Odd problem on IR Test.

[Raptor0014]

Doing some testing today on a Wet Room installation.

Fed from its own small CU. And consists of

2 x seperate 50A (for 10.5kw shower) circuits
1 x 20a radial for underfloor heating. This runs direct from CU to a fused spur and then from the fused spur to the UFH controller.
1 x 20a radial for 2 x sockets in a storage cupboard and a fused spur from this for the heated towel rail.
1 x 6a light circuit.

Everything fine until I got to the L-N IR test. It kept giving me a reading fluctuating between 0.08 & 0.37 MOhms. L-E and N-E just fine.

Nothing is going bang as it would with a L-N dead short. So it’s not that.

The 2 fused spurs have neons as do the 2 pull cords for the showers All of these I ensured were switched off as I know neons can cause an issue.

However this didn’t solve the issue on the global IR and I’m scratching my head.

The board is a Fusebox RCBO board. I tried doing the test on the supply side of each RCBO to see if that made a difference. But it didn’t.

I thought it may be the UFH controller so I disconnected that completely yet still got the same problem.

I have traced every single cable as it is still at the first fix stage. Sockets etc have just been added as the plasterer and tilers etc aren’t in for another month. Nothing is damaged. All the connections are sound and terminated correctly in the sockets and spurs, the shower cables have a wago on each conductor and are taped up. Towel rail isn’t yet connected but the supply cable is terminated in the fused spur.

What I can’t work out is why every single circuit is being affected by whatever is causing this fault.

Anyone ever experienced this? I could understand one circuit having the issue, but not every circuit.

 

[James]

I would disconnect all cables from the rcbo and test again.

 

[Raptor0014]

That’s the only thing I can think to do when I go back tomorrow. I’ve never had an issue before with Fusebox RCBOs and can’t see why it would cause an issue now.

 

[Ursanelec]

Might be the SPD giving you a false reading, if there is an SPD fitted

 

[Raptor0014]

Might be the SPD giving you a false reading, if there is an SPD fitted

There is an SPD but I had the SPD MCB switched off.

 

[GBDamo]

Are you doing a global IR? Or have the board fully terminated and relying on turning off each circuit to test?

For IV I'd always be individually IR testing circuits, before I teminate them into a board.

I too can't work out how this is happening unless you are.

And if you are think things like lighting drivers, fans etc.

Might be a faulty RCBO.

 

[Raptor0014]

Are you doing a global IR? Or have the board fully terminated and relying on turning off each circuit to test?

For IV I'd always be individually IR testing circuits, before I teminate them into a board.

I too can't work out how this is happening unless you are.

And if you are think things like lighting drivers, fans etc.

Might be a faulty RCBO.

Yes I’m doing a global IR.

Main switch off, SPD MCB off, all 5 RCBO on and all accessories on apart from the 4 switches with neons. There are no lights or fans in the installation. There are 2 standard MK sockets that are terminated correctly. The fused spurs are terminated correctly and the shower switches are terminated correctly. Only other thing is the UFH controller and I disconnected it.

Tested L-E, N-E & L-N from the isolated side of main switch.

I’ve never had any issue doing it this way before. Just seems odd with this board.

I then tried each individual RCBO from the load side and got the same result.

I’m back there today so will take out each circuit and test at the conductor. I had to leave yesterday to get the kids from school otherwise I would have carried on.

 

[GBDamo]

Yes I’m doing a global IR.

Main switch off, SPD MCB off, all 5 RCBO on and all accessories on apart from the 4 switches with neons. There are no lights or fans in the installation. There are 2 standard MK sockets that are terminated correctly. The fused spurs are terminated correctly and the shower switches are terminated correctly. Only other thing is the UFH controller and I disconnected it.

Tested L-E, N-E & L-N from the isolated side of main switch.

I’ve never had any issue doing it this way before. Just seems odd with this board.

I then tried each individual RCBO from the load side and got the same result.

I’m back there today so will take out each circuit and test at the conductor. I had to leave yesterday to get the kids from school otherwise I would have carried on.

Just a quick check with the board isolated is to look at the RCBOs.

With them off check incoming Live terminal to outgoing Live terminals for continuity and outgoing L-N. Shouldn't be any.

