Damn Safety Switch. Did its job.

[lograsso]

I have been absent from the eeVBlog forum for some years and then my account disappeared, but have been watching Dave since about episode 30 and am always excited to see what is next to come.  Hence, when I had a dreaded intermittent fault of my household electricity supply that was ultimately resolved I thought it would be great to share my experience on the forum.  I am also hoping those in the know might be able to assist me in understanding exactly what has gone on.
Here’s the low down.  Over the past few years my safety switch (RCD in the main board) would occasionally trip.  Only once every six months or so.  I would look around my property and find nothing unusual.  Switch it back on and, yay, continue on for months with no issues.
Last night, however, things got a little more complicated.  I awoke to a house of no power.  Actually I awoke to the alarm on my UPS telling me there was no power.  I again went to my meter box and discovered a tripped safety switch and no fuses tripped.  At mid-night my main concern was restoring power to two aquarium pumps and fridges so to not endure spoilt food and dead fish.  This time when I switched the safety switch it tripped immediately again.  I decided a good start would be to isolate the household circuits.  This particular RCD (safety switch) fed into five switched fuses.  All the household mains.  Lighting was on a separate circuit and had a separate RCD.  I quickly realised that by keeping fuse three off (the front of my house circuit), I as able to keep the RCD on and continue trouble shooting.
Next step was to go inside where I switched off all appliances on that circuit at the wall and then was able to reinstate fuse three.  I was then able to turn on my aquarium filters/heaters and return to bed knowing in the morning I had trouble shooting to do.  I was betting it was a refrigerator on this line.
On waking I resumed my trial and error, hoping to identify the appliance at fault.  I was eventually able to turn on everything on this circuit with no trip.  I thought, “Great, the dreaded intermittent fault.”  Then I went to turn my central heating on and bang, RCD tripped and no power.  Strange thing was that the heater was on fuse four.  I discovered this quickly as I could not turn the RCD back on without disabling fuse four.  Next step was to go in the roof and disable the central heating at the main source again.  It appeared I had success, for 30 minutes anyway.  Next the RCD tripped again and again could not be restored without disabling the fuse four circuit.  Here I was thinking it was fuse three.  More trial and error of turning things off at the mains only to discover the fault now appeared to be at the back of my house.
In between all this I had contact about nine local electricians with no reply.  Number ten was the lucky shot and finally I had a sparking on the way.  I was total frustrated with the apparent lack of logic that I could apply to the problem.  I knew somewhere I needed to review to total leakage current running through the RCD and this was going to be left to the sparky.  One strange thing I noticed in my testing was when plugging in a power board into another power board that was switched off.  This tripped the RCD.  Initially I was confused thinking that this was off and should not cause a problem.  This was my first clue once a level head returned.
My excellent sparky quickly measured the main power box and found a problem with leakage current.  He seemed surprised when I was relieved and I explained that I was concerned we would be continuing my trial and error and waiting for this dreaded intermittent fault.  He had me go throughout the house unplugging everything.  I suddenly realised where my testing was at fault.  I had not been disconnecting the neutral or earth. 
It was not until every plug was out that I got a message that the fault had cleared.  The last of these was a pond water pump.  The next task was to slowly plug in each device and await a good and bad message.  Good and I left it in.  Bad and I left it out.  This is where I am hoping the forum community to assist my understanding.  Eventually we discovered that every surge protected power board was contributing to the leakage current as well as my pond pump (which I had already suspected).  Everything else was fine.  The explanation provided seem to make sense.  A blasted power surge, pun intended.  It seemed to have taken out every surge protected power board in my home, but nicely protected all the attached appliances.  And I decided the pond pump would need review later.
Next I observed resistance measurements of all the suspect boards across neutral and ground.  Was this what is called a neutral-ground bond??  Essentially, most of my power boards had anywhere from 20Meg to 1Meg across neutral to ground and were deemed damaged.  I confirmed these by measuring them myself later.  Only three were passed and they were open circuit across neutral to earth.  My spark advised me two solutions.  One was to replace all the suspect power boards damaged by presumed surge and, two, to test with plugs out from now on.
I think what was happening was a cumulative effect which had not entered into my logic.  Different combinations of connections were giving random total residual current readings (leakage) and occasionally this would reach a threshold and trip.
I have since replaced all my suspect power boards and have had no further trips today.  When I measured the NG resistance I was shocked to find this one board with 1MOhm.  Seems high, but not when across the mains.  Even 1MOhm from live to neutral.
Please forum patrons.  Does anyone have any other thoughts on this?  Is my basic understanding reasonable?  My knowledge of electronics is reasonable, but I have always struggle with common mode issues.  If I am right about this then does this mean that when I next have a power surge my surge suppressors might work, then fail in this way eventually accumulating leakage again?  My now surge suppressor boards have an apparent light on them that is to go out when they are no longer functional.  Isn’t this when a total fail occurs and before that MOVs may slowly degrade.
Has anyone else experienced this and is measuring the neutral to ground resistance a reliable way a determining a failed power board?  Any suggested reading would be most appreciated.  Sorry for the long post.  Perhaps you found it intriguing.
Michael.

