Dead battery, alternator or both?

[kambites]

As well as holding a charge to start the car, the battery also acts as a bloody great smoothing capacitor. Removing it while the car is running is a very good way to test the spike sensitivity of every piece of electronics in your car. It should cope, but I wouldn't want to bet on it.

Even disconnecting one battery in a circuit containing two in parallel isn't a particularly good idea although the spike is much smaller and in practice everyone does it while jump-starting. Some car manuals explicitly say not to jump-start the car, presumably for this reason?

Edited by kambites on Monday 2nd March 18:54

[andyiley]

Virtually all car manuals say you should NEVER jump start a car, and every single one of us has done it/helped someone do it/watched someone do it/knows many others who have done it.

Contrast that to how many of us know someone, or even know someone who knows someone who has fried a cars ECU etc?

[Mr2Mike]

andyiley said:

Virtually all car manuals say you should NEVER jump start a car, and every single one of us has done it/helped someone do it/watched someone do it/knows many others who have done it.

Contrast that to how many of us know someone, or even know someone who knows someone who has fried a cars ECU etc?

You weren't talking about just jump starting a car, you were talking about disconnecting the vehicles battery whilst the engine was running. What you have advised is incorrect, and indefensible.

[andyiley]

The subject was changed by someone else so I chipped in again.......SORRY!

Just point out to me again where I changed the subject.............


I am very well aware of what I said, and I have not changed my mind.

Edited by andyiley on Monday 2nd March 19:25

[kambites]

As someone said above, car electronics should be designed to cope with spikes but I, for one, would prefer not to test it on a car I own.

Embedded systems are generally surprisingly robust, but that's not a reason to test them if you don't have to. I've never known a car have electronic problems due to voltage surges (at least that I know of) but I have blown up other supposedly protected circuits.

Edited by kambites on Monday 2nd March 19:43

[andyiley]

Like you, I too have blown electronics by accident, I have been in electronics/control systems in many different guises over the years and have never had any such issues on cars electrics over many, many hours spent on such systems.

Is it possible? Don't know.

Have I ever had such issues? No.

Have I ever heard of this happening under test conditions? No.

Have I ever heard of it under jump starting conditions? No.

[Gad-Westy]

A little baffled here. I left the battery disconnected for 5 hours this morning. Came back to the car, checked voltage across battery and it had barely dropped from 12 something. Drove the car for 30 mins or so, checked voltage with engine off and its still 12 something. Left the car for a further 5 hours but with battery connected and then checked voltage across battery and it was down to 2.5v.

So it looks like I have something draining the battery rather than the battery itself being an issue.

I'd noticed that when I check voltage across the battery when its disconnected it reads about 0.5v higher than when it is connected and I've verified that by then disconnecting it again and it jumps back up 0.5v.

So I started unplugging fuses 1 by 1 to see if any changed that. None of the normal circuits had any affect but when I removed the main 80a fuse for the car, the voltage went up 0.5v.

So then I dug a little deeper. If I disconnect the alternator multi plug, nothing happens but if I disconnect the alternators thick single cable, the battery voltage climbs up 0.5v.

So am I reading this right? The alternator itself is draining the battery when the car isn't running...

Edited by Gad-Westy on Monday 2nd March 21:25

[finlo]

Is the alternator getting warm/hot
When its not running?

[LordLoveLength]

Gad-Westy said:

A little baffled here. I left the battery disconnected for 5 hours this morning. Came back to the car, checked voltage across battery and it had barely dropped from 12 something. Drove the car for 30 mins or so, checked voltage with engine off and its still 12 something. Left the car for a further 5 hours but with battery connected and then checked voltage across battery and it was down to 2.5v.

So it looks like I have something draining the battery rather than the battery itself being an issue.

I'd noticed that when I check voltage across the battery when its disconnected it reads about 0.5v higher than when it is connected and I've verified that by then disconnecting it again and it jumps back up 0.5v.

So I started unplugging fuses 1 by 1 to see if any changed that. None of the normal circuits had any affect but when I removed the main 80a fuse for the car, the voltage went up 0.5v.

So then I dug a little deeper. If I disconnect the alternator multi plug, nothing happens but if I disconnect the alternators thick single cable, the battery voltage climbs up 0.5v.

So am I reading this right? The alternator itself is draining the battery when the car isn't running...

Edited by Gad-Westy on Monday 2nd March 21:25

May well be the case that the alternator is draining the battery - it is reasonably common. It is possible that one of the rectifier diodes is duff and the battery is discharging through an alternator winding. If the other diodes are good, it is possible that the alternator can provide enough current to put some charge into the battery and so *appear* to be working.

Simple way to test is to charge the battery off the car, reconnect *when the engine is cool* and wait 30 mins without starting the car. Put your hand on the alternator case - if it's warm it's shagged.
Your battery was probably ok - usually only one cell will fail and the voltage drops by about 2v. If yours has dropped to 2.5v it is being heavily discharged and won't survive. Sorry.

[kambites]

In essence a stationary alternator is a high impedance short across the battery, so what stops it from slowly draining the battery anyway? I assume some sort of logic in the rectifier stops current flowing backwards through the coil when the alternator isn't spinning so if the alternator is draining the battery I'd guess that's what's failed? I've no idea what sort of rectifier is usually used on an automotive alternator.

ETA: Ah it looks like the rectification involves a diode bridge which should stop a flow of current back through the system. I suspect that's got some sort of failure. Try putting an ohmmeter across the stationary alternator and see if the resistance is finite in both directions; if it is, your alternator is buggered (obviously the alternator needs to be disconnected from the battery at the time).

Edited by kambites on Tuesday 3rd March 08:04

[Gad-Westy]

Cheers chaps, very much appreciated. With a failed diode, would you expect to see a charge light on the dash on with ignition on but engine not started. Mine isn't illuminating.

[LordLoveLength]

Gad-Westy said:

Cheers chaps, very much appreciated. With a failed diode, would you expect to see a charge light on the dash on with ignition on but engine not started. Mine isn't illuminating.

There may be 9 or more diodes in there - depends which one(s) have failed.
Put a charged battery on a cold alternator and see if it warms up - it'll only cost you time

[kambites]

I have absolutely no idea without seeing a circuit diagram.

[Gad-Westy]

Thanks folks. Will perform some of these tests today once I get a chance.

Another little observation is that the alternator seems to be whining a little.

[kambites]

Gad-Westy said:

Another little observation is that the alternator seems to be whining a little.

I'd say that's probably completely unrelated. The electronics wont whine and the mechanical bits won't cause current to flow the wrong way through it. I suppose it's possible that the bearings are going and hence generating a lot of heat which has cooked the electronics, but that's stretching things a bit.

[Jimmyarm]

There is a lot of over complicated stuff on this thread.

Test the alternator output at the battery with everything connected, all lights and wipers etc on - should be 14v+

If it isn't, turn everything off, disconnect the big positive cable from the alternator, try not to shower sparks everywhere and check the output directly from the alternator with the engine running.


If your battery is draining when connected, you need to test current draw not voltage changes. Youtube etc has plenty of videos showing you how to check that. I can explain it if you need it