Car Battery voltage gone Crazy

[Jerry.Tan]

My Chevy pickup truck was getting hard to start on cold days. I took it to a service garage and had the guy put a battery tester on it. The results showed it to be borderline between Fair and Good. The alternator was fine. The guy said the battery should last till Spring. WRONG! The other day I parked at a local restaurant to use their WIFI. and plugged my laptop computer into my inverter. About 40 minutes went by, when the computer switched to it's internal battery and my truck's dashboard clock went to 12:00. That little 300W inverter actually drained the car battery so much the dome light was barely visible.

I got a jump and drove home. The next day I bought a new battery.

Anyhow, my reason to post this is because the truck has a built in volt meter on the dash. All my older vehicles had either an idiot light, or ammeter gauge. The ammeter would always indicate charging or discharge, but the volt meter is a little harder to understand.

But I'm a little puzzled by the erratic readings on that volt meter when I had that failing battery. Normally (and with my new battery), it's almost always reading about 14V (normal), and slightly less after starting the engine or if the lights were left on for awhile after the engine was turned off.

When I had that failing battery, I'd be driving down the road, and the volt meter would suddenly drop from 14v to around 11v. Then it would slowly creep back up. This was under a constant load, (not when I was braking and using brake lights, etc). This seemed to be happening more and more, and I would often notice the headlights would get a little dimmer when that meter dropped. That's why I had the service station test my battery and alternator.

Oddly enough, while the battery tested borderline between Fair and Good, it apparently was much worse than the test showed. What I dont understand is why the volt meter would fluctuate, drop from 14v to 11v and then go back up, and do so repeatedly...... My thinking is that one (or more) of the cells were shorting out randomly. But I'm just guessing.

Can anyone give a better explanation?

Note: With my new battery, the meter is staying at 14v all the time and everything is working well, so that shows the problem was the battery itself, not charging components or bad wiring or a bad volt meter.

[Meanie]

I think you answered you own question. Unless it was a loose cable or connection (which could also cause the same symptoms), the logical conclusion would be the plates/cells within the battery. Since the guy told you fair to good, I would believe it was more fair to bad, edging on bad and thus, your meter was reading the sporadic changes within the guts of the battery.

I'm sure someone will come along with a more precise answer explaining the chemical reactions of battery cells, but the general logic is obvious, IMO.

[CRNG]

On Sat, 13 Dec 2014 05:24:17 -0600, [e-mail] wrote in

Was the engine running while you used the inverter?

[philoÂÂ]

Since I was in the battery business I might as well jump in here.... and yes, it looks like you guys have figured out the probable cause.

There are insulated separators between the battery plates and one fairly common mode of failure is for a separator to slip out of place. Places normally shed active material which usually just settles down to the sediment tray at the bottom of the battery...but if a separator slips out of place, "mossing" occurs between the plates.

It is very possible for shorts caused by mossing to be intermittent.

If one cell is shorted the battery voltage indeed would drop to approx

11 volts. When the battery was put on the tester is was obviously in a non-shorted condition.

This is one situation where the service tech would have been right to advise the purchase of a new battery.

[Ed Pawlowski]

You either had a weak battery or bad connections. If the battery is more than a couple of years old, it would have been smart to replace it. Winter is not here yet and you had starting problems. Does it make sense to wait? Not for me. 40 minutes on the inverter could kill a battery even with just the draw of a laptop.

You were smart enough to have it tested, but they were not smart enough to tell you the battery was in trouble. Now you know what to look for a few years from now when the battery is approaching the end of life.

[Jerry.Tan]

-NO-

When the battery was still good, I ran that inverter/laptop for 3 hours once, and the truck still started, but just barely. It turned over very slow. But I know that was too long. Normally I'm not online more than an hour in the truck.

Of course, with this failing battery, combined with colder weather and shorter days, causing me to use the lights more and the heater fan when I'm driving, that 40 min. sure took it's toll.

Rather than burn up a lot of gas to use my laptop, in my vehicle, I think I'm going to get a small mower battery to carry around, and just put it on a charger when I get home. The built in battery on the computer is old and dont last more than 15 to 20 minutes. The built in batteries cost more than what I paid for the 8 year old laptop. It's cheaper to just buy one of those $25 batteries made for riding lawn mowers.

[Jerry.Tan]

Great explanation. Thanks. But I am left asking "What is MOSSING?"

I do agree that battery had an intermittent issue, which apparently was not occuring when it was placed on the tester.

One thing I wonder, does the movement of the vehicle such as bumps in the road, cause those separators to move, and thus cause intermittent shorts? There did not appear to be any pattern to the voltage drops, they just occurred randomly. Yet I did notice them more on one local stretch of road which tends to be a rough ride.

[clare]

I think you pretty well hit the nail on the head, and your Garage was using 1950's technology to test the battery (a simple load test).If they tested it with a Midtronics battery tester I can almost guarantee it would have failed.

[Jerry.Tan]

I'm not sure what a "Midtronics" tester is, but it was just a load tester. I once owned one of them, but it died and I never bought another one, since I rarely ever used it. A lot of places will test a battery for free these days.

[Meanie]

Agreed but having your own load tester is a plus to avoid pulling the battery and taking it in and in some cases, discovering it's good.

[philoÂÂ]

tery.

Yep, a loose or slipped separator would cause an intermittent problem on a bumpy road.

To me, that seems like the most likely explanation.

