A report on simple repairs for your amusement
2009-05-06 by Alan Probandt
Here is a list of some repairs to synths that you might find interesting. The tone module repairs have all been on MIDI synths from the late 1980's on. I could never and probably will never be able to afford pre-MIDI analog synth, so I'll probably never work on one.
-Alexis Nanosynth: Loud buzzing noise when powered on. Open main filter electrolytic capacitor in the power supply as if the cap were not there. Replaced the four electrolytic filter caps of the power supply.
- Kawai K1M: No power: no sound. Bad fuse in main power. Fuse looked OK visually but tested open with voltmeter. Replaced fuse.
- BOSS/Roland DS330 Sound Canvas synth. No power: no sound. Traced to bad power converter (wall-wart). This device has a custom power plug/connector. It looks like a normal 2.1mm connector, but the pin diameter is very wide (@ 2.8mm). I replaced the power jack on the DS330 printed circuit board with a standard 2.1mm jack and used a normal 300mA /+12V wall wart.
-Yamaha desktop tone module (forgot model) No power:no sound ; Some previous user must have plugged wall or mains directly to the input because the power supply diodes had exploded. Replaced diodes with a 100V bridge rectifier, the filter capacitors, and the 78xx voltage regulator chips.
-Roland MT-32 no response to MIDI input. Used MIDI controller set to channel 2. In what must be the most stupid thing ever done by a synth maker, Roland set the MT-32 to NOT respond to any input on MIDI channel 1, which is the standard default channel for all other tone modules by all other manufacturers. Channels 2-10 work fine.
Yamaha TX81Z no response to MIDI input. User plugged in a home-brew controller that had no 220 ohm resistor on its MIDI out. On the first version of the TX81Z, Yamaha had no current-limiting 220 ohm resistor on the MIDI in. Result was a burned-out MIDI IN opto-isolator on the TX81Z. Replaced with a similar device.
DOD FX9 digital delay unit for guitar ; No output signal. This is a guitar stompbox. Previous owner attempted to repair a broken input phone jack, but got the input and output jacks reversed because they were both the same part. I had no schematic and attempted to trace the signal from the input jack through the audio conditioning circuitry to the analog-to-digital converter. The audio input input went to the output of an op-amp and the unit's output went to the input of an op-amp. Reversed the connectors in the chassis box and it worked fine.
Zoom multi-effects for guitar loud buzz at long delay settings, other effects (chorus, distortion/overdrive, reverb, etc...) all fine and short delays (< 175 milliseconds) were fine.
One of the tiny pins on the data bus between the processor and the main memory was lifted due to a bad pick'n'place soldering at the factory. Looking at the outputs on all of the pins on the bus one-by-one with an oscilloscope, I noticed that the output from the processor was 0-+5V and the input to one of the memory chips for the same trace was +2.5V. Continuity test with ohmmeter showed broken solder connection. Put a drop of flux and quickly touched the pin with a hot (@400 degree F) fine solder tip to re-establish the solder connection between the tiny Quad-Flat-Pack chip's pin and the circuit board. Worked fine again. When the delay was short, the delay signal was never written to the section of memory that was activated by this broken address bit, so short delays were normal and longer delays had distortion from trying to access the wrong sections of memory.
Lots of simple things. Synths, tone modules, and guitar effects are made a lot better now than they were in the 1970s.