How safe is a wooden cabinet?

[boater sam]

Strange day today.

Sat in the workshop. diesel heater keeping me warm, raining outside and its miserable. tidying up the bench.
Shop radio, an old battered Ferguson 289, huge speaker, push-pull by a pair of EL33s, on Radio Stoke as usual. It has drifted off tune a bit possibly due to the heater.
I go to retune and the cabinet is giving me the finger tip tingle, but its wood, varnished, can't be live......can it?

Volt stick goes crazy as soon as it approaches the set, anywhere on the set. Chassis is open, never had a back, its live. So is the whole wooden cabinet and the plastic knobs! Never have I before had a wooden cabinet that is live. Its not wet or even damp but definitely conductive. 

Thinking to stop this I hastily attach an earth wire to the chassis ( there are 2, it has a separate transformer and rectifier chassis )  and change the 2 pin plug for a 13A  with 1A fuse. Plug in, it warms up, no more tingles. Settle back to tidying up.

Oh dear, there is a clunk and I am in the dark!

This has got to be sorted. Out with the power chassis, insulation test mains in to chassis give me 2megs.  Not bad but not perfect. 
Underneath is a 0.01uf 1000v DC  300v AC waxy made by Static, new one on me, mains live to chassis anti modulation hum capacitor. Whip it out and its now a 0.077uf cap with 16% loss, obviously past its best.

With it removed, no more tripping, normal Stoke radio reception restored. Volt stick quiet and not flashing anymore.

But the question remains, why was a wooden cabinet live enough to give me the tingle? I always assumed that a woody was safe to pick up live.
Heck, I'm old enough to remember wooden fuse boxes and open wire ceramic fuse carriers.
Every day a school day......................

Sam.

[ppppenguin]

That cap isn't 0.077uF. That's a consequence of measuring a leaky cap with a tester that uses charge time to do its job. Leaky caps will indicate above their marked value.

No idea why the wood was even slightly conductive unless it's rather damp.

Sam's experience is also a demonstration of why it's better not to have your lights on the same RCD as your bench sockets.

[boater sam]

I  do know why the 0.01 uf cap reads over, its a clue to the fact that its duff. The Magic Chinese box tells all.
Unfortunately the farm is a daisy chain of RCDs, we are never certain which one will trip first !
 The bench 30mA is fed from the workshop 30mA which is fed from the outbuildings 30mA which is fed from the Dairy 3 phase 100mA RCD which ultimately is supplied from the main incoming distribution 500Ma RCD, whew!
Its all TT supplies so the protection is lots of ground spikes around the place. The connected load is unbelievable. We have our own grid transformer.

Its the workshop one that tripped on me hence the plunge into darkness. I hate winter in the UK, should not be here at this time but its a silly year.

[ppppenguin]

Sam, you know why it read over value, but some readers might not, and wonder how a cap can increase in value. The point is that the magic box is telling lies of a sort that not all will understand.

Cascaded RCDs are bad news. There is no discrimination unless there are suitable time delays. I would expect the main incomer RCD to have a delayed response but beyond that who knows. At least you didn't take out the dairy 3 phase RCD. Unless somebody re-plans the protection and does it correctly you're rather stuck with the problem. The only instant fix is to supply the bench from a hefty isolating transformer, followed by an RCD, to create your own local TT supply which can't be "seen" by the upstream RCDs.

[Murphyv310]

I've felt the pulling on the back of the hand with wooden TVs with a reversed mains plug.
No leakage seen on a very high impedance meter but nonetheless you can feel it.

[Crackle]

I was surprised to find that some rubber pond lining was conductive, so much so it even scorched a hole in it.
I had to un-stuff the electrolytic where I had used it to stop the new caps from rattling.
Mike

[ppppenguin]

I was surprised to find that some rubber pond lining was conductive,.....

My first thought is astonishment. My second thought is that perhaps it's carbon loaded and that makes it conductive.

Do you know what type of pond liner it was? Low density polyethylene(LDPE), PVC, Butyl rubber or EPDM?

The conductivity of butyl rubber with carbon loading has been studied: https://www.researchgate.net/publication...Composites

[Crackle]

It was probably EPDM as that seems to be the common type of pond liner. Looking it up it seems it may have a carbon filler, and this rubber does have some shiny flecks in it.

Mike

[boater sam]

There's not a lot of carbon on my radio cabinet!
I tried measuring the insulation value to ground from the varnished surface with a megger, reads nearly infinity! So why does 240v AC get through?
Its not damp as far as I know, the workshop is heated and nothing rusts very quickly.

There is an oddity however. I had a bakelite radio set on fire last year and where the smoke went past the tool rack on the wall, tools had a thin layer of rust on them afterwards, more like tarnish but they never got hot.
This set was on the end of the bench then. Does burning bakelite produce a conductive smoke?

[ppppenguin]

Smoke will probably contain carbon particles, so potentially leave a conductive layer where it settles.

Random thoughts on conductive wooden cabinet. I wonder if somehow AC break down the insulation more than 500V DC from a megger? I was thinking of some kind of non-linear effect. Or is the "conduction" via capacitance somehow? Selectively conductive in some areas, such as where the chassis is touching it?

I don't believe any of what I've just written about the cabinet, just throwing random ideas about in the hope that I might hit the target.