Hello everyone, long time no see 
Hope everybody’s doing great, i have a lot of reading up to do.
I was working on a weird bias problem on my friends tube amp, and it turned out to be the glue that holds the electrolytic capacitor that became conductive, let me show you:
Now as you can see, this glue has become brown and hard. It should be yellowy and quite soft and chewy. The glue is right over a relatively large current track on the pcb, shown in blue. Also, the cap and glue are right next to two voltage regulators. One if which is known to become hot, as per the comment in the schematics we received from the manufactorer.
3 of the 4 resistors in a row, right next to the glue belong to the negative bias circuit of the power amp tubes, carrying about -11 to -17 volts probably. The capacitor in the red circle belongs to the POSITIVE 15v supply for the transistor circuits.
Can you believe it? The glue between the positive side of the capacitor and the junction of the bias resistors measured 400kohms against the 220kohms from this junction to the negative rail, bringing the bias up WAY too high and making the poor tube red plate in seconds.
I didn’t find a lot of sources about the glue becoming conductive, a lot of people in forums say it’s capacitor leakage. All the capacitors tested fine (this wasn’t the only spot with the glue that turned conductive).
I hope this will help someone in the future
Afterthought: the design of the pcb is weird, this 10cm square area of pcb is a recipe for disaster. But hadn’t the factory used this glue, it would have never been a problem at all
3 Likes
I have read about this before, but unfortunately hardly any proven remedies.
Ah yes: search for ‘brown pcb glue’.
1 Like
Yes i’ve read 2 or 3 webpages about the same thing, just not a lot. I did find a lot people saying it’s electrolyte leakage, which was clearly not the case.
Anyway i applied isopropyl alcohol with a cotton swab and carefully scratched away the glue and the amp runs fine now.
The solder mask was damaged in a few places, don’t know if it’s due to corrosion or the glue just ripping it off when little pieces broke off, but i applied a bit of solder to the tracks and applied a bit of nail polish to insulate.
The capacitor is now secured with a bit of hot glue, making sure it’s not on top of any tracks or components.
3 Likes
The goop* conducts mostly because it has not set fully. I use car brake cleaner for cap spill, it takes longer but is gentle to the mask and traces.
Never scrape or pick.
*trademarked technical term
1 Like
I can see what you mean by not scraping but the conductive glue is said to be corrosive anyway. I was able to pull off the softened bits and the glue that hadn’t turned brown but took pieces of solder mask with it anyway. While my gentle scratching made only two little scratches on the board, one of which isn’t over the pcb track.
Are you sure brake cleaner is less agressive than isopropyl? Isopropyl is quite an acceptable solvent in electronics.
Brake often contains toluene, xylene and i believe also benzene. Some are even chlorinated. I myself doubt that it’s less agressive and it’s certainly not better for your health. Possibly you use a different kind of brake cleaner
The amplifier was made in september 2016 according to the date code, i figure the goop* should have set by now. There seems to be a pattern where all the goop in the high heat area’s of this amp has turned brown and conductive, the areas near voltage regulators and dropping resistors.
I think the glue might also be (or become) slightly hygroscopic.
The capacitors that i marked red on top all had conductive brown glue, while glue in other areas is still yellow. Note the power resistors and plastic to-220 regulators
By the way i’m not intending to be negative or undermine everything you said, just thought it would be an interesting discussion