I'm thinking there may be a rouge RCBO that's causing all your issues.

Either that or there's something you've missed and disconnecting each circuit from its RCBO and IR testing will narrow it down.

 

[R-fur]

Are the neons on the load side of the switch? I have seen spurs wired with the in and out reversed.

 

[loz2754]

SPD neutral still connected?

 

[buzzlightyear]

Test the Tails only feeding out. On each circuit. Not test from the rcbos, if all clear then you know if rouge rcbo.

 

[Raptor0014]

SPD neutral still connected?

Yes it’s still connected. Don’t normally need to disconnect it, just turn off the MCB.

 

[loz2754]

Yes it’s still connected. Don’t normally need to disconnect it, just turn off the MCB.

Yes but if it is a faulty SPD, disconnecting line and neutral will eliminate the SPD as the source of the problem.

 

[timhoward]

Out of interest is this an EICR or IV? I'm guessing EICR as you have loads connected.

Tested L-E, N-E & L-N from the isolated side of main switch.

The board is a Fusebox RCBO board. I tried doing the test on the supply side of each RCBO to see if that made a difference. But it didn’t.

Were they the compact double pole RBCO's or the taller single pole RCBOs? In other words were you repeatedly sending 500v between every circuit's N and the CPC bar for each circuit you tested?

I'm also considering that the N-E test would have had been sending 500v to the N and PE terminals of the SPD I think, as N wasn't isolated? I currently can't see what harm that would do though.

My working theory is that they are single pole RCBOs, and the N-E test has damaged something connected in an unusual way.

My general fault finding rule is that if all else fails, start looking at what does work. It can help focus the mind! So in this case everything out of RCBO's, leave the CPC's in for now, and see what happens. If still no logic disconnect CPC's. If still no logic disconnect bonding.

For some reason I'm half expecting everything to test clear until you put it back in...!
All the best with it, I'm sure you'll crack it tomorrow.

 

[Raptor0014]

Been back today. Turns out to be 2 of the RCBOs.

With all circuits disconnected from RCBO they all get the expected >999 Mohm across L-N

The 2 x 20A RCBO however give a reading of 0.08Mohm, even with the conductors removed. The 2x 50A & 1x 6A don’t.

So I’m going to speak to Fusebox on Monday and ask why. I could see 1 being a faulty one. But 2 and from the same supplier so likely the same batch. It’s odd that they would give that reading.

 

[James]

Been back today. Turns out to be 2 of the RCBOs.

With all circuits disconnected from RCBO they all get the expected >999 Mohm across L-N

The 2 x 20A RCBO however give a reading of 0.08Mohm, even with the conductors removed. The 2x 50A & 1x 6A don’t.

So I’m going to speak to Fusebox on Monday and ask why. I could see 1 being a faulty one. But 2 and from the same supplier so likely the same batch. It’s odd that they would give that reading.

have you checked the instructions do not forbid insulation testing?

 

[GBDamo]

have you checked the instructions do not forbid insulation testing?

Having, scrat hed your head for a few days and fried a few RCBOs it just about the right time to RTFM.

 

[mainline]

Doing some testing today on a Wet Room installation.

Fed from its own small CU. And consists of

2 x seperate 50A (for 10.5kw shower) circuits
1 x 20a radial for underfloor heating. This runs direct from CU to a fused spur and then from the fused spur to the UFH controller.
1 x 20a radial for 2 x sockets in a storage cupboard and a fused spur from this for the heated towel rail.
1 x 6a light circuit.

Everything fine until I got to the L-N IR test. It kept giving me a reading fluctuating between 0.08 & 0.37 MOhms. L-E and N-E just fine.

Nothing is going bang as it would with a L-N dead short. So it’s not that.

The 2 fused spurs have neons as do the 2 pull cords for the showers All of these I ensured were switched off as I know neons can cause an issue.

However this didn’t solve the issue on the global IR and I’m scratching my head.

The board is a Fusebox RCBO board. I tried doing the test on the supply side of each RCBO to see if that made a difference. But it didn’t.

I thought it may be the UFH controller so I disconnected that completely yet still got the same problem.

