Loop tester different results

[marcuswareham]

Hello,

I have a dedicated loop tester (Martindale EZ2500) and also a Fluke 1662 MFT

The reason I have the additional loop tester is often at work I will need to measure P-E & P-N loop impedance only while I am in a harness climbing, or have walked for ages (large concert touring) the EZ2500 clips to my belt, the fluke is a bit big.

I noticed last week that testing the same socket at home with both testers I got different results.
The EZ2500 was much in need of a new calibration, and the fluke was just coming up to needing annual calibration so I sent them both off

Got them back today, both passed within tolerances.
Did the same test to the same socket and the results are more or less the same:

EZ2500:
P-E: before: 0.22ohm after: 0.24ohm (T-safe)
P-N: before: 0.58ohm after: 0.75ohm
Fluke:
P-E: before: 0.42ohm after: 0.33ohm (Non-trip)
P-N: before: 0.44ohm after: 0.33ohm

These results are different enough to make a big difference to my work BS7909 our testing is a little different from domestic testing, and getting PSCC & PEFC from loop testing is important to us, as we have very long cables. We dont do r1+r2 tests and most of the tests in BS7909 are done live (loop testes, RCD tests)

So why are these results so different and what to do about it, is loop testing a bit vague (or at least non-trip)?

Any ideas, Thanks Marcus

 

[Vortigern]

Have you looked in your manual for each of these readings? Mine says that the figure I read can be + or - 6 up or down. So loop reading is very variable. For instance your figure of 0.22 in that case could be 0.28 or 0.16. As we are talking about variables representing fractions of an ohm for at least BS7671 it is not of immense significance unless exceeding table figures. And with ATT on, then figures are more variable due to the lower current. Add to this we can change the read figure by switching on and off the socket a few times and putting the plug in and out a few times thereby lowering the reading significantly, yes a very vague thing having meaning only within the required tables and outcomes. Of course another factor is how you are putting the crocs on to the circuit. i.e. at the bottom of the jaw where the leverage of the bite is highest and pressing the jaw before reading ensuring the best contact or up a (something) with a harness, dangling around getting an even more variable reading as you are swinging about and can't quite do the above?

 

[Lucien Nunes]

is loop testing a bit vague (or at least non-trip)?

Yes. It's the least precise and least repeatable test of any normally carried out in electrical work. It is especially sensitive to electrical noise, because the tester is trying to detect small changes in voltage caused by the injected test current, while the circuit is live with a much higher voltage and possibly passing much higher currents. It is quite common to have to make inferences about which readings are in error and have to be discarded / repeated, which can be averaged, which are valid as-is.

That said, looking at your example readings, where you show the different test modes L-E and L-N, are you nonetheless measuring from the same terminals of the socket such that they ought to give the same reading? If not, i.e. the L-E test really uses the E and the L-N test really uses the N, the data is rather sparse because the high-current and no-trip test of each tester can't be correlated if the impedances are different. If both test modes are being applied to the same terminals, then I wonder whether the difference before and after cal for the Fluke is based on different test leads, because the consistency between high-current and no-trip is very good. The Martindale results are a bit scattered and random even for loop impedance. I would check the test lead connectors very carefully; as mentioned, contact integrity is paramount in getting a stable and repeatable reading.

 

[pc1966]

As a general point these days "calibration" is usually not an adjustment process of days of old, but a verification that all is within specification. Though in some cases they do update correction factors, etc.

The difference between meters is likely the "RCD uplift" effect, see if you can compare them on a non-RCD circuit and find the difference there.

Another factor might be the test probes, as already mentioned you are trying to measure very low supply impedances and MFT do it the cheaper way of incorporating a correction for the manufacturer's supplied test leads (similar to a zero-set for low R measurements). Only the fancy high-cost high current testers have 4-wire Kelvin style leads to bypass the cable resistance when making the test.

 

[static zap]

@Lucien Nunes ..
I too was thinking differing noise response ,
(try other times of day ?)
(as well as others "Inherent accuracy" --)
I have an unswitched socket !
.. Beginning to feel my self turning into JW..

 

[marcuswareham]

Have you looked in your manual for each of these readings? Mine says that the figure I read can be + or - 6 up or down. So loop reading is very variable. For instance your figure of 0.22 in that case could be 0.28 or 0.16. As we are talking about variables representing fractions of an ohm for at least BS7671 it is not of immense significance unless exceeding table figures. And with ATT on, then figures are more variable due to the lower current. Add to this we can change the read figure by switching on and off the socket a few times and putting the plug in and out a few times thereby lowering the reading significantly, yes a very vague thing having meaning only within the required tables and outcomes. Of course another factor is how you are putting the crocs on to the circuit. i.e. at the bottom of the jaw where the leverage of the bite is highest and pressing the jaw before reading ensuring the best contact or up a (something) with a harness, dangling around getting an even more variable reading as you are swinging about and can't quite do the above?

In this case, instead of using probes I was using the 13A test lead that comes with MFT. and same when working at hight I although mostly 16A or 32A cee form connector test leads I have made up, so you are reliant on the contact of the connector instead of clips/probes which should be good.

Although it is a fraction of an ohm it can make a big difference, at the end of a 300m cable on a type D MCB
[automerge]1590019055[/automerge]

@Lucien Nunes ..
I too was thinking differing noise response ,
(try other times of day ?)
(as well as others "Inherent accuracy" --)
I have an unswitched socket !
.. Beginning to feel my self turning into JW..

maybe noise, although if I turn all the loads of in my house and try the test, I still get different results between the testers, I guess also depends on noise from other houses on the same transformer, with loads of switch-mode power supplies putting harmonics on the PEN conductor

even if noise is the cause it doesn't help much as in the real world, when I can go back to work (live events, think it will be a while) there could be alot of noise at the time of testing etc

 

[static zap]

I wonder which makes the most ,X box One or PC gamers rig ..
.. Plenty gaming late night for USA excitement ..
Not much resistive heating -excluding kettles !