Testing questions (works discussion)

[gazdkw82]

Having a conversation at work about testing and a few things have come up that we all have different views on. Bare in mind we have all just passed 2391, it's surprising we have different views.

So, when performing a Zs on a circuit, or even a Ze, will the meter need to be zero'd?

Is Zs at light fitting now considered too risky? It's been mentioned that Zs for lighting is to be calculated only?

Thoughts?

 

[telectrix]

not sure about other makes of MFT ( are there any). but with megger you only zero leads when continuity testing (dead test). it's not risky to test at a light. croc on E and insulated probes on L an N. we're electricians, not fairies.

 

[Richard Burns]

Only a few MFTs have a zero function on the Zs/Ze test, if the function is there then it should be used, generally it would not be there and so would not be used.
I would not consider a Zs test at a light fitting to be too risky unless there are other problems with the access to terminals, etc.
However there are schools of thought that state that any live testing is too dangerous, which is approaching the level of total idiocy.
Nominally you could do the dead test R1+R2 (or R2), use the supplier provided value of Ze and calcualte Zs, and this would be acceptable under the regulations, however it would be the ultimate worst case and so could fail some circuits that would be compliant in normal circumstances.

 

[westward10]

Seems to be a recurring query no live testing for the Zs reading.

 

[gazdkw82]

Seems to be a recurring query no live testing for the Zs reading.

I can't think of anything more irritating.

Nothing better measuring Zs. Plus, I test Zs on the exposed conductive part (socket face of metal clad socket, boiler housing etc) because this confirms the full path and gives the real reading from the point of shock.

 

[westward10]

On your 2391 did you do live testing for this type of circuit.

 

[telectrix]

banning live testing??? the snowflakes need to get out of their cotton wool suits and get into the real world

 

[gazdkw82]

On your 2391 did you do live testing for this type of circuit.

Yes. In fact I remember being told Zs needs to be performed at every fitting

 

[Silly Sausage]

If one can't test an energised circuit safely, it's definitely time to re-consider one's career choice

What a load of b............

 

[westward10]

Yes. In fact I remember being told Zs needs to be performed at every fitting

Interesting. The Zs for a lighting circuit is normally only done at the end of circuit with the assumption if it is okay at the furthest reach it is likely to be okay prior to this.

 

[gazdkw82]

Interesting. The Zs for a lighting circuit is normally only done at the end of circuit with the assumption if it is okay at the furthest reach it is likely to be okay prior to this.

I think there is some literature to confirm this. Maybe GN3

 

[telectrix]

Interesting. The Zs for a lighting circuit is normally only done at the end of circuit with the assumption if it is okay at the furthest reach it is likely to be okay prior to this.

agreed, but the presence of cpc ( either by measuring R2 or R1+R2 should have already been confirmed at each point previously when dead testing.

 

[Ian1981]

Anyone read professional electrician magazine?
There’s an technical article inside which relates to the fact that for initial verification, then the Zs need not be measured by direct measurement or calculation and should be referred to as N/A to the appropriate column where there’s an upfront rcd in circuit.
Reason being is that if you test the rcd and the results are satisfactory then you will have proven disconnection of the final circuits protected by the rcd.
If a circuit is not rcd protected then Zs needs to be confirmed by measurement by calculation or directly.
I believe the regulations now state that were a circuit is protected downstream of an rcd,then as long as all the exposed conductive parts are indeed connected to the source of earth by a continuity test being performed, then no further testing needs to be done.
It’s in part 6 anyway 643.7.1

Thoughts?

 

[Ian1981]

 

[gazdkw82]

You can't put N/A on an initial verification can you?