[somewhereouthere (OP)]

Hi, hoping someone can answer this mystery

In my garage there is this old wall clock that has been lying down on a shelf for many ears, with the same AA Duracel battery, not in operation, due to the 'dead' battery.  But I have just been to the garage and I heard this ticking sound.  To my amazement I discovered the clock had come back to life.  How could this be?

I removed the battery and put it back in but it didn't work then. The strip on the side of the battery says it's dead.  I tried it in another clock and it didn't work.

The clock had not been disturbed recently, and there was no sign of tampering as it's covered in dust and there were no finger prints.

I don't know how long that battery has been in there but I think for over 5 years.

The temperature is about 7 celcius, and there's been some rain.

So do I have ghosts or something more rational?

[RD]

I don't know how long that battery has been in there but I think for over 5 years.

Gases can accumulate in (old) sealed batteries, causing them to be pressurised.

... there's been some rain ....

A change in atmospheric-pressure could then cause the electrolyte in the battery to move,
(like a barometer), and consequently change the performance of the battery.


[Changes in atmospheric-pressure alone are sufficient to power a mechanical clock]

[alancalverd]

The test strip on a Duracell places a load on the battery to test its internal impedance, which is what usually determines death. Problem is that the load from the test strip will be enough to kill a battery that is close to death anyway! 

[Petrochemicals]

As the battery degrades its not impossible to believe some charge may accumulate enough to move a clock for a few moments. Take the battery out and give it a rub.

[Bored chemist]

Take the battery out and give it a rub.

A genie may appear and grant you three wishes.

Warming the battery might also help.

It's possible that the clock has circuitry which shuts down when the voltage falls below some cut-off.
Variations in temperature or pressure may cause enough change in voltage to bring it above that threshold (briefly).
It's also possible that mechanical vibration (opening and closing to door as you entered the garage) shook the connections slightly and improved the connection.

[vhfpmr]

The test strips on the battery apply quite a heavy load, so that they won't indicate the battery is good when it hasn't sufficient power to operate the heaviest load the manufacturer thinks you may want to use it in. Clocks, on the other hand, use an absolutely minimal amount of power, which is why batteries in them last for years. I save all my AA and AAA cells when they're too flat to run the appliance they were used in, and then use them in the mantlepiece clock. A 'flat' battery will run the clock for many months, particularly if it had previously been used in a heavy load. The test strip is telling you nothing.

Batteries operate more efficiently when they're warm, so it is possible that a slight warming of the room might tip the voltage over the threshold at which the clock starts working. It's unlikely though, if it's been standing for years the self discharge of the battery would have taken the voltage low enough for it not to re-start with a slight temperature increase.

My money says it's almost certainly an intermittent fault, and not a flat battery. The clock stopped working because of the fault, started working spontaneously because that's what intermittent faults do, and then stopped again when you disturbed it.

Intermittent faults can be the bane of an electronics engineer's life, because if the symptoms won't present themselves when you're trying to find the fault, there's a limit to what you can do to root them out. Many intermittent faults boil down to loose/broken/corroded components or connections, in which case heating, cooling, thumping, prodding and vibrating are usually the first resort, not least because they're quick and simple. As well as thermal expansion disturbing a dodgy connection, temperature variation can also affect component parameters, as can supply voltage, but that's less likely in this case.

On a clock, the only external connections you'll find are the battery terminals, if they look ok and aren't corroded, you won't get any further without pulling it apart. That might be easier said than done though, they're not usually designed to be serviced, and might not come apart without breakage. Inside, you're likely to find components that are too small to repair without considerable skill and experience. A cracked chip component for example is likely to be a lost cause, because even if you had the skill to replace it, you won't have any means of establishing its value once it's broken, and without instruments.

[somewhereouthere (OP)]

*update*

I've tried the battery again with the 2nd clock and it is working, and has been for the last half hour, this is despite it not working earlier and the battery strip saying it's dead.

Im guessing the battery must b unstable and the weather has affected it, causing it to come back alive.

I should add the battery has been in doors all day where it is warmer, but when it was in the garage and started to work the air temperature was very cold (about 5c).

In the garage the clock had been in the same place for many years, rarely disturbed, and I don't think it would have picked up much vibration from things like the door being closed, as it was sitting on top of other objects on the shelf, and the shelf is attached to a breeze block wall, which are soft so absorb vibrations.

So it's definitely the battery, but why it's working again is the mystery.

*update update*

I have just noticed that battery has the date March 2018 on the side, which is apparently the expiration date.

The battery is an AA Duracel Ultra Power.