View Single Post
Old 28th Nov 2019, 9:52 am   #2
GrimJosef
Dekatron
 
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,314
Default Re: Repair nightmare's.

Intermittent faults can drive you crazy.

I had a pre-amp in once which would make loud clicking and popping noises (a very bad thing if it's followed by a huge solid-state power amp driving expensive speakers) starting about 15 seconds after switch-on and ending about 30 seconds later. If you switched it off immediately you could quite often get it to do it again. But after that you had to wait hours for the fault to return. I then discovered that if I took it from the workshop into my living room I might have to wait 2-3 days for the fault to re-appear. The circuit was hybrid with a totem-pole triode arrangement and 3 large-ish high-voltage MOSFETs, one of which was in the power supply regulator and the other two of which handled audio signals. The whole thing was built on a pcb.

It turned out that the water-based solution used to rinse the pcb after assembly had somehow been contaminated. When this initially dried out the contamination concentrated itself into a few tiny nearly-colourless dots on the pcb surface. You needed a bright light and a magnifier and, most importantly, the inclination to look for them before you even realised they were there. In this particular pre-amp one of these dots had happened to form bridging the copper pcb islands into which one of the MOSFETs' drain and gate terminals were soldered. There was about 200V between these islands.

When the contamination dot was dry all was well. But it was slightly hygroscopic, and if it was left in a cool environment for long enough it absorbed enough moisture to conduct in a noisy irregular sort of way. This was what was causing the noises. That MOSFET ran hot enought that it could dry the dot out in less than a minute which is why the noises would go away. My living room is warmer and less humid than my workshop which is why it took longer there for the dot to re-absorb enough moisture to start conducting again.

The cure was to take the tip of a cocktail stick and pop the dot off the pcb, which took a few seconds. But by this stage I had had the amp in the workshop for about 6 weeks and I had changed practically every component including, remarkably, the MOSFET with the problem !

Cheers,

GJ
__________________
http://www.ampregen.com
GrimJosef is offline