Zs Testing Tt

[Piratepete]

Hi Guys
Job - semi detached with a TT earthing arrangement.

I used the Z loop function on my MFT to check the earth at the DB. Reading 164 ohms - ok.
Same reading with earth reconnected.

However if I do the same test with the Zs RCD function, I get 204 ohms. Why? Do I have a MFT fault?
I need to use this function when testing individual circuits because the main switch is a 30mA RCD.

So far I've only had time to check the calibration sheets (all fine) and to do the same measurements on a TNC-S circuit, where both readings were the same (0.23 ohms).

Do I need to check the tester with a 200 ohm resistor or is there an obvious reason for the discrepency?

Cheers

Pete

 

[Richard Burns]

There can be a discrepancy with measuring on the low current test because any noise on the installation will affect the result.
If the noise is only a 10mA variation then when using high current (25A) testing the difference is negligible but when testing at 15mA the variations can be large.
In general the tester attempts to cancel out the effects of noise causing variation but it can still happen.
One time I was testing and getting variable results then the builder stopped using his SDS on the circuit I was testing and all was fine!

 

[Leesparkykent]

Carry out the alternative Zs test.

 

[Wilko]

A check at 200 ohms and/or another TT would be good. I am so stubborn I go back 'next day' and repeat the measurements in case I've missed something
I looked for explanations in my MFT manual but didn't find any for this 25% delta. I'll include the info in case it helps ... Manual says the test current will be restricted to 10mA in RCD mode (vs 2A in normal). And that if measuring near the transformer the phase angle of supply is a source of error (but looking at table it's not going to give you even half this error, and at 200 ohms you aren't next to tx anyway). General accuracy claimed to be +/- 3%, so no answers here - sorry.

 

[Gnasher]

Not sure what model MFT you are using but what settings do you set the left hand selector on for each test? On the 1730 the options are.

1/. L-PE
OR
2/. L-L. L-N.

And are you using a 2 or 3 lead test?

 

[Sparkingmad]

Did you try and use the same test for all tests even on the incoming supply cables to the board? I.e. If using low current test on out going circuits use same settings on incoming? What results did you get?

 

[i=p/u]

Lower current (3 lead) always seems 20-30% higher,

 

[Deleted member 9648]

As already stated the low current no trip test is notoriously inaccurate. I've brought this up both with megger and our NICEIC inspector. The stock answer is that if there is 30ma RCD protection then a 'ballpark' figure is good enough, as RCd's will operate at a far higher Zs than would be acceptable in practice. Just accept it and carry on.

 

[westward10]

As already stated the low current no trip test is notoriously inaccurate. I've brought this up both with megger and our NICEIC inspector. The stock answer is that if there is 30ma RCD protection then a 'ballpark' figure is good enough, as RCd's will operate at a far higher Zs than would be acceptable in practice. Just accept it and carry on.

What did they deem "ballpark", a figure you feel is acceptable and ignore the random excessive ones.

 

[Toneyz]

As I've put on here before I sent 4 testers away to be calabrated all came back passed well within spec/tol. allowed put on our test socket there was a great difference in loop readings thats why mostly these days I do Ze + (r1 + r2) As well as a low current no trip inacurate readings if we all got our testers calabrated today and met up tommorrow plugged our testers into a socket we would get a range of readings

 

[Deleted member 9648]

What did they deem "ballpark", a figure you feel is acceptable and ignore the random excessive ones.

It comes down to experience....you and I know what reading to expect every time we do a test. Sometimes I'll do several low current tests on the same point and take an average.
On a TT lets face it.....if the true reading is say around 100 ohms it hardly matters if the tested reading is wildly inaccurate, even by as much as 50%!
Where we rely on an overcurrent device of course an accurate reading is vital so it's just as well we don't need the low current test.