Solo Setups
Harmonicas
Pedal Keyboards
MIDI Guitars
Delays & Loops
Other Tools
Sounds

Index
What's New
FAQ
Forum


Revised
21 July
2000

Troubador Tech

Casio PG-380/310/300 Problems CURED:
It's the capacitors!

If your Casio PG-380, PG-310, or PG-300 has developed tracking problems, particularly after a period of disuse, you're not alone. My own PG-380 exhibited such behavior. Also, since I first published the Casio MIDI Guitars page in the summer of 1996, I have received many, many emails from other PG-380/310/300 owners with similar complaints-- so many, and so similar, that it became clear that the problem is both widespread and intrinsic to the PG electronics. If your instrument displays these symptoms, there is a very good chance it can be cured with a capacitor transplant. First, though, you might want to try juice therapy; it worked for me.

Symptoms: Here are the symptoms that I have experienced:

  • When first turned on, the fifth (A) and sixth (low E) strings won't trigger MIDI notes on.

  • After a few minutes, the sixth string will trigger notes on, but muting the string won't turn them off. The fifth string still doesn't respond.

  • After a few more minutes, the fifth string will trigger notes on, but muting the string won't turn them off. The sixth string begins to work properly-- that is, notes are triggered on and stop when they're supposed to.

  • Sometimes, but not always, after the instrument has been turned on for a long time, it will trigger notes on and off properly on all strings. In one experiment many months ago, I left the guitar turned on for 24 hours after experiencing the initial symptoms described above. After being powered up for that length of time, it worked perfectly for several hours, until I turned it off.

Scores of people have written to me describing the same or similar symptoms. Often, they report similar triggering and muting troubles on the fourth (D) string.

Capacitor transplants: Many people (probably around a dozen) have written to me stating that deteriorating surface-mount electrolytic capacitors are the source of the problem, and that replacing those capacitors cured their instruments. One of them, Tom, was kind enough to post details on the MIDI Guitar mailing list.

Incidentally, everyone who replaced the capacitors used non-surface mount devices.

The deterioration of electrolytic capacitors exactly explains the symptoms which I and so many others have experienced. The behavior of electrolytic capacitors is peculiar, but fairly well understood: They're chemically active devices which change properties dramatically, not only over time, but dependent upon whether a voltage is present. (Here's a lucid factual explanation of electrolytic capacitor "forming," in the context of a charming bit of fiction writing; here are details on electrolytic capacitor failure modes, and here's an excellent discussion of capacitors in audio roles, and their replacement.)

To make a long story a bit shorter: Electrolytic capacitors deteriorate about seven times more rapidly when there's no voltage applied than when they're in service. (You may have experienced this phenomenon with an old tube guitar amp that you haven't used for a while; "cap jobs" are as much a part of maintaining such an amp as tube replacement.)

This nicely explains why PG-380, PG-310, and PG-300 owners have experienced the same symptoms at different times, even though the instruments are all roughly the same age: The ones that get regular lovin' last longer than the neglected ones like mine.

It also explains why, in some instruments, including mine, notes would fail to trigger or stick on when power was first applied, but would behave properly after a "warmup." It wasn't a thermal issue, as I speculated, but a matter of the capacitors requiring time to form.

Juice therapy: I lacked the money to have the capacitors in my PG-380 replaced. I also lacked sufficient confidence in my eyesight and circuit board soldering skills to attempt the work myself. But the behavior of the instrument after a "warmup" suggested the possibility that leaving it powered up for a long period of time might allow the capacitors to re-form to a functional state. Since I believed there was a high probability that I would have to replace the capacitors anyway, I felt I had nothing to lose by trying it.

Caution: Even your normal nine-volt power supply can cause further damage to already deteriorated capacitors. Before trying this trick yourself, please read to the end of the page.

My music room includes a closet with an electric outlet inside, and that's where I put the PG-380 and plugged in its nine-volt power supply. Out of sight, out of mind: I powered the instrument up on 10 November, 1999; I began testing it on 10 May, 2000. Here is the record I kept of the results:

10 May 2000: tested without disconnecting power; all notes triggered, all notes stopped, no apparent problems driving VZ-8M. Powered off ~5PM

11 May 2000: Powered up ~10:30AM; all notes triggered, all notes stopped, no apparent problems driving VZ-8M. Tested internal synth sounds for first time: Output level very low, but triggered/stopped OK.

