Power Supply Mystery

Aug 24, 2024 Last reply: 1 year ago 27 Replies

A couple of hours ago the UPS in my computer room started beeping, I thought it was odd because the TV in the lounge was still working. Anyway went to check and found power off to the room.



Looked at the consumer unit and one of the RCDs, marked "ring circuit" has tripped.



Couldn't see any problems so I flipped it up and turned the computers back on. I thought I must check and find out what RCD serves what part of the house (been here coming up to 3 years and not thought about it before).



About 45 minutes later electricity company turned up with flashing yellow lights. Knocked on the door to ask if I had power issues, I mentioned the RCD but he said probably defective kit. He did say that two neighbours had lost power.



Is it possible/likely that some sort of anomaly in the supply tripped one of my RCDs and the main fuse/RCD in a couple of neighbours' houses? Just seems odd.


A fauly neutral in a 3-phase distribution system could cause some properties to have too low a voltage and others to have excessive voltage. John

RCDs are prone to ‘phantom’ trips due to various causes. When old fashioned incandescent bulbs were the norm, one failing could cause a trip. I’ve heard of cases where several fluorescent tubes being turned on can trip one occasionally.

We experienced a similar problem to you a few years back, although we had a ‘glitch’ which tripped one of the RCDs on the split consumer unit then, a little later, we lost power. It transpired there was a problem with the cable at the end of the road. The road was powered by generators for a couple of weeks while it was repaired.

I’m not sure if the design has changed but they used to be very simple devices. Basically, the Live and Neutral conductors form a coil. Each generates a magnetic field. The two fields should be equal and opposite, cancelling each other out. There is a lever which, if the currents don’t balance, is pulled by the now unbalanced field and breaks the circuit.

There is a smaller test coil to force an imbalance to check operation when you press the test button.

Yes. Any temporary increase in neutral impedance would increase the earth current via any suppression devices, making earth leakage trips possible, and any localised voltage surge/dip would have done the same and/or tripped any on the limit RCDs

It does. They don't call it electrickery for nowt....

The bit I din't understand Jeff, is if your computers are protected by a UPS why did you need to "switch them back on"?

Because it only lasts for a few minutes!

For a brief second the current draw wants to be infinite as resistance drops to near zero because an arc is very low resistance ...

It's why MCBs used to trip when an incandescent bulb used to blow

Maybe the UPS is only there to do a controlled shut down

Because the most common time to discover your UPS needs batteries is when you have a power outage.

Indeed.

I used to set my server to return to last known status when power was restored. That meant that after a power cut it would restart automatically.

After a power cut where the power was restored, failed and restored like the proverbial, with some impressive clunks from the disks, I thought better of it, and now leave it off after a power failure.

I had a UPS for 8 years that went though 3 sets of batteries without being used and decided I could live without.

For clarification, was this actually a RCD (with test button), a MCB, or a RCBO (both in one unit)?

formatting link
It is possible for a RCD (or the RCD element of a RCBO) to trip due to a number of external causes - especially if it already has a moderate amount of current imbalance on it (not quite enough to trip it normally, but it is said to be "sensitised").

Less likely that external influences would trip a MCB. That needs either a substantial "fault" current (i.e. typically > 5x the nominal rating of the device), or a prolonged overload - often lasting many minutes depending on how much it was overloaded.

Its possible...

A short from one phase to neutral can have the same effect, increasing the voltage on the other 2 phases.

Not sure, here's a pic on Dropbox:

formatting link
Some are fuses, some are RCD or similar?

Hey, actual wire fuses would blow. Given that open wire fuses are guaranteed to blow within four hours at twice the rated current, that's a pretty severe and prolonged overload.

Only if you bought cheap bulbs from supermarkets. Decent branded bulbs like Osram, were less likely to do this.

Ditto. Only time I lost a crucial boot hard drive was power cut-restore-brownout-restore-brownout-cut.

Now if they go down, they stay down until power is solid again

All the crucial s*it is on the server, and that is generally pretty idle disk wise. So in general after a bit of fscking around, it comes back up cleanly

Fair point. I never investigated to that level and now it's a thing of the past ..

Philips Hint of Pink were the ones I found to take out the breaker on every failure!

nib

I never found a single candle of any make that failed to trip the MCB .

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required