RCBO tripping problem that I cannot get to the bottom of.

[Phoenix]

Wonder if anyone can make any suggestions on this, I've been asked to get to the bottom of an RCBO tripping problem after others at the firm didn't manage to fix the problem, and as to date, I haven't yet.

Issue is within one of the treatment rooms (consult room but equiped to do minor procedures such as wart removal) at a Doctor's practice. The cirucit thats tripping is a ring final thats decicated to this room, serving 3 or 4 twin sockets on dado for the computer equipment, a single low level outlet for the medical fridge, a twin socket the other side for an adjustable bed (this wasn't even initialy plugged into this circuit) and a fused spur for an examination lamp. Cirucit is wired in 2.5 LSF twin and earth on basket tray and in conduit and dado. The set up is the same for 3 other rooms, all of which are fine.

The issue started following the RCBO being tested on an EICR and now it generally trips at least once a day, often more. Generally the issue will occur when the medical staff use the printer for the first time each day, the job will be sent and straight away the cirucit will trip before the printer has has any chance to make any noises, so its like its tripped just as its entered its warm up cycle (laser printer).

I wasn't quite sure what my colleagues had done, but I turned up, tested the cirucit for IR with LN strapped together and tested to earth (which was greater than >1000 on 500v), I proceeded to change the RCBO for a brand new fresh one (and also as a result it ended up upgraded from type AC, to A not that this should have a difference, other than its certain it'll be from a completly different production run, and its a nice to have thing to be upto the current standard) When I changed the RCBO, it was clear that from how the neutral flylead was routed that it had already been changed by a colleague, but ah well, I already had the board open and the new one ready so it went in. RCBO tested out fine, I tested the IT appliances through their IEC leads and all were fine (>1000) although that was actually a L&N to cpc test rather than to earth as a whole, because I was doing it with clip leads from the plugtop. The fridge wasn't tested at this stange, because I'd realised I'd left it plugged in and switched on when I tested the fixed wiring (so it was tested then, and this was to earth as a whole). The adjustable bed wasn't plugged into this cirucit when the problems started happening, instead being on a cleaners cirucit. The examination lamp was turned off that the double pole MK fused spur, hasn't been tested, but this is generally not on when the issue occurs.

I plugged the IT equipment back in and booted the machine up without any issues, I repeatly switched in a 1.5kw heating load as I knew laser printers have a fuser element, and wanted to rule out the load of that causing earth leakage through some N->E path which I was sure couldn't exist without me having found it, but wanted to rule it out, no issues. got the medical staff to log on, and print something, RCBO straight out, reset, subsequent prints all fine (this is what they say they get, always in the morning, often first print after lunch, and very occasionly a run of 2 or 3 trips on printing before it settles down)

They tell me they have had the printer changed (they changed it themselves, and called IT to remotly configure it - so they are certain its actually been done) Frdige has also been changed, and I think the PC may have been as well.
I've changed the RCBO and can find no IR thats less than >1000meg on cirucit or appliances.

I ran an extension lead through the suspended ceiling and connected the printer to the cleaners cirucit instead, expecting to see the tripping move to this cirucit instead, no, nothing tripped at all, I went back and added the fridge onto this as well just incase it was a combination of the two. I was going to try the printer and fridge on the room cirucit and the computer and screens on the cleaners cirucit, but they were due a CQC inspection and while it was not routed in a way that was a hazard, we needed to get rid of the extension lead, but I'm not sure whether I'd have gleaned anything useful from it.

I did find a bit of an tech note from Brother (who make this printer) about AFDD tripping in the US and they had altered the firmware to avoid this on affected models, now while its not an AFDD involved here, I did think that if the printer was switching something in a way that could be mistaken for an arc then that could be adding high frequency interference on the mains supply and increasing capacitive leakage in the cabling and other appliances, long shot, but I've seen actual series arcing faults casue a downstream RCD to trip through something like that (loose tail in henly block causing RCD main switch of the board it fed to trip). So I got them to ask IT to ensure printer was on the lastest firmware, which didnt change anything.

I'm running out of ideas, although as I typed this, it did occur to me that the the tripping after a period of inactivity is probably significant, I did consider the evironment, but its a modern building (6 years old) with mechanical ventilation from all the rooms and no signs of any dampness issues, but its just occured to me that the printer sits on top of the medical fridge, I'm wondering whether there could be an issue with the closeness of the printer to the condensate boil off tray on the fridge, but again a long shot, but I would have expected to be able to trip the cleaners socket circuit with the printer if that was the case.

