in late May, more components were cut out, the circuit board area underneath thoroughly cleaned, and components replaced, but this one is still not right, voices 5 and 6 are just out of tune in a random way every time the synth is turned on.
We monitored the Korg forum for any advice, and some was given, and it was good advice, but it either didn’t work or didn’t apply to this particular problem.

These kinds of repair are the long haul *uckers, but both of these are the result of buyers hunting for silk purse synths at sows ears prices.

We suspect that during the Analog Renaissance, a lot of the dross of the synthesiser world is being dug up and offered for sale, after being dumped into sheds and attics all across the world. As for the non-smoking studio use only line, all we can say is “please knock it off “, you are only fooling the purchasers, but not the service engineers. Everything that comes through our humble workshop gets scrupulously cleaned while it is dismantled, and the amount of Ammonia cleaner and cleaning cloths that get used tells us that no synth ever lived in a smoke free environment.
And another thing, what really grinds my gears is the insertion of the phrase “RARE” on every entry on E-Bay these days, (is a Juno 106 rare, or a DX7). In our opinion, there are less than two dozen synths that are actually rare.

Rant over for now, there are three more weeks in June to go…
Teisco a go go

A Teisco S110F arrived in a couple of weeks ago, apparently working fine until recently, but now only producing weird buzzing noises from the output.

The obvious culprit here was going to be the power supply, and indeed something about it was more than a bit flakey, the 15 volt positive rail was absolutely fine under all circumstances, but the 15 volt negative supply seemed to have a mind of its own.
On switch on of the synth, the negative voltage wanders from -9v to -13v as it warms up, it does get better as time goes by but it is never useable..
As the S110F is built from 3 circuit boards behind the front panel, the logical thing to do was to disconnect them all from each other and try to trace what was upsetting the power supply.

As all the circuit boards are single sided, there are a lot of wire links on the top side, and much care was taken carefully isolating one part of the circuit board from the next to attempt to track down a faulty component.

A hint was leading towards an OpAmp that fed the uA726 on the left hand board, this is VCO 1, when it was isolated, the -15 V seemed to hold up accross the entire circuit board, and as this board is responsible for all keyboard control including Portamento and Sample and Hold, this is a good place to get right.
Replacement of this Op-Amp did seem to lead to a temporary respite, we had good plus and minus voltage accross the entire circuit board, and for a while it seemed stable.

We knew that plugging in the second pcb would upset that again, due to some extensive initial testing, which led us to believe that there could be more than one dead IC in the synth. The same isolationist technique was used on the second pcb to try and ascertain which chip it was on this board. This board has VCO2, the mixer and the filter, so it is an important one for the synth.

As expected, the negative rail collapsed when the second board was plugged in, so for confirmation it was unplugged from board 1 and the system tested again with only the VCO 1 board plugged in.
The Negative rail had collapsed again.

The power supply pcb was taken out and had all electrolytic capacitors replaced and the big ones upgraded with higher capacitance and voltage rating as the originals are old, and no harm would be done.

Still not a lot of joy, a closer exam of the voltage rails on the scope showed a lot of ripple on the negative rail directly at the power supply itself, and as all the caps had been replaced it couldn’t be that.

A break of a day or so and we got back to this little baby.
Logic dictated that if it wasn’t the power supply capacitors, then it was likely to be the power transistors that were suspect, and it turned out one of them was. The measurements of the device out of circuit seemed to indicate it was ok but some kind of instinct kicked in and said that it might not be.
We didn’t have the exact transistor as a spare but substituted a European over specced equivalent just to be sure, and lo and behold, instant power supply joy.

We replaced both PSU transistors with a complementary pair of Euro equivalents and it was all good.

Soak testing for many hours and checking out all the functionality of the synth showed good results.

One slightly unusual aspect of this synth is in its filter, it is a Moog style Ladder filter built with the same type of matched transitor pairs used in the TB303 (only more of them), but the interesting thing is the size of the filter capacitors, as in the electrical value.
The value of them is 220 nanoFarad (usually abbreviated to nF and indicates the amount of electrical charge it can hold).
If you substitute the concept of electrical charge for the bucket of water analogy, then simply think of more nanoFarads being a bigger bucket. If you imagine that the bucket can only be filled by a hosepipe of a fixed size and that it also has a hole in it, then you are getting close to a rough capacitor model.
We are not going into a long diatribe on capacitor theory, but a few numbers may illustrate our point.

