IET logo
 
IET
Decrease font size
Increase font size
Topic Title: RCD trip
Topic Summary:
Created On: 02 May 2015 02:49 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 02 May 2015 02:49 PM
User is offline View Users Profile Print this message



bacong

Posts: 2
Joined: 25 July 2008

After few weekends of testing to see why RCD was inadvertent tripping. I found on ramp tests, it failed straight away at half current test. 21ma actual value. It is a 60 A 30ma. Have also changed it out to a new one which trips at 19ma an MK and my original is a Hagar.

Here's the thing, my neighbors RCD trips when mine does, which got me to start testing in first place, also losing freezer full of food. I have knocked some loads off, had two fridges and small chest freezer in garage. I have measured a constant 80ma of earth leakage in the supply earth to the D board. It is still there when I knock the main supply off. UKPN been out as well, could not find anything wrong on my side. I have TN-s to supply sheath, however they found the neighbor had removed his earth clamp from the sheath and put it around some metal armoring! Which they corrected. Im the end of the run as last house built. testing my loads and all coming uo fine resistance wise and continuity. It is oHL supply to short cable run, about 30m or so from poles.

I still have the earth leakage, bills a re expensive, which is why I contacted the supply folks. Any Ideas if 80ma is acceptable and would be causing teh RCD's to trip early, would PME earth system help resolve issues? No keen on increasing RCD to higher value ma.

Thanks
 02 May 2015 03:43 PM
User is offline View Users Profile Print this message



alancapon

Posts: 6881
Joined: 27 December 2005

The requirements for testing RCDs are that they should not trip at 50% of their rated current, and must trip at 100%. Anywhere between the two, they may or may not trip. As such, your values of 19mA and 21mA for a 30mA device are both acceptable, as they are more than 15mA and less than 30mA.

Regards,

Alan.
 02 May 2015 05:50 PM
User is offline View Users Profile Print this message



slittle

Posts: 3678
Joined: 22 November 2007

I'd be thinking there's a dodgy neutral on the network somewhere if both RCD's are going at the same time.

Have you tried a loop test P-N or tried measuring N to real earth and put a good load (kettle & oven) on whilst doing it ?

Stu
 02 May 2015 07:00 PM
User is offline View Users Profile Print this message



bacong

Posts: 2
Joined: 25 July 2008

Thanks Alan, some comfort there.

Stu,

I did do a loop test and got 0.35 ohms, which is acceptable. I have not done the kettle test though , something to try next. Thanks for this
 03 May 2015 08:18 PM
User is offline View Users Profile Print this message



IronFreely

Posts: 322
Joined: 06 November 2014

I'm fairly sure it's acceptable for a 30mA RCD to trip anywhere between 15 and 29.99mA, at 30mA it must disconnect, below 15 it must not be affected.
With that in mind 80mA of leakage us bound to cause problems.

I'd be inclined to think that the problem is current leaking through some shared pipe work or similar as your neighbour seems to be affected, have you tried using your clamp meter around copper work in both your houses? I'd check the water main first. Failing that it'll be a bog standard earth fault somewhere, probably N - E fault as its intermittent.
I'm not sure that a problem with supply N would cause this as the RCD would detect no Imbalance unless the problem with the N conductor was load side of the RCD, I'm happy to be re-educated by brighter sparks but I doubt RCD's ever are affected by imbalances supply side of them.
It all sounds quite Shakespearian, "A curse on both your houses!"
I'm not surprised that your bills are high if you're loosing current perpetually, it all adds up over time, even 80mA is roughly 1 kilowatt hour of wasted energy ever two days or so.
 03 May 2015 11:27 PM
User is offline View Users Profile Print this message



aligarjon

Posts: 3852
Joined: 09 September 2005

RCD's aren't supposed to operate with supply side faults. however i had a customer who's rcd kept randomly tripping during bad weather. the supply company took some convincing but it was traced to trees blowing against the overheads up the valley somewhere. Once they had been sorted the problem disappeared.

Gary

-------------------------
Specialised Subject. The Bleedin Obvious. John Cleese
 04 May 2015 10:26 AM
User is offline View Users Profile Print this message



mapj1

Posts: 9366
Joined: 22 July 2004

An RCD detects an imbalance in the two coils and although that is normally a load side thing, and implies a leakage to earth or cross-coupling to some other circuit, for transients this can indeed be due to things happening on the supply side.
The load side may have perfect insulation to earth on both L and N, but there will be incidental capacitance to ground from both wires and some loads that have filter capacitors L-E and N-E as well,- which normally creates a milliamp or two of earth current at 50Hz, which has no ill effect.
However, if for example the neutral bounces relative to earth suddenly by a few volts (as it might from a line side step change in resistance at a scratchy joint on a pole for example) then the capacitance to ground on the load side will have to near instantly charge to a different voltage. To do this a large differential current flows though the RCD briefly, the step being superimposed as a spike on the normal 50Hz sinewave. Unless there is an identical transient on the line at the same time, the RCD will see this as a short duration imbalance (and a time delay type will ignore it, while an instant trip type won't.)
Filtering on the supply side may help, but the real solution is to eliminate the source of the transients.
Such problems are rare, but perfectly possible - and waveform recording apparatus capable of seeing sub-millisecond events is the most reliable way to prove it.

