RCD tripping dilemma

joe j · Jun 7, 2012

SPARTYKUS said:
2bobsparky, whereabouts are you in the country?

Now that's what this forum is about. Well done Spartykus

SPARTYKUS · Jun 8, 2012

joe j said:
Now that's what this forum is about. Well done Spartykus

Joe j

Don't follow your train of thought please explain further?

dwardy · Jun 8, 2012

if its a office you might not need all the circuits on an RCD. so can eliminate a few straight away. how is the lighting wired, also are there any outside lights.

SPARTYKUS · Jun 8, 2012

SPARTYKUS said:
Joe j

Don't follow your train of thought please explain further?

Still waiting joe j.

Just to clarify OP, I was asking where you are based in case someone reading this thread could pop in and have a look for you. Not quite sure how that fits with joe j :rolleyes2:

Tonkatoy27 · Jun 8, 2012

How many pc workstations are fed from the circuits protected by this RCD? If it's a 30mA RCD you can only run abot 10 workstations before it trips and thats without leakage from any other equipment. Have they recently added any PC's? There's only one real way to fix it and thats dead with insulation resistance tests at the CU with all the neutrals disconnected from each other.
If you're trying to fix it trial and error by removing bit's and pieces from the RCD then you be weeks instead of hours.

I hope the office is shut at weekends or you'll need a free night and a big torch!
Good luck

Darkwood · Jun 8, 2012

This thread is a week or so old, all this ground has been covered and the OP has been made fully aware of design leakage by lighting computers etc, the limitations of removing the RCD from circuits and what needs doing if it is done for fault finding purposes, also he has been made aware of the testing sequences and the quickest way to identify problem. As recently mentioned its just going around in circles.

note what hasn't been mentioned here is if the leakage is designed high earth then the RCD could be changed to a Main switch and a SI unit 30mA put up front of board, this recognises and ignore certain design frequencies and allows for a greater no' of computers etc.

In reply to Tonk a design allowance of 3mA per work point is advisable thus making a circuit only six workstations at max, 10 workstations leaves you open to the occasional nuisance tripping.

Darkwood · Jun 8, 2012

With regards to my SI-RCD unit suggestion, its may be a quick fix but only if it is high earth design currents, but the installations is poor by design and its this that should be addressed, with all the hours spent searching the fault, a few hundred quid to fit rcbo's and front end with main switch would be cost effective and also will identify the problem circuit if genuine fault leak exists.