Insulation resistance problem

[davesparks]

So if there is still a load connected and I test at 250v would my reading be >200 Mohm meaning everything is good.

You see it's questions like this which just wind me right up, just stop and think about what you are asking for a minute!
What kind of load would have a resistance of >200 Mohm? A resistance of 200M would equate to a load with a running current of 1.2 micro amps, it's not going to be a very useful appliance!

 

[davesparks]

Thanks for that. I have a fluke 1654b, not sure I it has a Kohm setting but will look into it.

That is a terrible meter to use for fault finding, an old school analogue insulation resistance and continuity tester is far more useful.
Personally I favour the robin K3131 but alternatives such as the megger BM7 are equally good.

Digital meters tend to focus your attention on the fine details of an exact figure to however many decimal places and can be easy to misread, an analogue meter gives you a far quicker and easier to identify yes/no.
Also with experience you can learn useful,information about. A fault by watching how the needle moves. Damp or water ingress will cause the meter to move differently to a fault of a fixed resistance or a loose/floating connection.


My favourite for this was a fault in a designer pendant with a 3m long chain. I found the fault when I accidentally knocked it with the ladder and my mate shouted that the needle on the meter had started moving back and forth, in fact it was moving exactly in time with the pendant swinging back and forth. The fault was the lamp holder being somewhat toasty

 

[davesparks]

Thanks dave, I realise now what you are saying but I was thinking if a conductor was just very close to another but not actually touching somewhere would that not be similar to having bad insulation or am I still being simple.

I didn't have time to answer this properly earlier.
If you look up the dielectric strength of free air then you will see that the gap between conductors would have to be fixed at approximately 0.1mm for a dielectric breakdown to occur at the 500V test potential (this is some very rough approximation maths on my part, it will depend hugely on things like humidity and temperature and shape of the conductor)
If this breakdown does occur you will get an arc between the conductors, which is basically ionised air sitting between the two conductors across which energy will flow. I cannot say exactly how this would appear on a meter but I would expect it to either read as a dead short of it can sustain an arc or be an unstable value as the arc strikes and extinguishes randomly due to air currents cooling and distributing the ionised gas (the heat from the arc will create air currents)

But this is assuming you apply a constant 500V such as you would get from an analogue tester, but instead you are using a digital tester.
Now I am sure you have taken the time to actually find out and understand how your tester performs its tests and don't just push the button and read the numbers. So you will already know how the tester will vary the applied test potential based on the actual load and so the reading you get on the display I this situation should be obvious.

 

[telectrix]

So if there is still a load connected and I test at 250v would my reading be >200 Mohm meaning everything is good.

well, obviously a reading >200Meg would indicate that there's no load resistance apparent.

 

[sparky1508]

well, obviously a reading >200Meg would indicate that there's no load resistance apparent.

Well that's what I thought so what would be the point of testing at 250v rather than 500v of its still gonna give me a reading as though a load is connected. Will be doing some more tests today anyway so will post results and hopefully find the problem.

 

[Octopus]

Thanks
No alarm connected.
Will do methodical as Dave suggested.
Forgive me for this but the only reason to test on 250v as far as I am aware is to protect sensitive equipment but I had every plug disconnected from the sockets and had nothing left connected. I know I don't need to unplug but just went round and unplugged everything anyway. Is there another reason I should test on 250v?

Just be aware that some sockets only have sp isolation so you DO need to unplug everything to be sure.

Back to the op why are you doing this test? Is there a problem you haven't told us about? It's always best to post ALL information to avoid assumptions and guesses.

 

[telectrix]

because a reading of > 200Meg. @ 250V tells you that there isn't a load connected. if it reads something like 0.1Meg. or less, then that says there's probably a load connected and by using 250V, you won't have fried it.

 

[sparky1508]

Just be aware that some sockets only have sp isolation so you DO need to unplug everything to be sure.

Back to the op why are you doing this test? Is there a problem you haven't told us about? It's always best to post ALL information to avoid assumptions and guesses.

No problems just practicing with my tester and those were the readings I got so it's good for practice to find out why and do some problem solving in my own house.

 

[Octopus]

No problems just practicing with my tester and those were the readings I got so it's good for practice to find out why and do some problem solving in my own house.

Ok. So methodical testing will find the issue or appliance. All you need is time, patience, a pencil and some paper.

 

[telectrix]

Ok. So methodical testing will find the issue or appliance. All you need is time, patience, a pencil and some paper, a supply of smokes and beer, and loads of patience.

modded that for you murdoch.

 

[davesparks]

Well that's what I thought so what would be the point of testing at 250v rather than 500v of its still gonna give me a reading as though a load is connected. Will be doing some more tests today anyway so will post results and hopefully find the problem.

A lot of people prefer to do the 250V test as it is less likely to write off delicate equipment. I'm not entirely convinced that sensitive equipment is going to survive a 250V DC test any better than a 500V DC test, but then the kind of sensitive equipment I am used to encountering is along the lines of vintage studio mixing desks worth tens of thousands of pounds. I always specify that it is the customers responsibility to switch off and isolate, or at least supervise me isolating, the equipment.

Personally I prefer a standard resistance measurement first to identify any connected loads of i am in doubt.

 

[davesparks]

OP, where abouts in the country are you?

 

[sparky1508]

OP, where abouts in the country are you?

In Andover, Hampshire.

 

[sparky1508]

Had a busy day today but managed to do a test on the circuit.
r1= 0.60ohms
rn= 0.61ohms
r2= 1.04ohms

Calculated R1+Rn= 0.3025ohms
Actual R1+Rn= 0.34ohms

Calculated R1+R2=0.41ohms
Actual R1+R2=0.45ohms

Insulation resistance test
Single circuit @ 250v
L-N = 0.01Mohms
L-E = >200Mohms
N-E = >200Mohms

Split the circuit and did as you said and found that it's the last run of cable between the last socket and CU but looking up the side of the boiler there is a single run of T&E leading to another fused spur but not had time to investigate it yet. At least I know where it is now.
Thanks will post what I find.

 

[sparky1508]

Sorry should have mentioned its on a 32a type 1 breaker with max calculated resistance of 1.38ohms.