Arp String Ensemble

Arp String Ensemble

Arp String Ensemble

I have had three of these come into the shop in the last 6 weeks, and the last of these was picked up yesterday by the owner. I’m seeing failures of rare and expensive divider chips such as the SAJ110 (QDIP pin layout but the only NOS parts I can find are DIP so an adapter socket needed to be made), and SAH190 can. The Mullard C280 series “tropical fish” striped capacitors are falling apart on touch (you see cracks appearing at the ends), but wherever possible I leave them alone in the signal path as newer mylar types may have a different tone (the C280’s are sought after by vintage pedal builders for there tone, which I guess means they are non-linear!).

An SAJ110 on one machine was sucking many amps (it measured 12 ohms) and hadArp String Ensemble Internal view caused the power supply to fail, on another it was a Arp String Ensemble capacitorsmissing output. I recapped the power supply on two units, and for one of them we did a mass replacement of electrolytic’s as part of a restoration (and a sample of physically tired looking capacitors showed they were drying out with severely reduced capacitance.

The third unit had one dead key, and the owner had already changed a TDA0470 to no avail.Mullard C280 capacitor failed Looking with a scope the replacement chip was bad, and in fact it was ordered over Ebay from a vendor in China so who knows what it actually was. The original was put back in, and after much probing and deciphering of component layout the bug was traced to a bad resistor that had oxidized where the leg was cut off (they seem to have used some sort of pin flattening cutting tool in the factory which makes it hard to get components out), and then it had been re-soldered with a blob of solder over the top of the joint hiding the issue.

These keyboards sound great, but there are a lot of supporting components and rare chips that went into it so reliability is an issue.

Roland Juno 106

Roland Juno-106 800017a chip removed

Roland Juno-106 800017a board with chip removed

A number of these have come into the shop and for most of them it is a 80017a voice module issue. Roland put 3 chips (including the 3109) and supporting components onto a module, then covered it with a moisture sealing epoxy coating. There is a belief that the coating was required in order to protect surface mounted resistors from humidity change in order to stabilize their value, and/or protect their IP. This coating is thought to become conductive after 20 years or so. The only source of replacement parts is through chip pulls from scrap machines, and they are rare so costly. A typical symptom is a hanging voice, and/or crackling. I recently invested in a de-soldering station in order to get these Juno-106 Main cardchips out cleanly. The keyboard here was purchased on Craigslist by the customer for a good price but when he got it home he found it crackled badly, and a voice was hanging (you can isolate which of the 6 voices it is by putting the keyboard in test mode). For this unit it was not obvious which chip it was but careful testing showed it, and it could be verified with an oscilloscope. A replacement was obtained and fitted; all seemed well. After 3 hours of just sitting warming on the bench another chip started to crackle and had to be replaced. Speaking with the chip vendor this scenario is not unusual. The keyboard was not such a deal now that the cost of two replacement chips is factored in but the customer loves it and will get good use from it.

I have been using the acetone strip method to try and re-use a bad module before Roland Juno-106 Stripped 80017a Modulesscouting a replacement. It takes 3 days of soaking and careful work to get residue out from between the chip pins. I silicone seal the module after washing it in order to seal the above mentioned resistors from humidity. This method has been pretty successful with crackly modules, but has not been good for completely dead ones, or where the VCF is drifting as we are looking at chip failures in those cases. It is a sad fact that all of these 800017a modules will eventually fail, and one unit eventually did have all 6 replaced. I have not tried the aftermarket replacements for the 800017a, but would advise folks to consider that. Also there are places that Roland Juno-106 on the benchyou can send the card and they will acetone strip all 6 modules; I’m not sure what the deal is when the strip does not work on a module though.

The bottom line is that if you are thinking of buying a Juno-106 then be aware of its issues and that they are not cheap to repair.

via This Old Synth – Vintage Analog Synthesizer Repair Portfolio Page.

Roland SH-101

Roland SH-101 in red

Roland SH-101

This is the red version, and this one had the hand-grip. It had several issues, the most common of which was a touchy power switch (causes oscillator drift). I go 2 ways on this repair, either wiring across the switch so it is permanently on (no-go if batteries are being used), or order a replacement from TechnologyTransplant; the customer voted for the latter.

 The switch arrived and was fitted.   Switch cleaner was applied to all pots, switches, and sliders. I then re-soldered tracks around audio out and modulation connector, restoring switch operation and getting rid of crackles, and also re-soldered around the frequency and filter sliders as things were intermittent there.

