RCBO passes 30mA but fails 150mA but in a strange way. Any ideas?

[slipnduck]

Morning everyone, this is my first post - although i had an account years before that seems to have been lost.

Can any of you shed some light onto a situation I've come across.
I have a newly installed RCBO which is failing the x5 test with a reaction greater than 40 mS.

Now, I don't have a problem with that as some devices do fail straight from the box but.... it's the manner in which it fails that I find perplexing.

It's a Type A b curve CP Fusebox RCBO at 40A - so admittedly budget kit.
When tested on the testers AC setting it tests fine - 30mA and 150mA both causing sub 40mS activation at 0 and 180 deg.
However on the type A setting it activates at greater than 40mS on x5 BUT it activates at 36mS on x1 at both 0 and 180.

How can it be that it can activate at the lower fault level in a time that would satisfy the higher fault level test but then go on to fail the higher fault level test?

I'm not sure exactly what the tester does for the Type A tests but I assume it just introduces a DC current whilst the test is in progress along side the simulated AC fault current. I thought then that the AC current is increased for the x5 test whilst the DC component stays the same. If this is the case then the result seems very confusing.

Mybe the DC component of the test is also increased by the tester for the x5 test but I wouldn't have thought that would be the case as they are for up to 6mA DC and you'd think that is the current used by the tester for both x1 and x5 tests.

I don't mind swapping the device out but I'd just like to understand why it's happening.
Any ideas?

It's a new tester BTW - Metrel MI3152 - which I'm still getting to know. But the settings used in the tests are the standard ones for a 30mA General use RCD on both type A and AC.

Thanks

 

[slipnduck]

Are you testing the RCD with no load connected?

It was still annoying me so I've just been back there (got the key to the place).
When I tested it on Friday it was from an energized socket outlet with no other loads connected.
As I said I've just been back, I re-tested at the load side of the RCBO with it isolated downstream.

Same result when I do it on auto sequence - my tester does x1 first then x5 then x1/2 then ramps

at x1 (30mA) on the type A setting it gave - 37mS on the (+) of the cycle, 28mS on the (-)

then at x5 it reacted in 57mS on the (+) of the cycle - it then halts the test as it has failed.

I then did it with non auto sequence at x5 IDn - it lets you chose which side of the cycle or you can chose both one after the other.

It's always the (+) that it fails on and on the (-) it reacts in the same time as if you were testing it as an AC device -ie without the DC component 28mS.

It kind of makes sense I guess as the DC would shift the AC signature in one direction (+) ????
But why is the reaction time slower at x5 than at x1??

Here is another strange thing I did it on the double test a few times (+&-) after failing first time at 57mS if I repeated it straight away it would react at 28mS on both (+ & -)
But I couldn't get it to pass on consecutive occasions, it would pass then next time fail at 57mS on (+) then next time it would pass at 28mS on both (+ & -) the next time fail again on the (+) side.

There was an audible difference in the sound it made when it failed to when it passed, only slight but noticeable- not quite as crisp a click.

Anyway I've removed it but I still want to know the 'What and Why' I'm not content with it being a 'gremlin in the gizmo'.

 

[slipnduck]

Have you ramp tested the RCBO in question?. I've had quite a few RCD's / RCBO's where the trip current has exceeded the 30mA level fresh out of the box
While it is not a requirement I always ramp test RCD's and RCBO's now when I install them and also on EICR's, I started doing it after I was asked to do some other work at a property 6 months after I had replaced a CU, when doing the testing on the work I noticed the readings from the RCD tests were a bit different to 6 months earlier dug a bit deeper and found that on a ramp test the RCD tripped at 36mA. Since then I have found quite a few that would pass the prescribed tests but fall short when ramp tested

I was about to say yes then but now I think on it, the ramp tests come at the end of the auto sequence unless performed as a single test and seeing as it always failed at the first x5 on (+) it halts the sequence. Now I can't recall if I did it as a single test being so caught up in the weird times it activates in.

 

[slipnduck]

It almost sounds like an AC type being tested with a A waveform and the sense transformer saturating!

Yes, it's marked as an A. But why would saturation be greater on the x 5 test? This is where my understanding of how the test works comes up short.
Is the DC component increased from the x1 test to the x5 test therefore increasing saturation and thus reaction time?? I'd love to know that.
I'm gonna try and talk to someone from Metrel next week and ask exactly how the test is performed.

 

[sound circuits]

I'm having exactly this problem with the same gear: a Fusebox 40A Type A RCBO, Metrel MI3152 tester. I replaced the RCBO with a brand new one and got exactly the same results. It a cooker circuit feeding an oven and induction hob. Isolator switched off for the test.
I'm getting 38.6ms on + 1x and 28.4ms on - 1x, then 54.7ms on + 5x then FAIL. I've tried doing individual tests instead of Auto and I can get a reading of 28.6ms on the - 5x but the + always fails. ??????????? Wish I had another MFT to compare!

 

[pc1966]

Are you seeing the same if tested on 'AC' type?

 

[sound circuits]

Are you seeing the same if tested on 'AC' type?

No, it's passing when tested as 'AC'. Set to AC I get: + x1: 29.0ms / - x1: 38.9ms, +x5: 28.3ms / -x5: 18.1ms

 

[newfutile]

Try testing at 250 ma , i had this trouble with a Hagar who told me to test at 250 ,also saying the regs state test at more than 5 X

 

[i=p/u]

Theres videos about it. They telling you to use higher trip current...

 

[sinewove]

All so,in the May PE mag it says test type A 30ma rcds at 250ma,Nicey advice no less.
S

 

[sound circuits]

Ok - just watched GSH's & John Ward's videos on testing type 'A's at 250mA - I'd bet any money this is happening here, however I can't get a reading from my Metrel. I've tried setting it to 500mA at x0.5 but it just fails it without giving an actual time. How annoying. Oh for a Variable setting like on the Meggers. I've just emailed Metrel to see if there is a firmware / software update available.