"UPS" dead after power cut

[JohnW2]

In a spin-off in anotherthread ( click here ) ...

Don't rule out something as prosaic as a fuse blowing when a bunch of old batteries suddenly demanded to be fed when the power came back on.

No fuse (or equivalent) in sight. However, the battery is shot. Nominally 12V, it has 9V off-load, but that falls to about 2.5V when I connect (mains) power to the unit. That in itself is rather odd!

I would have thought that the 230V output sockets would have been directly connected to the 230V input, via NC contacts of a relay (there are 5 PCB-mounted relays in the unit!), when the supply was live - in which case there should be output whenever there is a live input, even if the electronics were non-functioning (due to fault, or if, for example, they derive their power from the battery), hence all relays in their non-energised state.

Some further and more systematic investigation is therefore required. I guess the starting point will be to see what happens if I temporarily give it a decent, charged, 12V battery!

Kind Regards, John

 

[bernardgreen]

An intelligent UPS may refuse to give any output to the load if it finds it has a flat battery even though mains is available.

The logic behind this is to ensure that critical equipment does not receive power after it has shut down, possibly in a uncontrolled way due to loss of power. Human intervention is needed to ensure the equipment is in a safe state to be repowered.

These UPS may have a manual over-ride to connect mains to the output to enable equipment powered by the UPS to be made safe.

 

[deadshort]

If the battery holds the relays on during 'normal' 230v operation and the battery is knackered when it was required to operate the battery is completely flat due to the demand then surely, when the power (230v) is reinstated the relays will remain open as they have zero power from the battery until they are fully charged or replaced.

Kind Regards,

DS

 

[JohnW2]

An intelligent UPS may refuse to give any output to the load if it finds it has a flat battery even though mains is available.

As I implied, that is obviously one possible explanation, and hence the first issue I am going to investigate (with a non-flat battery!). However, if that is the case, then the (presumably constant 'trickle') charging must have fooled it into thinking that the battery was OK - since it was providing output right up to the moment of power failure.

The logic behind this is to ensure that critical equipment does not receive power after it has shut down, possibly in a uncontrolled way due to loss of power. Human intervention is needed to ensure the equipment is in a safe state to be repowered.

I'm not sure that I completely follow that.

Kind Regards, John

 

[JohnW2]

If the battery holds the relays on during 'normal' 230v operation and the battery is knackered when it was required to operate the battery is completely flat due to the demand then surely, when the power (230v) is reinstated the relays will remain open as they have zero power from the battery until they are fully charged or replaced.

Sure, if the relays worked that way around, what you say would be correct.

However, as I said, I had imagined that they would be 'the other way around' - so as to be 'fail safe' in the sense that output would be maintained whilst there was power on the input, even if battery and/or UPS electronics failed. In other words, I had imagined that the relay would be unenergised whilst power was available, hence powering the output sockets via its NC contacts, and would only become energised (and disconnect output sockets from input ones) if/when the inverter fired up in response to a power failure (I suspected that this might be why there were so many relays!).

However, it's starting to look as if that is not the case, and that what you suggest may be correct (and there may be a good reason, such as bernard has suggested). If that is the case, the device gives protection against power failures whilst it is functional (and with a working battery), but not against failure of UPS electronics or battery whilst a supply is available- so I guess one has to decide which is the more probable risk!

BTW, I test all my UPSs (by confirming that they respond correctly to removal of power) approximately 3-monthly, and the one in question worked fine when last tested on 3rd October. However, my tests don't usually involve leaving the power disconnected for a long period of time, so maybe not long enough for the ailing battery to go flat. If that transpires to be the problem, I will need to modify my testing procedure!

Kind Regards, John

 

[ericmark]

Some times safety is built in which we don't expect. I remember when first working on PLC's a thermostat controlled heater would not power. I tried shorting out sensor and it did not power up so was convinced it was something to do with the program.

However it turned out the guy writing the program realised the temperature in the factory should never drop below 15°C so in case sensor got shorted out wrote it into the program should the temperature drop below 15°C to turn heater off.

A shorted cell in a battery may cause hydrogen sulphide or other nasty gases so if the UPS is designed to switch off at 11.2 volt should the battery drop to 11 volt it may consider there is a fault so refuse to charge the battery.

Step one would be swap battery for a good one and see how it works even if temp borrowed from car. (Assuming it will not upset radio if disconnected).

 

[JohnW2]

Some times safety is built in which we don't expect. ... A shorted cell in a battery may cause hydrogen sulphide or other nasty gases so if the UPS is designed to switch off at 11.2 volt should the battery drop to 11 volt it may consider there is a fault so refuse to charge the battery.