Teisco S110F capacitor 220nF
Moog Taurus capacitor 100nF
Minimoog capacitor 68nF
Micromoog capacitor 33nF
Moog Prodigy capacitor 27nF
Moog Source capacitor 10nF

The pattern is obvious, the size of the capacitors in the ladder filter got smaller as time progressed, yet the 2 Moogs at the top of the list are regarded as the ultimate Bass machines, does the Teisco have a secret weapon that has been kept hidden for a few decades?

Not for us to say, but the bass end is certainly good, but not quite like a Moog, which considering the filter architecture, is a little surprising…. getting late more to come…

Room 101

As anyone who has read the book or seen the movie of George Orwell’s 1984, room 101 is where you meet your greatest fear.
Normally a Roland SH101 should hold no major fears for a synth tech, there are several quite common faults with these, the blown power supply, the dead Curtiss oscillator chip, the odd Op-Amp failure in the LFO or elsewhere, but nothing that will tax the competent soldering iron wielder.
And then… from beyond the grave… lurched in this one…

It all began when the phone was answered, “I have been offered a dead SH101 on and I am thinking of putting in an offer”. We naively answered in the Top Gear idiom, “How hard can it be?”

The synth was purchased, and a week later came our way to be returned to its former glory and then when examined, the true horror began to unfold.

The legend was that this synth was bought a long time ago as faulty, but never got repaired as the original owner just bought another one to replace the dead one, and never got around to it.

We are seeing this story or something like it more and more on auction sites these days, along with lines like
“I never got around to it” or “it should be an easy fix”. There is no more profitable way of clearing your attic than this kind of chicanery, and if you say it is faulty, and sold for spares, then you are not liable for any counterclaims by the subsequent new owner, and the current greed and lust for anything analogue at a cheap price is only adding to the misery.

The power supply in this one was faulty, the main power transistor in it had already been replaced by a previous solderer, and the transistor type used (not an original) seems to be a good choice. But another transistor and diode in the power supply was also dead.
The Curtis Oscillator chip was actually missing, none of the LED’s worked, sending noise through the filter produced no output. Further testing showed the filter chip to be also dead.
Hunting down why all the LED’s were dead led to the discovery that all the transistors involved in the LED switching were blown, as was the IC that also drives some of them, and one of the LED’s too was blown.
Having replaced 7 other transistors on the main pcb and replaced any necessary IC’s, we finally had some semblance of a functioning 101. The synth powered up, the correct LED’s came on, the function switching for the Arpeggio, Hold, and Sequence Load and Play were now all working, The CPU chip also seemed to be working, the only survivor on the board so far.
After getting this far, time to see if the unit will calibrate its DAC and keyboard interface, there is a test diagnostic routine built into the CPU firmware, you just hold a couple of buttons down at switch on.
Guess what? the keyboard CV’s were all over the place, and the demultiplexer chip also seemed dead.

The only parts of this synth that actually work are the LFO, the noise generator, and maybe the CPU, (I say maybe because one of the pins is producing a strange waveform unlike any of the others, and it does pertain to the DAC system.

We are waiting for a new CEM3340 Oscillator chip before proceeding any further, our inclination was this one was a true scrapper, only the keys of the keyboard are worth anything, although the slider pots are also good.
However the client wants it restored, and it won’t be cheap, but it can be done.

As of now the timescale is November, but for continuity purposes the following text is here.

The new Curtis chip was installed and worked fine, the new (?) filter chip was also fine, the only remaining problem seemed to be that the calibration still didn’t work, and one Octave of the keyboard wouldn’t play.
Guess what, the CPU was faulty after all, only one pin of it but it controls the missing octave of the keyboard as well as one section of the DAC.
A replacement is available from Vintage Planet ( but is €50 and means the SH101 will never run on batteries again, unless you own the Duracell company.

This synth has been an absolute dog from start to finish, over 75% of the transistors in it were faulty as well as 80% of the chips (including all the expensive ones), a couple of rare and valuable slider pots had to be replaced and the obligatory new power switch.

The reason for the long delay between May and November, there was an option of a free badly damaged 101 which the owner was giving away which we hoped had a working CPU chip. Alas the owner of said SH101 has decided that he wants it restored, the ridiculous prices on E-Bay for this little plastic POS have changed his mind to the extent that it is worth spending some money on, so we will probably see this one soon.


Leave a Reply

Your email address will not be published. Required fields are marked *