Do circuit breakers die of old age?

RCDs do sometimes fail, but more often than not they lose sensitivity, not gain it. MCBs usually tend to last for ever, unless abused.

What type (maker), and what tripping characteristic - Type 1, 2, 3, B, C or D? The MCB trips you're getting are puzzling. One of the following must apply:

- you have a faulty piece of equipment with an intermittent short, etc., or a similar fault in the fixed wiring;

- the total load on the ring is near the limit (~7 kVA) so that the switch-on surge of one more item causes a trip. It would take an awful lot of PC equipment to get anywhere near that;

- the MCB is over-sensitive (not something I've ever experienced).

BTW changing the MCB to 40 A could be dangerous unless anyone's done some design calculations to show that it's OK. It makes the whole circuit non-standard and unlikely to comply with the wiring regs if the circuit's wired in 2.5 mm^2 twin & earth, as is usual.

- I'm tempted to get the whole board updated.

That sounds like a good idea in this case. The whole-house RCD approach is now a strongly deprecated practice, and is always going be problematic in your case, where any trip (for whatever reason) will be a major nuisance.

Upgrade to a modern split-load board with a separate non-RCD protected socket circuit for your office. The sockets will need to be labelled as not suitable for portable equipment outdoors, and Section 607 high-integrity earthing should be applied. All other sockets in the house should be on the RCD protected side. (This assumes TN earthing. If it's a TT-earthed installation a 2-RCD set-up will be needed, or a supply upgrade to PME if available.)

Lets get this right. You have an RCB which is the "main" incoming switch for your consumer unit and an MCB for the ring circuit.

That on occasions either the MCB or the RCD will trip if you turn on a workstation.

First of all when you say "turn on" a work station, is power already going to the PS and you're hitting the (soft) start button, or are you turning on the switch on the PS itself? Secondly does the MCB trip with just one or 2 workstations connected? Are all the workstations the same?

I've had an RCDs trip when I've switched on at the PS which I assumed was due to suppression within the PS causing asymmetric current spikes on live and neutral. Hitting the soft start button on the front has never caused me any problems. Also, if needed, switching on at the wall helped because on mains socket outlets only the live is switched.

For the MCB to trip it must be sensing some large current. The only thing I can think of is that some old power supplies had awful power factors, making the sensed current higher.

The breakers are all Type 2 - i'm not aware what this signifies but it did cross my mind that they might be simillar to standard fuses insomuchas they can be slow blow, quick blow etc - at any one time i might have 10 workstations plugged in but not powered up - however i assume power is still going to their power supplies - i would imagine there to be a bit of a "peak inrush" (i forget the exact term) - demand if all of them had powered applied at the same time (this is what happens when i try to reset the trips) - maybe i need a slow blow type breaker?

Do all new installations/upgrades include RCD main breakers ?

Correct

Correct

I have a number of workstations, any of them has been known to trip either breaker - they are less prone to trip if they are turned on via the wall socket and then via the soft start.

All have 350 - 450W internal PSUs.

Yes, that's the idea. MCBs have two distinct tripping mechanisms - magnetic for "instant" tripping at high current, and thermal for smaller overloads. The "Type" determines where the transition between the two lies. Type 2 (now obsolete) will trip instantly at somewhere between 4 and 7 times the nominal current, so a minimum of 120 A for a 30 A unit.

The current Type B will go at between 3 and 5 times then nominal current, are more sensitive to inrush surges. For your office a Type C might be appropriate (5 to 10 times).

Yes, see above. Even then, switching that lot on in one go might continue to be a problem. Big CRT monitors - any of those? - add to the inrush current, due to the degaussing coils. Staggered switching is the name of the game here. You can buy multi-way IEC distribution blocks with built-in staggered turn-on - they're used in the broadcast industry for switching banks of TV monitors.

No, a split-load board has a main switch, then a bank of MCBs for the circuits that don't need RCD protection, then an RCD that only protects selected circuits. The selected circuits should be only those which feed socket outlets for general use (in particular those likely to be used for portable appliances outdoors) and, if you want to pre-empt the new 17th edition regulations coming in in 2008, everything in a bathroom.

On 9 Nov 2006 01:30:03 -0800 someone who may be snipped-for-privacy@easyrestore.co.uk wrote this:-

Do remember that these "breakers" are operating for different reasons. The main one is operating because it detects a lack of balance between the current flowing in the live and neutral conductors that is above the set value. The ring one is operating because it detects a current flowing on the live conductor for a time that is above the time/current characteristics.

Did you, or anyone else, calculate whether the wiring is suitable for this first?

Intermittent faults are always the worst to diagnose. It needs proper recording equipment to find out what is going on. Guessing what is happening may work, but it may not.

What is the trip current for the RCD? If 30mA then the RCD problem is easy to explain.

Computer power supplies tend to be of the switch-mode type rather than linear. Anyway the point is that switch-mode power supplies generally have a relatively high earth leakage current (typically 3-5mA). This means that it isn't at all surprising that you can trip the RCD occasionally as you must have 2 or 3 power supplies per workstation if you include printers, fax, monitors etc.

To fix this you need not to have an RCD seeing the combined leakage current, i.e. get rid of the RCD incomer. Alternatively you may be able to swap it for an RCD with a higher trip current to avoid nuisance trips (100mA).

Bear in mind that this will then mean none of your circuits is protected by a 30mA RCD and so additional RCD protection is required for any circuits where equipment may be used outside. This can be fixed by using RCBOs for other socket circuits.

As to why the MCB trips, no obvious reason, should only be overcurrent. I'd either test all the power supplies or at least try to identify whether a particular psu is causing the problem by switching them one one at a time (with all the others disconnected). Testing is better.

Fash

What calculation would that be ? - I've simply exchanged a suspect 32A breaker for a known good 40A one as part of a process of ellimination - the same issues were found so i changed it back.

I have 4 TFTs - no CRTs - a small laser printer and couple of hub/routers + a max of 8 workstations.

I normally use the soft start buttons to power up - therefore all of the internal PSUs are being fed from the ring but taking no real load - very often i have all workstations running without issue and can turn on/off via soft start without tripping yet at times even if i just have a couple connected to the ring it blows.

Looks like a non RCD ring + indivdual on/off for each workstation is the way to go ?

I did half come to this conclusion a few weeks ago but have managed to soldier on since - are 100ma RCDs commonly used/easy to obtain? - all i have found are 30ma versions.

On 9 Nov 2006 10:38:46 -0800 someone who may be snipped-for-privacy@easyrestore.co.uk wrote this:-

The rather obvious cable sizing calculations. Had the wiring caught fire during this "process of elimination", which presumably lasted for some time, you might have been in a difficult position.

Sounds like you need the specialist advice of an electrical designer.

Possibly.

The leakage current isn't affected by whether there is load or not from the power supply. And yes it is fairly easy to get hold of 100mA RCDs. Try

Still worries me that the mcb was tripping.

Fash

It would concern me too. There is no reason why a PC power supply should trip a 30A breaker. I would like to know what these workstations are so I can avoid them.

As for a solution either..

fix the PSUs in the workstations

or

use one or more UPS to power the workstations.