[sdiy] SPV-355 -15V rail issue

brianw brianw at audiobanshee.com
Sat Jun 17 23:51:18 CEST 2023



On Jun 16, 2023, at 10:12 AM, Todd Sines wrote:
> I have an SPV-355 that was behaving adequately fine until recently. Unit will power on, (and LEDs illuminate correctly) if the -15V LM7915 is unplugged, when engaged, it shorts and blows both fuses. 
> 
> Wonder if anyone’s encountered this and if so, what did you do, or where did you look first?

Has anyone responded to help?

#1. I'd be concerned about vintage synth chips that don't like to be powered with only one of the two rails. I can't remember which chips, but some will self-destruct if they only get one of the two rails. I mention this because you might want to avoid feeding +15V only without -15V. I haven't looked at the schematic to see whether the SPV-355 uses SSM, CEM, or other synth chips. I realize that it's a reasonable test, and you've learned something by that technique, but I'm suggesting caution before repeating this test.

#2. You might be able to test the supply in isolation by connecting a fake load. You'd need to know the current draw of a healthy SPV-355, or the maximum current that the supply can handle. You'll also need power resistors to dissipate the heat of a full load. Anyway, if you can set that up, you can confirm that the power supply is healthy when feeding a normal load. I suspect that the supply is good, and the main board is shorted, but testing in isolation can help confirm.

#3. Sounds like the main circuit has a short on the -15V node. You might be able to check for this with a volt meter set to measure resistance - although it can be very difficult to test a circuit this way due to the bypass capacitors storage charge. Many meters will "beep" when detecting a very low resistance, and that allows for quick checking around for shorts. If you do see a short, and there are removable (socketed) chips connected to the -15V node, then you can remove them to isolate the problem. If there's still a short on the -15V node with everything removed, then you might have to start desoldering likely components to determine which is bad. If the short goes away when removing components, then it should be straightforward to determine which component(s) is/are the culprit(s).

I had one piece of gear where a component in the design was supposed to protect against surges, and was connected between power and ground. That part sorta did its protection at some point (lightning strike), and afterwards shorted the supply to ground, so fuses were constantly blowing. I got lucky by examining the circuit to see that there was a component between power and ground, and tested my hypothesis by desoldering that part. Sure enough, it was causing the short. The circuit worked without it, but was unprotected, so I purchased a replacement to restore protection against future lightning strikes. You might not be so lucky finding the shorted component - and in my case it was on the supply, not on the main board - but it's worth a look around.

#4. You can always look for components that have a tinge of brown or black, showing that there was overheating at one point.

Sorry for the very vague suggestions. I'm hoping someone has specific experience who can help you.

Brian Willoughby




More information about the Synth-diy mailing list