12 May 2000: Powered up ~2:30PM. Again, no problems with VZ-8M, but extremely low output from internal synth. With amp at impractically high gain setting, sounds seemed to be OK-- just too quiet. Tested iPD/Mix output with two cables, both good with guitar output, ruling out defective cable. Tested Guitar/Stereo output with known good stereo to dual mono cable, ruling out either output jack as source of problem.

14 May, ~10.00 PM: Despite being without power for ~53 hours, PG-380 triggered and stopped flawlessly with VZ-8M. Left powered up until ~5PM 15 May.

19 May, ~9:20 AM: After ~88 hours without power, triggered and stopped flawlessly with VZ-8M. Internal sounds not checked. Power off ~10:20AM.

24 May, ~1:20 PM: After ~123 hours without power, it just worked. Internal sounds not checked. Power off ~1:55 PM.

03 June, ~11:30 AM: Nearly 10 full days without power; all strings triggered and tracked properly upon powering up. Internal sounds still detectable, still triggering and tracking properly, still unusably faint.

20 June, ~1:30 PM: More than 17 days without power; everything worked properly upon powerup, except that internal sounds were still unusably faint. Power off ~1:55 PM.

21 July, ~7:15 AM: 31 days without power. It worked, just as before.

Conclusions: I was lucky. The application of 9VDC re-formed all of the capacitors involved in the pitch to MIDI conversion process. However, it seems reasonable to assume that one or more capacitors in the output path of the internal tone generator either had deteriorated beyond rescue by this method, or (and I suspect this to be the case) the application of 9VDC caused the capacitor breakdown in this circuit. Determining what happened with certainty is beyond my means right now.

I think it highly likely that it was not necessary to leave power applied for six months! I suspect that one to two weeks would be sufficient-- but I have no way of knowing.

A safer approach: Hindsight makes it easy to see that I should have tackled the problem more conservatively, by beginning with a much lower voltage.

The ideal approach would be to use a continuously variable voltage power supply, and to begin with a very low voltage-- on the order of 0.1 - 0.5 volts. If I were doing it, I would increase the voltage in 0.5V increments every three or four hours; I suspect this would re-form any capacitor capable of being rescued within a matter of days, rather than months. I don't know that, though, and I welcome any corrections.

Caution: I have not tested this procedure; try it at your own risk. If you do attempt this method, be sure to connect the power supply to the instrument with the proper polarity. Even a fraction of a volt with polarity reversed could damage all of the electrolytic capacitors beyond any hope of re-forming them, requiring their replacement. Also, take care not to apply more than nine volts.

Probably the easiest way to insure correct polarity (as well as the easiest way to connect a bench-type variable-voltage power supply to the instrument) would be to connect alligator clips to the battery connectors inside the guitar's body.

In the absence of a variable-voltage power supply, one of the readily available power supplies which produce multiple fixed voltages would be safer than my method of applying nine volts, faith, and hope. Typically, these units produce 1.5VDC to 12VDC in 1.5V increments. Also typically, the output cord of these units terminates in multiple connectors, one of which might fit the Casio power connector. If I were doing it, I would begin at the lowest available voltage, and increase the voltage in the smallest available increments, maintaining each for several hours.

Caution: I have not tested this procedure; try it at your own risk. If you do attempt this method, be sure to connect the power supply to the instrument with the proper polarity. Most such units have a polarity switch; be certain that it is set correctly before applying power to the instrument. Even a fraction of a volt with polarity reversed could damage all of the electrolytic capacitors beyond any hope of re-forming them, requiring their replacement. Also, take care not to apply more than nine volts.

If anyone knows of any reason not to attempt these procedures, please say so, soon! Also, I welcome any feedback from anyone who does attempt them, regardless of results.


Casio MIDI Guitars | Troubador Tech Index

W3C HTML 4.0