AXIOM 49 with dead midi output

I posted this unit on the blog because it's failure was caused by a most unlikely culprit: a failed ceramic capacitor which was used to filter out noise on the midi output. Although I encounter countless failed aluminum capacitors, I rarely (if ever) have encountered a failed ceramic capacitor.

Nevertheless, this was the case here.

The unit came to me because the midi out was failing, it was completely dead. Of course I checked the jack and the connections around it, and all seemed fine. I reset the software.

Just for a lark, I removed the tiny small value capacitor that filtered the midi out line to ground, and found that the midi out came right back.

Although the capacitor did not read as shorted, it's capacitance must have drifted and increased significantly, enough to actually be filtering out the midi signal, rather than just noise.

It was a 220 PF cap, shown below, and the unit worked fine when it was removed, and continued to work fine when a new one was installed. The second picture shows the area where the cap was located... I believe labelled C4 on the silkscreen.

Roland KR650 repair: bad electrolytic capacitors

This particular model suffers from leaky electrolytic capacitors on the main board. By leaky, I don't mean electrically; I mean there is some nasty liquid leaking out of these things! The result is all kinds of distortion in the outputs, low volume, intermittent problems, etc.

I have repaired a few of these. The repair has always been permanent, however on most instances the liquid that leaks from the capacitors burns through traces on the board. On this one, the liquid burnt out the V- supply to one of the op amps near the DAC chips.

So, the repair consisted of removing all the surface mount electrolytics on the main board, cleaning up all the liquid, replacing the caps, and repairing a trace that supplies the V- to one of the op amps near the DACs.  Other traces may be damaged too, so if it still doesn't work when you replace the caps you will have to check the traces that run near them.

I used standard through hole caps instead of surface mount on this job. I bent the leads and attached them to the pads; there is plenty of room for this, and I applied a bit of hot glue on the leads to hold them in place. If my memory serves, they were all 3.3 and 10uF caps.

The instrument sound great now. I also replaced all the tact buttons. There sure are a lot of buttons on that thing!


Finally, the contacts on this one were good. However, I think this was a transitional model, and the key contacts are no longer available. This may be a consideration in judging the cost effectiveness of repairing this model; if the contacts show signs of failing, it would be worth trying to repair them first (perhaps by recoating them) before investing time and money in fixing the other issues.

The highlighted area is where I found destroyed traces, supplying the V- to one of the SMT op amps near the DACS

.