Lucien Nunes - Results of IR tests on equipment

[FreeLec]

Hi Lucian. I hope you see this post that I said I was going to put up in response to your information.

After what you said about being surprised about such a low resistance with equipment and because I hadn't given this too much thought, I had only assumed that some equipment had very low IR I have tested some over that last 2 weeks.

After testing various bits of equipment L+N->E I found that a lot were >999MΩ.

Many were lower than that but still in the hundreds.

The same type of equipment can have different result. Fridges 550MΩ, 250MΩ and the lowest 2.5MΩ

Same make of hand dryers next to each other; one at >999MΩ, the other 8.7MΩ

Frogeye emergency light at 11MΩ

The lowest one was an Andrews movable air conditioner at 1.75MΩ. It is a very old one with mechanical controls so no electronics, only 2 fan motors, a compressor and a condensate pump. The pump and condenser motors are shaded pole and the compressor is capacitor start. Not sure about the evaporator fan but i think it's shaded pole as well.
However, this has not been run for a while and it's possible the windings have some damp. When I get a chance I will put it in a warm room for a few days and test again. Then run it for a while and test.

An emergilite fitting which normally has >999MΩ had 0MΩ. However, this one was faulty.

I found a few bits of equipment took longer than you said for charging capacitors although still less than 4 seconds.

RCBOs 0.11Ω

RCDS N-E is 0.33 MΩL-E is 0,17MΩ.

However, an RCBO/RCDS capacitor (a ceramic one) is not on a filter circuit. It functions to provide the electronics with an alternative neutral in case the neutral is lost. I have read that this is not really necessary as the the chance that the neutral is lost at the same time as a fault on the circuit is negligible, hence, some RCBOs don't have an earth tail.
I don't agree with that. If you loose the neutral the circuit will appear dead. Someone can then touch the line conductor when investigating a piece of equipment forgetting to unplug it.

I tested various capacitors directly and found electrolytic are very leaky at DC. However, you did say that they are not used across L-E and N-E so that wouldn't be an issue.

In conclusion, it could only have been faulty equipment or a fault in the circuit under test that could have caused very low IR readings.

However, a circuit that should have a high IR reading but has one that is 1 to 100 just because of equipment doesn't really show the real condition of the circuit itself.
Therefore, as far as I'm concerned, all circuits that has plugable items should really be unplugged and 2 pole switches (as in FCUs) should be switched off.
As far as lighting is concerned there is no choice but to accept a lower reading if this is caused by any equipment. Switching the single pole light switch off won't help as when you connect L+N to test you are still connected to the L through the load.

So you were absolutely correct. Your information was invaluable and had got me thinking more about the circuit and things connected to the circuit.

Thanks.

 

[Lucien Nunes]

Freelec - thanks for taking the time to do the tests - I'm pleased you got some good measurements of a range of items - they sound fairly representative to me. Can't write more now but I am about to do a thing on identifying loads from the resistance L-N prompted by another thread. But I won't be doing all of those at 250V!

 

[Marvo]

Sounds interesting, are you talking about NILM? http://publik.tuwien.ac.at/files/PubDat_200382.pdf

 

[Lucien Nunes]

No, I meant identifying likely possible connected loads when doing an IR test.

 

[HappyHippyDad]

Freelec - thanks for taking the time to do the tests - I'm pleased you got some good measurements of a range of items - they sound fairly representative to me. Can't write more now but I am about to do a thing on identifying loads from the resistance L-N prompted by another thread. But I won't be doing all of those at 250V!

Hi Lucien,

Did you get anywhere with this?

 

[Lucien Nunes]

Oh, sorry, lost the thread, will do it later.

 

[FreeLec]

Lucian

I hope you will catch this thread.

I have been doing some condition reports so have been coming across a lot more items of equipment.

There have been 2 that have had readings like 0.35 Meg from L+N->E.
They tend to be electronic equipment. I don't know exactly what they were as they where in data racks with lots of stuff.
At first I thought that maybe they had internal surge suppression but quickly realized that this could not be the issue as I test at 250V at first if there is a risk of vulnerable equipment.

What do you think? Apart from RCBOs and SRSDs are there some items of equipment that have that low an IR?
Or do you think the items have a fault?

 

[Lucien Nunes]

If there were two giving the same value it suggests something intentional rather than a fault, and probably not VDRs either. I would not be surprised to find it was one of those annoying 'earth present' neon indicators (wired L-E) that sometimes turn up in rack power distribution units. If that was the case then it would read open-circuit at any voltage below the striking voltage of the neon (around 70V), falling to a value in excess of the series resistor (typically 0.25 - 0.5 MΩ) above. A test with a multimeter would therefore likely read infinity.

I should have mentioned these polarity check / earth present indicators as another exception in the earlier thread except that I didn't think of them at the time, I was looking for anything that could give a reading in the tens of kΩ and these will be hundreds. I have seen some with 'OK / Reversed' indicators, in which case L+N->E will light both and give the resistance of the two in parallel. There are some with the 3-light combo used in nasty check-plugs, they will probably behave similarly.

 

[FreeLec]

Thanks Lucian

That didn't occur to me. I haven't seen an earth present indicator in a data rack before, but then I don't look at them that closely.
I looked up PDUs for data racks but didn't find any with one so they must be rare.

I did an IR test on a Fluke test plug with LEDs and a Martindale with neons. The first gave 0.02 M and the second gave 0.06.

Still, another thing to keep in mind when I get low test results (although I do try to unplug everthing first last week I kept coming across locked data cabinets with spurs that were inside. Too much time to get the key.)

Thanks for the info anyway.