Yes, that is one of the many possibilities. However, as I said, if that's the case then it must have been satisfied that the battery was OK so long as it was being 'trickle charged', since the device was working fine right up to the moment when the power failed. It's obviously possible that the battery then rapidly 'went flat', leading to the present problem

Step one would be swap battery for a good one and see how it works even if temp borrowed from car. (Assuming it will not upset radio if disconnected).

Indeed, as I've said, that will be the first test - but not with the battery from my car, since that would, indeed, upset the radio (and perhaps other things as well - oh for the days when cars were simple, and almost entirely 'mechanical'!!).

Kind Regards, John

 

[ban-all-sheds]

You could try charging the batteries with a car battery charger - I did that once with a UPS which appeared to be dead, in case the gubbins inside did need a bit of life in the batteries to work relays etc.

And lo, they did. Managed to get the batteries up to enough for the UPS to come to life. The batteries were still kaput, but at least I knew that if I bought some new ones I'd be in business.

In your case though you are probably sure enough that the UPS is basically sound, the batteries are knackered, and replacing them is what's needed. Time to give Mr Yuasa some of your hard-earned.

 

[JohnW2]

You could try charging the batteries with a car battery charger - I did that once with a UPS which appeared to be dead, in case the gubbins inside did need a bit of life in the batteries to work relays etc.

That is precisely what is going on (or meant to be going on) as I speak/type - but at the last count didn't look as if it's going to work, since precious little current is going from the charger into the battery! If that fails, I'll dig out another (healthy) 12V battery from somewhere to test with!

And lo, they did. Managed to get the batteries up to enough for the UPS to come to life. The batteries were still kaput, but at least I knew that if I bought some new ones I'd be in business.

Yes, that's the reassurance I'd like before buying a new battery ...

In your case though you are probably sure enough that the UPS is basically sound, the batteries are knackered, and replacing them is what's needed. Time to give Mr Yuasa some of your hard-earned.

Indeed - but, as above, ideally I'd just like the reassurance that the UPS itself is 'basically sound' before I do that.

Kind Regards, John

 

[JohnW2]

And lo, they did. Managed to get the batteries up to enough for the UPS to come to life. The batteries were still kaput, but at least I knew that if I bought some new ones I'd be in business.

Yes, that's the reassurance I'd like before buying a new battery ...

OK - I lost patience with the 'minimal current' charging malarky ....

With a decent, charged, batter, the UPS springs to life and works fine ... so that answers most of the questions about 'how it works', and indicates that I simply need to buy a new battery. I also need to do longer tests on all my others, to make sure that I haven't missed any other 'nearly dead' batteries!

So deadshort was presumably right (and my guess wrong) in terms of 'which way around' the relays work (i.e. that they send mains-derived power to the output sockets when their coils energised, which requires there to be ~12V across the battery), perhaps for the reason suggested by bernard (although I didn't totally follow it).

Thanks for everyone's interest and contributions.

Kind Regards, John

 

[Monkeh]

The simple answer is that the UPS control circuitry runs off the battery, and only off the battery.

All relays should be normally open.

 

[JohnW2]

The simple answer is that the UPS control circuitry runs off the battery, and only off the battery.

That much I always realised - it would make no sense to have a PSU for the electronics which was separate from the 'battery charger' (and floating the battery across it probably makes PSU design easier in some senses). However, in terms of what I was speculating about how things worked, the control circuitry would not have been required in order for power to get from input to output (i.e. it would go through NC relay contacts).

All relays should be normally open.

I now know that to be the case in my UPS, certainly for the one which sends mains-derived power directly to the output sockets. You seem to be implying that there is some reason why it 'must' be like that - do you perceive a safety-related reason, or something of that nature, for not using NC contacts?

The implication of all this is presumably that the benefits of having a UPS diminish when the battery is approaching the end of it's life - since the risk of loss of output due to battery death might then become greater than the risk of loss of power due to a power failure?!

Kind Regards, John

 

[Monkeh]

You seem to be implying that there is some reason why it 'must' be like that - do you perceive a safety-related reason, or something of that nature, for not using NC contacts?

A UPS in the off state must not consume substantial current (holding an NC contact open..), nor must it connect its internal inverter and the incoming supply together. This pretty much requires all relays be NO.

Schematics for common UPSes are educational and quite easily available (speaking Russian is usually to your advantage, however).

 

[ban-all-sheds]

It does seem wrong that a UPS with a dead battery can't be plugged in and start the battery charging.

 

[ban-all-sheds]

speaking Russian is usually to your advantage, however.

I used to think I knew how to say "Cheers!" (as in with a drink), but I was mistaken. Common mistake, apparently.