Tweet
An Ashdown MAG 600 came in with the sub-octave not working. On this amp its a satellite board mounted on a bracket. Just an in/out, ground and +15v. Plug in and replicate the fault. Fine.
Open up the amp and release the board and rest it on a piece of cardboard. 100hz test signal in the input. The thing starts working as soon as I get my hand within 2" of it. Like a theremin. The same with a test lead or scope probe bought into close proximity. I'm thinking there may be a grounding problem, but there isn't. So the thing is a paradox; how do you trace a fault when anything that comes into close proximity causes it to work? Static component tests on the passive components looked fine.
So the next thing I did was to use a fine piece of wire about 6" long - so thin it didn't generally trigger the proximity effect - and probe the board to see if I could make it work. Strangely the circuit would kick in when the wire barely touched the back of the 4013 IC and replacing this was the fix. So I'm guessing some capacitive effect at work here and that was a lucky break.
This reminds me of my mainframe engineering days. I used to look after a bunch of Prime machines and they were really tricky to work on. We had to use extender boards to get to the components - they were large boards that slotted in like oven shelves. Except the extenders along with the test gear changed the timing pulses and screwed them up so it was all futile. Most places just had a stack of replacement boards, but they were so expensive it was worthwhile stripping the pulled boards down and re-populating every component just to get them working.
Open up the amp and release the board and rest it on a piece of cardboard. 100hz test signal in the input. The thing starts working as soon as I get my hand within 2" of it. Like a theremin. The same with a test lead or scope probe bought into close proximity. I'm thinking there may be a grounding problem, but there isn't. So the thing is a paradox; how do you trace a fault when anything that comes into close proximity causes it to work? Static component tests on the passive components looked fine.
So the next thing I did was to use a fine piece of wire about 6" long - so thin it didn't generally trigger the proximity effect - and probe the board to see if I could make it work. Strangely the circuit would kick in when the wire barely touched the back of the 4013 IC and replacing this was the fix. So I'm guessing some capacitive effect at work here and that was a lucky break.
This reminds me of my mainframe engineering days. I used to look after a bunch of Prime machines and they were really tricky to work on. We had to use extender boards to get to the components - they were large boards that slotted in like oven shelves. Except the extenders along with the test gear changed the timing pulses and screwed them up so it was all futile. Most places just had a stack of replacement boards, but they were so expensive it was worthwhile stripping the pulled boards down and re-populating every component just to get them working.
Comment