High Voltage circuit design failures? attempt to modify ... help sought.

[jaunty]

So - firstly i am posting HERE (not under repair) because i am trying to sort out some problems with an electrostatic panel driver board (for electrostatic loudspeakers) that i have been having some problems with. I do not know for sure that the problems lie specifically with this board but there are definitely things here that require attention.

1. traces under conformal coating appear to be oxidizing or possibly damaged due to HV electrolysis (?)
2. board itself has very poor HV design (look at all the hard right angles on the traces!)
3. a few of the solder joints appear to have had deep pitting (I am assuming due to coronal discharge activity - or possible bad soldering to begin with and joints originally were done as fillets (which as i understand is not great HV soldering technique - one should have 'domed' or 'balled' solder joints contrary to normal low voltage technique to avoid coronal discharge etc) so I have begun redoing joints on the board.

please note - yes the joints look like hell right now otherwise (mostly due to the thick layer of 'corona dope' on the top layer of the board.

Despite the 'not great' pictures (ipad, sorry) i hope the spotting underneath the conformal coating on the traces is visible. It is MUCH more prevalent on the traces leaking up to the 6KV output terminal - which is what makes me wonder if this is due to electrolytic activity (I've seen this kind of thing happening before with 400V tube circuits and leading to the yoke of a CRT).

My thoughts are to remove the right angled sections of the traces, remove conformal coating on all traces and then lay in 20-24 awg copper wire and reflow the new beefed up traces with solder.

what say you? any thoughts? anyone dealt with this kind of thing before?





oddly the picture links do not seem to be working - here are direct URLs

https://photos.app.goo.gl/4U25CHDzFv6ueM876
https://photos.app.goo.gl/wBsM9PyqL3NoxGUi7
https://photos.app.goo.gl/FpC1DEC2AU9qLaNEA

[PartialDischarge]

I would do the following: get rid of all the green soldermask, reball the joints as you mentioned in a domed shape, solder the traces, cut slots with a dremel disc in between caps for example(the more the better) , apply good corona lacquer ( 2 or 3 layers with drying in between)

[jaunty]

hmm .. 'slots'? in the board? what orientation? i'm assuming you mean to prevent any possible conductance or path between caps? i have never heard of this happening on these boards (between caps) though i am probably more worried about discharge between any HV point on the board and nearby grounded metal parts ... i was thinking of potting the whole thing in silicone or RTV which doesn't have a bad dielectric strength. maybe corona dope first though

[PartialDischarge]

Slots are a must. Fibers inside the board will also leak current. By “in between” I mean in the middle, a cut tranversal to the joints. Look for other similar pcbs and see how this is done.

[jaunty]

Slots are a must. Fibers inside the board will also leak current. By “in between” I mean in the middle, a cut tranversal to the joints. Look for other similar pcbs and see how this is done.

like so?
https://photos.app.goo.gl/UcA6FBfTdGivhfKc9

if that's wrong maybe if you could provide me some kind of picture reference i would be grateful

[PartialDischarge]

To be fair that board is designed very badly

[jaunty]

that was actually my first thought when i saw it ... though i got nothing but resistance when mentioning it to the audio (amateur tech) community who involve themselves with these things. This is an aftermarket board. The original uses rounded traces (no corners or hard edges) - i think it's tricky enough using a PCB with razor thin traces where coronal discharge is concerned - no matter how smoothed out the pattern.