-------------------------
regards Mike
 05 May 2015 12:31 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 15838
Joined: 13 August 2003

constant 80ma of earth leakage in the supply earth to the D board. It is still there when I knock the main supply off.

That doesn't sound like "leakage" from your installation, but diverted N current from the supply. It's most usually found on PME supplies (values of several amps aren't uncommon), but as most TN-S networks have been repaired using TN-C-S techniques, it's likely you have a PME supply in disguise.

Try clamping around both L+N supply tails together rather than the earthing conductor - that should tell you what's really leaking (on the same cancelling principle RCDs themselves use).

- Andy.
 05 May 2015 10:00 PM
User is offline View Users Profile Print this message



IronFreely

Posts: 322
Joined: 06 November 2014

The reason I love this forum is I learn so much detail that you just don't pick up in a few years of training. Mapj1 once again you have enlightened me.
 06 May 2015 01:57 PM
User is offline View Users Profile Print this message



lyledunn

Posts: 1107
Joined: 13 August 2003

I agree Ironfreely, if Mike also gave golf lessons I would be with him in a shot!
We should be grateful for the experts like Mike on this forum who often give comprehensive comment. I subscribe to a fire safety forum where the so-called experts often snipe and sneer from the sidelines. We should also be grateful for the less expert lads and lassies on the forum who bare their souls in pursuit of knowledge and opinion. Altogether, a great forum!

-------------------------
Regards,

Lyle Dunn
 06 May 2015 02:49 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9366
Joined: 22 July 2004

I feel flattered, as I'm sure do other regular commentators, but a word of caution - the most seasoned poster is no more infallible than the next guy, and it is quite common to get the wrong end of the stick when a complex problem is being explained in three lines initially, sometimes resulting in spectacularly odd mis-advice.

Please never switch off the sense that this is no more reliable than a chat with a guy at the bar in the pub after work, and so, as you would then, don't hesitate to chip in with the 'are you sure - I thought it was more like XYZ ?' if it appears to be so much tosh.

Speaking personally my day job can sometimes involve having to be very careful not to say anything too committal, (*)and I enjoy the sense of anarchy here. I certainly learn a lot of funny stuff I would not otherwise, it has been a long time since I did much domestic for example, and there are folk popping up here doing things I never have, so this is a fun way to keep up with some current trends/problems. In 'payment' I offer my opinion on stuff I think I know about. If this chatter also does not send others to sleep, and/or improves their knowledge then that's a good thing too.

(* "without prejudice or commitment" gah !)
Edit:
So much for sentiment, back to business.
One way to detect or monitor such 'snap crackle and pop' events is to have a specially tricked filter.
Here is an example from a DIY article from the late 1980s (I actually used to read this very magazine, but by 1988 I had stopped - there being a limit to the number of Christmas tree flashers and lights on reminders one can sensibly build.)

Whole article on web - circuit description in mid page
The circuit cascades two sets of classical ferrite ring and capacitor delta filters, and dirty enters on left, clean leaves on the right, but there is a pick-up overwinding on the first ring, and this is used - almost RCD like- to detect and indicate fast out of balance transient events.
In this case they flash some LEDs, and I would not recommend this exact circuit - the 12V power supply for the electronics in particular is toe curling (those LEDs are another example of a mains derived FELV sans isolation - the things we used to do).

Modern high tech versions would replace the LED bargraph with a micro controller and some totaliser displays " XX events exceeded so many volt seconds since last reset"
However the basic idea of a filter that has been tricked to allow monitoring of the fast edged currents either in the inductors or the capacitors is re-used as the basis of 'arc fault indicators', which are commonly used in countries with lightning problems or a history of scratchy infrastructure.
Simpler detector/ noise tracing variants have an amplifier and put the picked up noise onto a small speaker - so you can hear the noise, as this often provides a good clue as to the origin.

-------------------------
regards Mike


Edited: 06 May 2015 at 11:45 PM by mapj1
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2017 FuseTalk Inc. All rights reserved.

..