A more obvious problem would have been an internal "hard short" such as broken plates permanently touching. In that situation the battery would have been about 10.5 volts no matter what

[dpb]

On 12/13/2014 12:28 PM, [e-mail] wrote: ...

A link again to a set of lecture slides on Pb-acid btteries...

Includes some cutaway photo's that make it pretty easy to see what can happen...

[Ashton Crusher]

I don't know about Chevy but I know that FORD runs their voltmeters from the computer and they ALWAYS read at the 14v mark until something is SERIOUSLY wrong and then they drop down a lot lower. It's not really a voltmeter, it's just a go-no go indicator just like an idiot light. It tells you you have "good voltage" OR you have "bad voltage". That is probably why you saw it fluctuating like you did, because the battery was bad, but not because it was intermittently bad, just because it was running right at the computers switch point between bad and good. The way to verify if your voltmeter is a REAL meter or just an idiot-meter is to run the car at idle, see where the voltmeter needle points, and then turn on the headlights on high beam. If the needle doesn't drop at least a volt or two it's just an idiot gauge. A lot of cars temperature gauges are similarly set up although not as go-no go but still computer controlled so the computer tries to hides bad news till its really bad by restricting the needle so it's not linear with temperature rise.. They are better then just an idiot light but not as good as a true gauge.

The average driver thinks he wants gauges but he really doesn't, the fluctuations make him nervous and result in extra warranty work when someone complains "my oil pressure keeps dropping". Most oil pressure gauges are also just a go-no go nowadays.

[Ralph Mowery]

That may explain it. I did not see how the voltage could go below 12.5 volts, and really how it could go much below 14 volts with the engine running unless there was a loose connection. Unless the alternator was bad, but with a new battery and no other changes, it would tend to eliminate the alternator.

[clare]

Not if the battery was any good. The laptop won't draw more than 75 watts - most run closer to 50. At 75% efficiency (and virtually all current inverters are better than that) you are looking at a maximum of 100 watt draw, which at 12 volts is less than 8.5 amps - with 40 minutes being .66 hours =12.88 amp hours

[clare]

A "midtronics" tester is the brand name for a patented transconductance? tester that can check the condition of a battery even when discharched. It measures internal resistance and voltage recovery rate and a whole bunch of other things that translate to the capacity of the battery to store and deliver power.

They are not cheap, but MOST battery wholesalers and retailers worth their salt have one and use it religiously. If a battery passes on a midtronics, it is good and is not replaced under warranty. Start looking for another electrical problem that is drawing the battery down or not charging it properly. If the midronics says it is bad, it WILL fail to start your vehicle very shortly if it has not already.

No garage should be selling or servicing batteries without one today. Period.

[clare]

In my long lifetime working with automobiles and auto electrics I have never run across a voltmeter, ammeter, temperature guage, fuel guage or speedometer that works as an "idiot guage" or "go no-go" indicator. Inhave seen MANY oli pressure "guages" that do. In fact the vast majority of so-called oil pressure guages just tell you if the pressure is higher or lower than 4 or 5 psig.

[N8N]

Did the problem go away after replacing the battery or no?

Maybe the brushes in the alternator are worn, and the test did not reveal t his because it just happened to be working OK when tested. How many miles on it?

nate

[Ashton Crusher]

In my 99 Mustang I can put the whole cluster in TEST mode and drive the car with it in TEST mode. I've done that. The first thing teh computer does when it goes into test mode is to cycle the needles on ALL the gauges thru their full range and then they go back to their "normal" position. In other words, ALL the gauges are run by the computer, none of them are actually directly connected to the sensors. In addition, when it's in TEST mode the Trip Odometer displays the actual sensor readings and you can cycle thru them all as well as some P codes. Each time you push the odo reset button it cycles to the next sensor.

So you can display the actual temperature on the Odo display while watching the temp gauge. And you can see that the temp gauge needle climbs as it warms up and when fully warmed up sits right exactly in the middle. When it's warmed up the Odo reads 90. But on a hot day in traffic you can see the odo readout of temp go up to 100 but the needle for the temp gauge doesn't move off center. It's clear that FORD had programmed the temperature gauge to read actual temperature until it gets to 90 degrees and after that it holds the needle at mid scale until presumably some much higher temperature is reached when, again I have to presume as my car doesn't run hot enough to get to this point, the computer will kick up the needle.

The voltmeter seems to be similarly computer controlled. You can watch the voltage vary on the digital odo display as you rev the engine and turn stuff on and off but the needle on the gauge never moves. For the oil gauge there is no sensor, there's just the go-no go switch that goes on at about 7 psi. The odo readout is either 0 or

255 representing an open switch or a closed one. If it's closed the 255 tells the computer to tell the gauge to go mid-scale. It never goes higher, never goes lower except for when it first goes into TEST mode where the computer makes it go all the way from 0 to full scale in a nice smooth sweep and then back to zero. (it goes to zero if the engine isn't running)

The gas gauge and Speedo and Tach are also run thru the computer but I have no reason to think the readings are being "faked" like the temp and volt readings get faked. It wouldn't make sense to fake the gas and speed and tach readings....

As I understand it, if the IP cluster ever goes bad I have to have any replacement programmed at the dealer to match my cars VIN or something or it won't communicate with the computer and won't work.

[Jerry.Tan]

Yes, the new battery solved everything. It starts great and the voltage is constant.