Insulation Resistance Sensitive Equipment?

[Casterbridge]

Are smoke alarms and extractor fans considered senstive equipment?

 

[AndrewDavie]

No, but they _are_ a load, so you need to isolate them to get a proper reading.

Sensitive equipment usually means surge protectors or neons.
Test at 250V first to make sure you've not got any loads left in before retesting at 500V.

 

[Taylortwocities]

so its OK to squirt 500v up the live and neutral of a device that is made for 230volts is it?

I'm not letting you in my house!

 

[SimonH2]

IMO, for a 240V supply, the equipment must already be capable of withstanding 340V - and that's with no allowance for surges and spikes. 500V is just under 50% on top of that (not the double that saying it's 230V equipment infers) - and I'd suggest that anything not capable of that is marginal to be connected to the supply anyway.

But if testing with stuff connected, why not just test at 250V, or 340 if the tester has the option ? If anything can be damaged by that, then it deserved to be killed and thrown in the skip where it belongs.

Does testing at 500V really demonstrate anything that won't be shown up at 250V ? I know that some faults don't show up at low voltage but we're not talking a few tens of volts, I'd be interested if anyone has ever found a fault that showed up at 500 but not at 250 - apart from the two extension leads I have that were thrown out for failing due to their design (surge protectors) ad the sparkies insistence on testing them at 500V.

EDIT - and that's L&N together to earth. The only devices that should be present across that in normal equipment should be filter capacitors, and perhaps some high value bleed resistors.

 

[JohnW2]

Does testing at 500V really demonstrate anything that won't be shown up at 250V ? I know that some faults don't show up at low voltage but we're not talking a few tens of volts, I'd be interested if anyone has ever found a fault that showed up at 500 but not at 250

I very seriously doubt it - unless something has happened which reduced the insulation to (literally, or the equivalent of) a micron or two's thickness of PVC. Indeed, I actually suspect that voltages considerably below 250V would also not usually fail to pick up things which would be found at 500V[/quote]

Kind Regards, John.

 

[Casterbridge]

OK - Do I assume that standard practice is to leave them out or connect them up? My thinking was if they are not sensitive and connected up, then I would not only confirm IR for the circuit but also the equipment if reading >299Mohms.

 

[JohnW2]

OK - Do I assume that standard practice is to leave them out or connect them up?

If you're talking about L-N tests, then all loads need to be disconnected or else you'll just get silly (low) answers. If you're joining L+N and testing against earth, then it doesn't matter so much, but it's possible that L-E and/or N-E components or leaks within the load could muck things up.

My thinking was if they are not sensitive and connected up, then I would not only confirm IR for the circuit but also the equipment if reading >299Mohms.

What do you mean by 'connected'? If you mean "connected and 'on'", and if you are talking about L-N tests, then, if the load does anything at all, then the measured 'IR' is usually going to be very low, making a nonsense of your IR measurement. However, again, if your talking L+N vs. E, then 'connected' is usually OK (but could occasionally give you a 'false' low reading, due to components/leaks in the load).

Kind Regards, John

 

[SimonH2]

However, again, if your talking L+N vs. E, then 'connected' is usually OK (but could occasionally give you a 'false' low reading, due to components/leaks in the load).

Indeed, and this has come up before, in fact in a thread I started about doing some "quick" tests between tenants.

From observation, in the flat I can do an IR test (L+N to E) at 250V fine with one polarity, but not the other - the RCBOs show a low reading one way but not the other. Thus I can leave everything one, open the main switch, and quickly test the IR - if it's reasonable then I see no reason to go any further, and in particular there's no reason to start disconnecting circuits in the CU (which has the potential for introducing faults which weren't previously present.) However, I can't do it with the MFT as the test isn't long enough to charge the caps in some equipment* which itself gives a false low reading - but Dad's old IR test set works fine as you can hold the button for a few seconds and watch the reading settle.

* Eg some outside PIR lights - which if you disconnect them at the CU means you don't test that part of the installation at all, making a mockery of the testing IMO.

 

[JohnW2]

However, again, if your talking L+N vs. E, then 'connected' is usually OK (but could occasionally give you a 'false' low reading, due to components/leaks in the load).

Indeed, and this has come up before, in fact in a thread I started about doing some "quick" tests between tenants.

Indeed so, and I recall that discussion.

From observation, in the flat I can do an IR test (L+N to E) at 250V fine with one polarity, but not the other - the RCBOs show a low reading one way but not the other. Thus I can leave everything one, open the main switch, and quickly test the IR - if it's reasonable then I see no reason to go any further ...

I'm inclined to agree, particularly in the context of a 'quick test'.

I can't do it with the MFT as the test isn't long enough to charge the caps in some equipment* which itself gives a false low reading - but Dad's old IR test set works fine as you can hold the button for a few seconds and watch the reading settle.

Indeed so - and, at least in the case of my MFT (Fluke 1652) one can't even overcome that by conducting two or more tests in quick succession - in the name of safety, the Fluke has a built-in 'auto-discharge' facility, and it won't let one doing a second test until the circuit has been 'discharged'!

The manual for my Fluke says that IR measurements are 'operable' for capacitable loads up to 5μF, but they are obviously talking about a fairly 'pure' capacitive load (with no resistance in series with the capacitor, hence very rapid charging). The killer' situation is obviously that of a fairly large capacitance in series with a fairly large resistance, which will be slow to charge, hence giving falsely low IR readings. I've just done an experiment with a 0.47μF capacitor. Tested with the Fluke at 250V it gives, as expected, an IR reading of >200MΩ. However, if I put a 1MΩ resistor in series with the capacitor, the IR reading falls to about 6MΩ. If I played with the amount of resistance, I could probably get the reading lower.

Kind Regards, John