 The bender was only operating in one direction, the LFO push function did not work, and the unit needed adjustment as putting the VCO amount slider forward caused a frequency change. The last 2 items were cured by setting up the position of the bender on the pot shaft. The big issue was that the LFO switch was bad, and they did not make this repairable. I was unable to find a replacement (although having stripped that area out on a Juno-106 yesterday I see it looks similar). I managed to remove the old switch and modify it so that a micro-switch could be installed, and this works fine.

The unit was finally set-up to spec and sounds great!

Rhodes Mk1 73 Stage Piano

Rhodes Mk1 73 Stage Piano

Rhodes Mk1 73 Stage Piano

This unit was suffering from multiple dead keys, and things had been getting progressively worse. There were other mechanical issues, and once they were resolved the electrically dead notes could be focused on. By tapping the coils center pole 12 pickups were found to be dead, and on removing one it could be seen that the inner end of the winding had rusted through where it meets the pin. There is not enough wire left exposed for me to attach a bridge wire to, so the coil was deemed beyond repair.  As I researched the Rhodes Mk 1 Stage 73 I found out just how big a problem that rust on the coils was, and that most used parts on Ebay etc are remnants of machines  scrapped due to this problem, with statements in the ads like “rust cleaned off”. Continue reading

Roland SH-5

Roland SH-5

Roland SH-5

This SH-5 came in with LFO-2 not providing modulation, and the led was dark, also the bender action was jumpy. The LFO-2 issue was due to a  bad CA1458 operational amp which was replaced and restored operation. The bender is metal on this unit and we were seeing a little grittiness due to wear and lack of lubrication. I stripped out the keyboard control panel and cleaned, lubricated (pot cleaned and a small amount of grease applied to metalwork), and aligned it. Tuning and scaling was a little off on both oscillators, and was adjusted to spec.

MiniMoog Musonics

Musonics MiniMoog

Musonics MiniMoog

This early unit (#242) needed some further work after another tech had worked on it. The keyboard had been re-bushed, but the bushing supports had not been straightened, which coupled with some badly stretched key return springs had left it unplayable. This was resolved, and once the set-up was completed (it had been tuned from C on the far left when in fact the first key is an F!) it played great (taking into account the temperature drift you get on a version 1 oscillator card!).

via This Old Synth – Vintage Analog Synthesizer Repair Portfolio Page.

Emu Emulator II

Emu Emulator II

Emu Emulator II

This one had a bad power supply, and the customer had been trying to fit more modern replacements. Re-soldering cold solder joints bought the base power supply back to life on the bench, and when fitted to the Emulator the root issue was shown to be a shorted tantalum capacitor on the input to the -15v regulator. The harness was in a sorry state where it connected to the power supply, and a new connector was fitted. Continue reading

Moog MiniMoog

Moog MiniMoog

Moog MiniMoog

The owner of this unit was having trouble finding someone to repair its dead oscillator and other issues because it had a home-brew CV interface installed that was not working. He had taken it to another repairer who would only work on it if the mods were first removed, but would not take on the mod removal task. It therefore came to me for mod removal and repair, and thankfully I had another MiniMoog in that I could compare the mod and original wiring with. Continue reading

Arp 2600 with bad VCO

Arp 2600 with a bad VCO

Arp 2600

This just came in with “squirrelly” keys, and VCO 2 was drifting badly when warm. The keyboard contacts were cleaned, and VCO 2 was checked with an oscilloscope. The VCO was found to have drifting frequency output, but stable inputs. The 2600 had sealed VCO’s (these early modules were encapsulated in solid epoxy, later ARP ones had an epoxy sealing layer and can be opened and repaired) so an aftermarket replacement was ordered from CEM. The replacement part arrived after some months and was duly fitted. The keyboard contacts were cleaned, and the unit set-up per the service manual.

via This Old Synth – Vintage Analog Synthesizer Repair Portfolio Page.

E-mu SP1200 Drum Machine

E-Mu SP1200

E-Mu SP1200

This was purchased by the customer some time ago, and he had stopped using it due to boot problems. It had sat in storage for some time and gradually degraded. It was a missing large touch-pad button, plus the boot diskette had been lost. Without the diskette there was no way of finding out what else is required, and the customer went off to find one.

A boot diskette was finally sourced, and after many attempts it finally booted. The boot issue needed to be addressed, and it showed a need for pad button switch cleaning. The LCD back-light was also out. Continue reading