measuring 12V 12Ah Lead-Acid battery

forget about measuring the capacity, and just connect it up to your load, and unplug the power cord aqnd see how long it holds the load. The rule of thumb is that if the battery is 5 years old or more, it should be replaced. This time length could be a lot less if the battery has had to supply heavy load current often. In other words, the battery gets 'used up' when it is deeply discharged a lot. If you depend on this UPS for backing up your important computer, then don't mess around. Get a new battery.

Most of my failures have been due to increased series resistance. Just replaced a set yesterday. 6V cells easured almost 7V open circuit and quickly dropped to 4.3V under load. Stick on enough light bulbs to match your load and measure the battery volts. mike

--
Return address is VALID.
500MHz Tek DSOscilloscope TDS540 $2200
http://nm7u.tripod.com/homepage/te.html
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
http://www.geocities.com/SiliconValley/Monitor/4710/

The battery is probably not loading the charging section of the UPS, and it is probably not passing the load test that the UPS does. You must load the battery at about 2 times its amp rating, and determine the time the load should run, in order to properly test the battery. Your car headlamp is nowhere near the proper load test required.

Generally speaking, the battery should be replaced about every 3 years in a UPS. Sometimes I have seen them go for about 5 years, but this is rare. We replace them every 3 years, because it is cheaper than waiting for a total failure, and then having a more expensive data or computer loss.

--

Jerry G. ======

Upon first measurement, I measured 13.78V (holy smoke!!) across the battery. (AC was disconnected for a few minutes before). Figured that's the surface charge, so I hooked up a 12V domelight from my Cutlass until it dropped to about 12.7V in a few minutes. Disconnected light, and now it's back to 13.08V.

So, how do I measure the capacity of the battery, or how do I determine whether it's good, OK, marginal, or bad? I have a couple DMMs, and got a sealed beam headlight I could put on as a more meaningful (though slightly non-linear) load.

The battery doesn't appear bad to me, though I have a feeling that it may be a bit overcharged.

TIA, aur

I read in sci.electronics.design that aurgathor wrote (in ) about 'measuring 12V 12Ah Lead-Acid battery', on Tue, 14 Dec 2004:

Well, you've proved for yourself that open-circuit voltage is no use as a measure of capacity.

Use the headlight to time how long it takes to discharge the battery to

10 V on load. Compare headlight watts/12 (= nominal current) x time in hours with the ampere-hour capacity marked on the battery. If the two figure are in the same ball-park (you won't get better accuracy without a far more complicated test), the battery is OK. For example, if it runs a 60 W light for 2 hours down to 10 V, you have 5 A x 2 hours = 10 ampere-hours. For a 600 W UPS, that's a good enough battery.

Unlikely, unless your UPS is faulty. Make sure that the battery connections are clean and tight.

Sometimes, an APC UPS will give a false 'battery low' indication if you try to start it with a load connected. Disconnect all loads and then start the UPS.

--
Regards, John Woodgate, OOO - Own Opinions Only. 
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

That's normal voltage.

The charging voltage is OK. 13.60V - 13.80V is normal.

Your problem can be bad connection on the battery terminals. You have to tightening the terminal contacts. The current can be over 50A from the battery and you need a very good contact on the battery terminals to prevent voltage drop!

Hmm, this is definitely a possibility, though I'm not exactly sure how do I tighten up the contacts because the battery is equipped with F2 faston terminals, and F2's plus F1's are used pretty much everywhere on the PCB. I assume I can replace the two main connectors on the PCB with something better if the contacts are the culprit, though I'll probably try contact clean, or burnishing first.

BTW, to much of my surprise I found the battery's PDF:

so I'll do a load test on it tomorow; umm, err, later today. ;-)

Thanks for the replies.

Yes, I do know very well that the open circuit measurement may not be very meaningful, that's why I'm asking for help. I guess I'll just put up some load and measure how long it takes run it down to 10Vas you said, and calculate capacity from that. Although I think there exists some equation or formula which allows capacity can be computed from a few measurement points taken in time under some reasonable load.

While some bad battery warning happeneed when I tried to start my PC on a non-fully charged UPS, the irritating ones seem to happen randomly after some time (usual;ly weeks) of running. Turning everything off, then restarting the UPS first usually fixes the problem.

This is simple, but it also puts in the conversion losses.

Nothing is very important that really require UPS protection, but I have quite a few brownouts and dropouts usually lasting for a few seconds, and they can be very irritating.

That's very possible, but the thing I don't understand that it can run for a long time just fine, then fail randomly. The first time it happened was about a year ago, afterwards I lessened the load on the UPS, but it happened again after a while. I'd say the frequency is about 1 warning every other month. If I had daily or constant warnings, yes, I'd say the battery is probably no good.

Very nice, but I don't think I can justify it right now. :-(

Will do that, but I need to get a 0.1 ohm or so resistor to measure current first.

aur

Measuring with resistors is problematic. 0.1ohm and 15amps is

1.5V. that's a lot in a 12V system. Clamp-on DC currrent probe is your best bet. But you don't really care about the 12V current. Load the AC side with the watts you need. The 12V current is whatever the current is. You can't change it anyway. Just measure the voltage for an INDICATION of how sad the batteries are. But even that is moot. It runs for X minutes before shutting off. If you need more, change the batteries. If you don't need more, it ain't broke.

Batteries are often rated at a 20 hour discharge rate. You can't expect to get half the time at twice the current. And you sure can't expect

1/40 the time at the typical 40x current. Search the lead-acid battery vendors. You'll probably find typical curves. Consumer grade UPS systems SERIOUSLY abuse batteries. They draw too much current when on and they charge too much when off. I'd even bet that the cheapo batteries they use in consumer systems don't even meet their spec at 20 hours. Most of us are WAY more concerned about saving an extra buck than whether it runs 6 or 6.2 minutes. mike

--
Return address is VALID.
500MHz Tek DSOscilloscope TDS540 $2200
http://nm7u.tripod.com/homepage/te.html
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
http://www.geocities.com/SiliconValley/Monitor/4710/

You're probably right, but I could always get by without one. ;-) How much do they normally go for?

The problem is not running time -- it's the occasional beeping and battery bad light with seemingly no good reason. (the UPS is in my bedroom) If the battery is indeed bad, I will replace it, but I want to make sure it's really bad, and it's not something else that causing the problem, with bad contacts being another possible cause.

May I suggest the original poster to take the battery to his friendly service station. They have a battery tester that can apply a lot of up to perharps 400 or 500 A.

--
Jean Castonguay
Électrocommande Pascal

Well, it could be the battery after all. I hooked up a headlight that started out at 2.6A, and

1 hour 20 mins later the battery reads only 9.50V (loaded) while the current draw is down to 2.3A. Looking at the terminal voltage/discharge time graph, this battery appears to be around 50%.

PS: it's now down to 8.9V in just a few minutes, so I guess I'm at the knee, and it's time to stop this before I run it down completely.

and it

the

load

is

in a

rare. We

total

Yeah, my boss was complaining because we had to spend several thousand on UPSes to replace those over 5 years old. And it took almost two years, so some of then were more like 7 years. And now we found that the batteries on the 48VDC rectifier for the PBX say that they were installed in '95, so they're over 9 years old. Those are gonna cost some big money to replace.