I have traced every single cable as it is still at the first fix stage. Sockets etc have just been added as the plasterer and tilers etc aren’t in for another month. Nothing is damaged. All the connections are sound and terminated correctly in the sockets and spurs, the shower cables have a wago on each conductor and are taped up. Towel rail isn’t yet connected but the supply cable is terminated in the fused spur.

What I can’t work out is why every single circuit is being affected by whatever is causing this fault.

Anyone ever experienced this? I could understand one circuit having the issue, but not every circuit.

My guess would be that L-N IR dc test damaged the rcbos internal circuitry

 

[Raptor0014]

I guess I’ll take this on the chin then.

It does indeed say to remove the connections from the RCBO before IR testing.

 

[timhoward]

If it makes you feel any better I managed to kill a freezer once doing a global IR test. For EICRs and board changes I stick to L+N to E!
I was fitting Hager ones yesterday and they have a little sticker on them saying you can IR test at the outgoing terminals....
So thanks for reminding us to check the instructions.

 

[Lucien Nunes]

I always think that a requirement to disconnect something for an IR test is a sign of cheap or poor design. Clearly it is possible to make devices that don't need to be disconnected and unnecessary fiddling around with cables leads to wear and tear, errors and the possibility of introducing faults that the test was supposed to detect.

There are two aspects - whether the device will spoil the test result and whether it will be damaged. Anything that is connected to the mains can expect to see an RMS voltage of 253V and hence a peak voltage of 358V continuously for decades. Film caps ought to withstand 500V or else they are not fit for a noisy environment that can see spikes of a few kV. VDRs will give spurious test results but won't be readily damaged by the current-limited output of the IR tester. The weakest links would likely be 400V electrolytics and transistors in SMPSUs, but I don't like electrolytic capacitors hard-wired into my electrical installation e.g. in sockets with USB ports.

Back in the days of big, tough stud-mounted thyristors in theatre lighting dimmers, there would often be notices on the dimmer modules warning not to carry out insulation tests. Even if the thyristors were rated for 800PIV and the suppression capacitors 500V DC, and the primary of the mains transformer would thwart any attempt to test L-N with a resistance of only a few hundred ohms, the warnings would be there.

 

[freddo]

I was fitting Hager ones yesterday and they have a little sticker on them saying you can IR test at the outgoing terminals....

I thought that was great when I first saw those stickers, until I realised you still have to disconnect the neutral as it is still connected through when the RCBO is off. Apparently Wylex now do compact 1 module RCBOs with 2 pole switching where you can test without disconnecting.

 

[Raptor0014]

I thought that was great when I first saw those stickers, until I realised you still have to disconnect the neutral as it is still connected through when the RCBO is off. Apparently Wylex now do compact 1 module RCBOs with 2 pole switching where you can test without disconnecting.

Fusebox do 1 module mini RCBO which are DP.

I normally fit those which is why I think I’ve not normally had an issue. However on this job they were SP.

Just need to remember to read the instructions in future.

 

[mainline]

I always think that a requirement to disconnect something for an IR test is a sign of cheap or poor design. Clearly it is possible to make devices that don't need to be disconnected and unnecessary fiddling around with cables leads to wear and tear, errors and the possibility of introducing faults that the test was supposed to detect.

There are two aspects - whether the device will spoil the test result and whether it will be damaged. Anything that is connected to the mains can expect to see an RMS voltage of 253V and hence a peak voltage of 358V continuously for decades. Film caps ought to withstand 500V or else they are not fit for a noisy environment that can see spikes of a few kV. VDRs will give spurious test results but won't be readily damaged by the current-limited output of the IR tester. The weakest links would likely be 400V electrolytics and transistors in SMPSUs, but I don't like electrolytic capacitors hard-wired into my electrical installation e.g. in sockets with USB ports.

Back in the days of big, tough stud-mounted thyristors in theatre lighting dimmers, there would often be notices on the dimmer modules warning not to carry out insulation tests. Even if the thyristors were rated for 800PIV and the suppression capacitors 500V DC, and the primary of the mains transformer would thwart any attempt to test L-N with a resistance of only a few hundred ohms, the warnings would be there.

It's all down to some little china man selling them the pcb's fully loaded for a £2.00 a go
Oh and we do cases btw for £2.00 with your logo