Strange rcd results ...any views

[WELLSIANA]

had to test an existing rcd (wylex) during a modification...3 circuits protected only...sockets 1 sockets 2 & garage...original tests produced 59.9 on x1 and off scale on x5.....so i thought it was up the shoot and changed it...results then 36.6 on x1 and 39.9 on x5 ..but on one of the x5 tests it came up error...so i disconnected all circuits and tested the rcd alone...came up with perfect results...so i added a circuit one at a time and no matter the rcd became ok to pass but erratic...like with just the garage connected only results went x1 ..29.6 & 0.00 and then on x5 30.0 & 0.00...it always seemed erratic on the second part of each test ...yet first tests where fine....test button ok etc and all electric on and working...my intial thought was there must be a very small fault but as i separated and tested induvidually i just can not accept that view...welcome your thoughts

 

[Taylortwocities]

You should always test RCDs with all loads disconnected or you may get misleading results.

What did the IR tests look like on each circuit - that would show anomalies.

 

[Iggifer]

I call a load causing the erratic results, as TTC says, test with no loads

 

[Spark123]

I call a capacitive effect of the cable or a capacitor.

0.00mS?? sounds like your tester is due a new set of batteries or a re-calibration too.

 

[WELLSIANA]

the IR was fine and as was r1 +R2...i will admit as it was a property where to get to every socket would have been difficult i did test it with items plugged in still...its just i have never had it before ...so as i left items in what would i or could i pick up to make it erratic ...its annoying as in all my time i have not expierienced it before so i want to know what could have caused it and NOOOOO its not the batterys or my tester ..hmmm

 

[flyingsparks]

What makes you think it's not your meter? Have you tried it on another RCD at a different site?

 

[securespark]

Is it a Metrel?

 

[Spark123]

I can't imagine any mechanical device disconnecting in 0.00mS hence I'd be questioning the test instrument.

 

[PrenticeBoyofDerry]

the IR was fine and as was r1 +R2...i will admit as it was a property where to get to every socket would have been difficult i did test it with items plugged in still...

So can I assume then, that the above tests where carried out with plugged in loads then? So it is also possible that your r1+r2 results are incorrect as some appliances will add resistance to the readings and how did you perform the IR test when items were still plugged in?
The only tests I perform with items plugged in and on load are the functional ones!

 

[WELLSIANA]

the tester is a calibrated fluke 1652...has never had a problem b4...and when loads where disconnected at con unit it proved rcd working fine...it was only when you added a load...i do take on board your views on items plugged in but you have to understand and realise in this domestic property it was very not practicle to get access..like built in units / bookcase etc....i only added to sockets in a sun lounge and as the results are erractic am just looking for general views of cause with the fact to take away all loads i would have needed a builder...just general opinions and thoughts..

 

[PrenticeBoyofDerry]

i do take on board your views on items plugged in but you have to understand and realise in this domestic property it was very not practicle to get access..like built in units / bookcase etc....

I do understand, it's the nature of the job and come across oh so many times, but although but you must be aware that some items of electrical equipment will give you false reading on both r1+r2 and IR tests, also IR testing through loaded items (baring in mind just switching of at plug may not disconnect all conductors) could also damage vulnerable equipment!

only added to sockets in a sun lounge and as the results are erractic am just looking for general views of cause with the fact to take away all loads i would have needed a builder...just general opinions and thoughts..

In my opinion if you remove all loads from the RCCB then test at board, this should give you the correct readings, but if removing the conductor from busbar terminals and MCBs, then I would again test Zs to confirm they are terminated as before RCD tests.
Is there a second RCCB on the board and have you tried ramp testing?