HV Safety...

What people are working on at the moment
User avatar
jack
Thermionic Monk Status
Posts: 5646
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#1 HV Safety...

Unread post by jack »

Here's a little cautionary tale...

#2 son picked up some nice modular power strips from work - a 12 way and an 8 way - aluminium body with with steel end-caps. Nice units.

4 of the sockets were broken, so the 8-way was cut down on the chop saw to a 4 way and the spare good modules were used to replace the broken units in the 12-way - he wants to fix that under his workdesk (made from 25mm wooden strip kitchen worktop on a motorised rise-n-fall mount so he can work standing or sitting).

HOWEVER... occasionally the ring RCD would trip when using the 12-way...

He quite sensibly asked me to PAT or equivalent test them as their provenance was essentially unknown. Both checked out OK with a good quality (Keysight) multimeter - infinite resistance between the LNE circuits and combinations thereof - looked great (no damage to insulation, fixings tights etc.)... or maybe not...

I dragged out my insulation tester which goes up to 4Gohm @ 1000V. Later versions of this go to 5 Gohm

The good, 4-way, strip on 1,000V checked at roughly 3 Gohm. That's fine...

The 12-way strip which had tested fine with a multimeter, tested at 177kohm at 250V & 173kohm at 1,000V, i.e. about 17,000 times lower resistance than it should be. It maybe only about 1.5mA now/345mW, but something's not right...

Dismantling the strip, I isolated the issue to the non-live side of the mains switch. Further investigation showed that there'd been a bit of arcing in there and over time the inside of the switch had carbonised a bit, resulting in the low readings at HV. This will only get worse over time, causing other hazards... Not entirely sure why none of this showed up during LV testing with the DMM. My suspicion is that previously someone had a fan heater plugged into the strip and was turning it on and off from the strip, not at the fan. There's no snubbing in the strip either, so that wouldn't have helped at all.

Solution is to replace the switch, re-assemble and re-test at HV. Repeat until good. Possibly add snubbing... but tell son to avoid inductive loads anyway. Cleaning the inside with IPA got the earth leakage right down - bit of carbon, bit of dust...

The lesson here is not to trust even very good and expensive DMMs when it comes to testing the live/HV side of kit. A decent HV insulation/leakage tester is about GBP 60 or you can just make one with an HV supply, a current-limiting resistor and your DMM. There's a reason why PAT testing goes up to 500V or so...

PXL_20230804_091520490.jpg
PXL_20230804_091520490.jpg (213.12 KiB) Viewed 21929 times
PXL_20230804_091442089.jpg
PXL_20230804_091442089.jpg (161.8 KiB) Viewed 21930 times
PXL_20230804_091455664.jpg
PXL_20230804_091455664.jpg (167.44 KiB) Viewed 21930 times
Vivitur ingenio, caetera mortis erunt
Post Reply