RCD tripping at 1/2 test

[jeremy]

Testing some new cabinets today which had in them a PDU with a 30mA rcd upfront as main switch. ( We build among other stuff, communications cabinets for highways, railways, etc). First one tested tripped on 1/2 fault current. Tried another cab, same thing. Did a ramp test, (tester is Kewtech kt64), tripped at indicated 18mA, though it felt like it was going at 15mA, but the tester indicated 18 before it stopped. ( I think they test too fast for accurate diagnostics?) I would expect a 30mA RCD to hold until about 21-24, from experience, and would have been disappointed with 18, but for 11 to fail at 1/2? One the one hand, I hope I've missed something ever so simple so that we can ship these cabs tomorrow, but I think I've covered everything. Any thoughts?

 

[Sintra]

Did you test at the RCD with no appliances connected? If they are ramping at 18mA then your tester might need calibrating as at 1/2 times it should be outputting 15mA.

 

[thekingiam]

Tried another tester? Try a different make rcd? Disconnect loads and just test rcd?

 

[R G]

Just a thought , was the test carried out with all loads disconnected/ isolated ? , could be some leakage in the equipment connected ..

 

[telectrix]

might be some leakage on the load, as sintra said, can you test with no load connected?

 

[R G]

Did you test at the RCD with no appliances connected? If they are ramping at 18mA then your tester might need calibrating as at 1/2 times it should be outputting 15mA.

Beat me to it ..

 

[trev]

Just a thought , was the test carried out with all lads disconnected/ isolated ? , could be some leakage in the equipment connected ..

The lads were on their teabreak Ray
Sorry mate

 

[R G]

The lads were on their teabreak Ray
Sorry mate

Its those lads again , hahaha,

 

[NickD]

What I think, right, is test it with no loads connected.

 

[jeremy]

Tested with probes, ( there's something not right about the word "probes"!) at RCD load side, tried another tester, same make, 2 years younger, tried a different test supply circuit, tried swearing at it, a lot, ( that usually works). They're A type , in anticipation of a rectifier being fitted by the client, but that shouldn't affect the test. The way I see it is no matter what the circumstances of the test, whether it's a simulated situation or real, it shouldn't trip at 1/2, it should trip at x1 and x5. Trip times for x1 and x5 were the same if that's any use? Shouldn't be too bothered as I've had enough of running a workshop and am going back into the field ( got a job as a scarecrow!), but strangely I care what leaves the 'shop whilst I'm there.

 

[R G]

Although you have tested with probes , was the load completely disconnected , i.e. was the neutral and live to the load disconnected ,

 

[jeremy]

Admittedly, no, there was still an IEC dis strip connected. The only load there in theory is a neon indicator. I'll re-test when I get in early tomorrow. Thanks for all the ideas, sleep well chaps!

 

[Hawkmoon]

Is the RCD definitely 30mA? (easy mistake to make)

 

[thekingiam]

Got caught out by that in the past!

 

[shanky887614]

Is your tester set to 30mA?

A work colleague had his set to 10mA and was wondering why it didnt trip XD

 

[GLENNSPARK]

Testing some new cabinets today which had in them a PDU with a 30mA rcd upfront as main switch. ( We build among other stuff, communications cabinets for highways, railways, etc). First one tested tripped on 1/2 fault current. Tried another cab, same thing. Did a ramp test, (tester is Kewtech kt64), tripped at indicated 18mA, though it felt like it was going at 15mA, but the tester indicated 18 before it stopped. ( I think they test too fast for accurate diagnostics?) I would expect a 30mA RCD to hold until about 21-24, from experience, and would have been disappointed with 18, but for 11 to fail at 1/2? One the one hand, I hope I've missed something ever so simple so that we can ship these cabs tomorrow, but I think I've covered everything. Any thoughts?

27-27.5 is typical for a healthy 30mA RCD

- - - Updated - - -

Is your tester set to 30mA?

A work colleague had his set to 10mA and was wondering why it didnt trip XD

your kidding....

 

[SJD]

They're A type , in anticipation of a rectifier being fitted by the client, but that shouldn't affect the test.

Just to check, on the KT64, on the RCD position, that you've selected "A" from the soft key, rather than "AC", otherwise you won't get the right results.

 

[jeremy]

Thanks again for all the above;hopefully tester definitely set up correctly, as I was showing the newbie how to set it up! RCD is a Schrack 40/2/003-A

 

[shanky887614]

27-27.5 is typical for a healthy 30mA RCD

- - - Updated - - -

your kidding....

Nope, the penny dropped after 5x test

 

[R G]

If the isolation of all equipment shows a correct test reading , I would then assume that you if you have computer equipment involved there is leakage from each piece , I would then look at the idea of separating the equipment to RCBOs , Another point you would have to consider is what these cabinets are being connected to when on site, I have connected PDA cabinets in the past and usually on their own circuit Identified as having high earth leakage currents , and the necessary precautions applied..

 

[jeremy]

Thanks Ray, these cabs are empty at the minute apart from an IEC distribution strip.This morning I tested one of the RCDs on the bench, same results, then tested an ABB one right next to it and it was fine. They're getting returned. Thanks to all for your input.

 

[jeremy]

A quick update. We got a new batch in and still the same results when testing on Auto, BUT test at 1/2 only, not Auto, and it passes.Doesn't seem right to me that it fails during one procedure, but passes in another. I'm going to get in touch with Kewtech and see if they can shed any light.

 

[Plonker 3]

Not read all the replies, but have you checked your meter against a known reading on auto or is it due calibration?

 

[R G]

Thanks again for all the above;hopefully tester definitely set up correctly, as I was showing the newbie how to set it up! RCD is a Schrack 40/2/003-A

I notice it says 003-A at the end of the description , I am thinking thats 3 Ma as 30Ma is .030 A , best to check with manufacturers description , could just be the part number,,

I tried looking up Scrack but theres not much info , as most of the sites are in Polish , I would stick with what you know to be O/K with the ABB you tried ,,

 

[Richard Burns]

Thanks again for all the above;hopefully tester definitely set up correctly, as I was showing the newbie how to set it up! RCD is a Schrack 40/2/003-A

just from Ray's post I notice that Eaton / Moeller supply RCDs with the same part number, could well be Schrack are rebadged Moeller ones, I know nothing about Schrack but might be worth checking and getting some Moeller ones instead.
View attachment Moeller_Datasheet_FI-40_2_003-A_279187.pdf