Battery puzzler - 9V

[David]

Just had the mains powered alarm chirping every now and then. Puzzling in itself because although this should be a sign that the backup battery is at the end of its life, why just chirp now and then? Been doing it for the last couple of days, but only two or three chirps every few hours.

Anyway, turned of the power to the alarms and took this one down. Took the battery out and checked it with my old analogue meter and it seemed to be at 9V.

Dug two possible spares out of the sod it drawer and tested them as well. One showed around 9V and the other nearer to 10V. Anyway, shoved the higher voltage one in and refitted and powered up the alarm. No chirping. Will have to wait a day or so to confirm no intermittent chirping.

Decided to check the other batteries again with my cheapo ScrewFix digital meter. One showed a solid 8.85V. Further checks showed that it was out of date in

2010. The one from the alarm initially showed 9.04V and 9.05V with the reading flicking between the two. Checking again, it showed 8.85V and the voltage started to slowly decline as I kept the meter on it.

Puzzled, I tried by analogue meter on it again. Solid around 9V and no decline.

Tried the digital again and it showed 9.04/5V.

Tried it again and got the declining reading.

So it seems the battery was at the end of life and declining slowly, but a spot reading indicated it was fine. I thought when they went, they went down and stayed down.

Checking what the nominal voltage should be (I assume that this is the same as that read by a meter) WikiPedia suggests 9V for Alkaline and Zinc Carbon with 9.6V for Lithium. So the solid 8.85V is probably toast for anything that is very voltage sensitive although the same article says "nearly dead" is around 5V and most devices are designed to cope with this voltage range.

As you can gather, I am puzzled by the erratic behaviour of this particular battery. Is this normal for 9V batteries?

Cheers

Dave R

[Ash Burton]

It's fairly normal.

--- news://freenews.netfront.net/ - complaints: [e-mail] ---

[Toby]

A digital meter presents almost no load on the battery, where as the analogue one presents a load - if you test the battery with the digital and analogue meter at the same time, it will almost certainly show thwe voltage dropping on both.

So at no load, the voltage seems OK-ish, but when presented with a load, it drops.

[spuorgelgoog]

Because chirping uses up electricity which the detector is conserving for detecting and alarming purposes.

This allows it to remain operational and present an intermittent alarm over a longer period of time. No point in it warbling madly for an hour then dying completely if you happen to be out for that hour.

Owain

[Cursitor Doom]

Or worse still on holiday.

[Graham.]

Yes.

What can happen is the ambient temperature drops, causing the terminal voltage to droop below the "chirp" threshold. The initial chirp draws current causing the voltage to reduce more and the chirps continue in a vicious circle. A rise in temperature may be enough to break the loop, but not always. I've seen this particularly when lithium 9V batteries are used in smoke alarms.

[Malcolm Race]

Not only Litium. 20 approx years ago my elderly mother was disturbed at night by chirping noises which she thought came from her phone. She reporrted a fault to BT and one of the first questios asked was 'Do you have a smoke alarm?' She did, close to the phone. The problem was an alkaline battery dropping in voltage as the temprtature dropped at night.

Malcolm

[alan_m]

In my experience when the (alkaline) batteries start to fail in my smoke alarms the chirping always starts at 2am/3am. Often ignored at night the offending alarm cannot be identified in the morning after the sun has come up and/or the central heating has kicked in.

The OP shouldn't overlook that there is also a battery in his digital multimeter that may also be failing BUT readings tend to go high as a DMM battery starts going flat.

[tabbypurr]

nominal is not actual

only if very poorly designed

yup

no chance

NT

[Brian-Gaff]

You need to measure them on a load, not just open ended. OK so it will use up the battery a bit but this will give you a better handle on whether the battery is on its way out. Basically with any multiple of one cell, the usual way they go is that as the cells are in series, if one cell goes down before the others, it in effect gets reverse charged through the load, further knackering it and making it, effectively into a resistor, which drops volts the more you load the battery. What does puzzle me though is how any dry battery goes down if its just for back up and not used very much. It implies it is being used, or that the battery is a heap of crap in the first place! I have also had pp3s that drop volts in the little rivet which holds the spring grip connection to the battery. corrosion or something obviously. Brian

[whisky-dave]

Well batteries are made from chemicals which age just like everything else. It;s lioke my grommet draw where teh grommets have 'metlted' the plasic draw and sort of stuck to the plastic, wierd.

Virtually everything ages except gold it seems.

[David]

Except that the analogue meter showed no slowly dropping voltage - it was the digital one which showed the strange behaviour.

Thanks to all for the explanations.

Cheers

Dave R

[Johnny B Good]

That's *always* true on account the ADC reference voltage accuracy becomes compromised by lack of voltage. The drop in the reference voltage makes the test voltage appear higher than it really is. Normally, the DMM will display 'Low Battery' well before the voltage reference becomes compromised by an expiring battery.

I think most (if not all) DMMs choose to carry on displaying 'erroneous' readings (along with the low battery warning) rather than simply refuse to take any measurements under low battery conditions altogether as the 'Lesser of two evils' with the low battery warning offered by way of mitigating this particular evil by warning the user that the accuracy may well be compromised into a falsely higher reading, rather than leave the user totally blind to the presence of (a potentially lethal) voltage in the circuit under test.

ISTR that the 'standard' DMM test circuit impedances are 10 and 11 megohms on the DC voltage ranges (it may be higher on ranges above the

200v mark) whilst typical moving coil analogue meters can range from 5 to 50 kilohms per volt of the selected scale (20v FSD setting representing a test meter impedance of 100 Kilohms to 1 Megohm) with 20 Kilohms per volt being the more common (400 Kilohms meter impedance on a 20v FSD setting).