[chickenHeadKnob]

The general symptoms you describe are exactly the way MOVs fail under service. Your surge suppressors were likely hit by some kind of damaging transient. Not unusual at all.

See:https://en.wikipedia.org/wiki/Varistor
and:http://www.littelfuse.com/~/media/electronics_technical/application_notes/varistors/littelfuse_transient_suppression_devices_and_principles_application_note.pdf

[helius]

In the US we call them "power strips", not power boards, but I trust they are electrically similar. The surge protection elements in them are MOVs, metal-oxide varistors. They normally present a very high resistance (megaohms), but when the voltage across them exceeds a specified level, they begin to conduct and shunt the surge energy. Typically they are used in threes, one across live and neutral, one from live to earth, and one from neutral to earth. One big surge event or many small spikes will damage the MOV so its resistance drops below nominal: it gets leaky. If either the live-to-ground or neutral-to-ground MOV is leaking current, it will contribute to the imbalance seen at the RCD.

edit: the MOVs can also present a fire hazard, as when they are leaky they can heat up to ignition temperatures. Several brands issued recalls because their power strips were starting fires. The correct mitigation is for the MOVs to be bonded to thermal fuse elements so that the power is disconnected if they get too hot.

[lograsso]

Thank you for the feedback.  So my suspicions were correct.  I guess one power strip/board failing might not be such an issue, but having about twelve of them all feeding back to one RCD was in my case problematic.  My understanding of my new strips/boards with the indicator light was that it would go out only when the thermal fuse failed.  I guess my practice might be to monitor MOV resistance over time, or just update/replace my power strips/boards every so often.  I am also considering getting an "in-power-box" surge suppressor.  The sparky suggested then just one would be at fault in a surge/transient rather than multiple failures.

Thanks again,

[helius]

Central surge suppression MOVs are a good idea, but they cost more because they must be sized to absorb a lot more energy. The good news is that large MOVs can survive many more transient spikes before they become leaky.

[MosherIV]

Hi, yes, to add to what everyone else is saying......check out this



Big clive does some really good videos. I think there is another one about surge protected strips where he explains how the green neon indicator work (or not).

[Rerouter]

Normally a sparky would test at voltage, e.g. 250V DC on a "megger" to not turn on surge arrestors for appliances, but 500V DC for the wiring.

These megger units are effectively a charge pump, they generally only output between 0.5-5mA so that you cant easily zap yourself with one. but the general unit will let you test 250V, 500V and 1000V, if your in an area with 120V mains, they sometimes have a 125V option.

If your power boards ever measure 100K on a megger, what I would look for is a neon or led status light, the screwy manufacturers generally fit them from live to ground, had to replace about 7 of them at my house, replaced with resettable circuit breaker ones,

for exactly 1M across any of the wires, I would suspect EMI shenanigans, If they built in a basic X and Y cap emi filter, they may have included the optional resistor across live an neutral.

[tautech]

Central surge suppression MOVs are a good idea, but they cost more because they must be sized to absorb a lot more energy. The good news is that large MOVs can survive many more transient spikes before they become leaky.

That's exactly what I did a while back, stuck a big SOB at my incoming mains. Piece of mind for all the electronic componentry in a house these days. It's got a flag in the front panel to alert you when it's suffered too many hits.
It wasn't too expensive IIRC ~NZD60 trade so certainly cheaper that a bunch of antisurge power strips or a UPS that some use for the same purpose.
Location and your local powerco's supply quality will have a bearing on need as in an urban location the power is usually more reliable and more often underground so lighting storms will offer less risk than a rural location.

[helius]

Cool, I've been looking into doing the same. It's important that the central MOV device triggers at a lower threshold than the ones in various power strips (and in computer power supplies too) so it protects them. In bigclive's video, the small MOVs trigger at 300V RMS, so a central device should be rated at 275V RMS or thereabouts. The big MOV is more robust and easier to change when it does fail, so it should take the majority of the spikes and leave the small appliance MOVs intact in case of catastrophic strikes.

[tautech]

Cool, I've been looking into doing the same. It's important that the central MOV device triggers at a lower threshold than the ones in various power strips (and in computer power supplies too) so it protects them. In bigclive's video, the small MOVs trigger at 300V RMS, so a central device should be rated at 275V RMS or thereabouts. The big MOV is more robust and easier to change when it does fail, so it should take the majority of the spikes and leave the small appliance MOVs intact in case of catastrophic strikes.

This is the fella I put in:
http://www.kvc.com.my/StorageAttachment/Kvcsb/datasheet/1051/obo-V20-C-0-280.pdf
A single 280V 20KA unit.
This replaceable cartridge fits into a bigger enclosure ~5" tall, a couple of inches wide and prob 3" deep.
My board is all face mount HW (old style) and I had plenty or room to fit it on the board, you may not.
Be sure to check you have enough room to mount it.