If anyone has any ideas, I'd be grateful for any help here before I tear what little hair I have left out!

 

[binky]

It would clearly seem to be a function of the printer from what you describe.

Could you swop printer for a similar unit from one of the other rooms.

Question is, could you not dispense with the RCBO and just fit a MCB?

 

[Sharpend]

Have you done a ramp test on rcbo? Have you measured earth leakage on circuit?

 

[UNG]

Had one a while ago where the actual fault was on another circuit and not on the one that rcbo was tripping and the RCBO on the faulty circuit didn't trip although under test it was ok. Narrowed it down in the end to a cumulative earth leakage issue and a neutral earth fault on the circuit that wasn't tripping

 

[Marvo]

IT equipment with switch mode power supplies can cause common mode noise. Depending on the capacitive value of the EMC filters and any PFC circuitry and the frequency of the noise it can cause RCD/RCBO's to nuisance trip and this tripping tendency can be amplified if it's a ring final circuit with uneven impedance values or dissimilar conductor sizes on the L and N legs. It's not possible to measure this type of noise with electrical testers so usually you'd track it down using deduction.

If the printer has an external power supply I'd start by replacing it with an alternative.... customers lie almost as much as IT guys who replace printers so assume it could still be a problem plus even if they replaced the printer they could have used the original power supply if it had the same plugs. Same with any other PSU's that are easily swapped out, try replacing them one by one. Maybe also try an identically spec'd RCBO with a similar form factor from an alternative manufacturer although if it's a 'type tested' distribution board this may be against your regs as a permanent fix.

 

[Phoenix]

In answer to the questions:

1) I'm not sure I could swap the printer with another one from another room without IT invovement as I think the terminals may be locked down to specific printers (I'm guessing, but they have extra trays on for prescription blanks with a code lock on, so I preseume security is tight) and the printer has definatly been either swapped or renewed already

2) I am sure I ramp tested it, but cant remember the figure, other than it was pretty typicalish, 24mA rings a bell, but cant swear to it, that was with no load on the circuit

3) I haven't measured the Leakage on the cirucit, I should have done, but I'm not sure how reliable my leakage clamp is because it has got minor damage to the interleaving laminations of the core of the CT on the jaw. I'm not sure how much leakage is likely to be present in a standing state with only one workstation though, but its certainly soemthing to try when I go back. It would be nice to try and catch the increase in leakage of whatever is tripping it, but I think its going to be too fast to register

4) MCB woulnd't fly, not only would I struggle to find anyone at the client who would be able issue a risk assessment, I can't imagine anyone would be willing to do so for sockets that could be used for general use, in an area the public are present within, plus, I think we are bordering on section 710 with these areas anyway, plus it just seems like a daft thing to have to do, whats on the cirucit shouldnt cause issues.

5) Printer has internal power supply and takes IEC C13 kettle lead in, and they tell me thats been changed to (and I've IR tested it and found no issues)

 

[Mike P]

I had a very similar problem, it turned out to be a laptop power supply, even though the laptop was not connected. It was class 2 and obviously had 2 core from the plug. The Megger DCM305 found the issue (brilliant bit of kit). unplugged power supply all went away.

 

[bryanh]

If plugging the devices into another room via a lead and they not tripping ( I assume they being fed via a rcbo yes?) it must be the circuit in the room not the devices?

 

[Geoff1946]

I'm wondering about high frequency effects from harmonics due to the printer power supply.
The effect of these could vary between similar installations with different wiring layout and coupling between adjacent cores in the trunking, etc.
I would be trying connecting the printer via a filter.

 

[infinityelectrical]

Daft question but have you done an insulation test with the earth's connected to the earth bar, and live to neutral?
If it's only serving one small room with sockets that could be unplugged and double pole spurs then all equipment could be suitability isolated for testing L-E N-E. Got to be worth a shot at this point.

I would also put an earth leakage tester on the circuit and plug the equipment in an item at a time and watch the readings, could be that with all the items plugged in the RCBO is very close to its threshold and doesn't need much to push it over the edge!
Sy