disconnect battery = kill alternator, question: Grassroots Motorsports forum: Grassroots Motorsports Magazine

Subscribe get a full year for only $23.95 Sample Free Test Drive—no strings attached

Oct. 21, 2011 7:33 p.m. foxtrapper SuperDork

The claim has been around for forever. Disconnect the battery and the alternator will burn itself out. I've never really understood the basis for this claim, have had my doubts about it, but never bothered to pursue it.

Now, due to snowmobiles and some dirt bikes, I find this needs pursuing. For I find a number of them using alternators, with just a regulator (giving 12 volt ac power) and no battery. As well many running an alternator regulator and rectifier, still with no battery.

They don't burn out, and they have no battery.

So, I find in them some rather solid support showing alternators do not inherently burn out simply because you disconnect the battery, and that this is probably an old wives tale.

If I'm wrong, can you show me where?
Oct. 21, 2011 7:46 p.m. jstand New Reader

I've used disconnecting a battery while the engine is running as a quick and dirty way to see if the alternator is working (at least enough to run the engine electronics).

I'm sure it wasn't the safest way to test, (for me or the ECU).

I try to avoid checking newer cars this way.
Oct. 21, 2011 8:01 p.m. ransom HalfDork

My impression (not to be confused with an understanding) was that disconnecting the battery while it was running was likely to kill the alternator.

Something having to do with a voltage spike, or maybe flyback from the suddenly decreasing current?

I'm not actually qualified to use the word "flyback".

In any case, I would think that the motorcycles would do fine with a big capacitor smoothing things out, and the alternator and other wiring would be happy with the cap there to smooth things out.
Oct. 21, 2011 8:12 p.m. Streetwiseguy Dork

The issue is uncontrolled voltage. Most dirt bikes and such have a system designed to absorb excess voltage. Cars use the electrical mass of the battery to do this. There is no particular guarantee you will harm anything by disconnecting the battery, but should you run the vehicle at higher speeds, the odds of a voltage spike are pretty high.

In modern stuff, the alternator is actually the least of your trouble. I once ate the heater control module in a V70 Volvo just by hooking the battery up with the key in the "run" position. $1400, thank you very much. When we change batteries in modern vehicles now, the car is run into the shop, shut off, the key is placed on the roof of the car, we dissappear for half an hour, then disconnect the battery.
Oct. 21, 2011 8:40 p.m. TRoglodyte HalfDork

I just had a similar issue with sons Mazda3. Unhhooked jumpbox from battery and alternator wouldn't keep car running for any amount of time. Why? Old Chevy's would and this was a quick and dirty test of the alternator.
Oct. 22, 2011 12:38 p.m. AngryCorvair SuperDork

a typical automotive alternator senses Vbatt and adjusts the output accordingly. disconnecting batt with engine running makes the alt think that Vbatt is zero, so it cranks up the output to charge what it thinks is a dead batt, and that's what kills them eventually. i don't know how long it takes.
Oct. 22, 2011 1:24 p.m. noddaz New Reader

The guys at work replace batteries in cars all the time with the car running so the do not have to re-set the radio code or some of the other memory functions...
Oct. 22, 2011 3:01 p.m. iceracer SuperDork

It is one of those things that maybe it will, maybe it won;t. Why take the chance when voltmeters are so cheap. Alernators are not. if the alternator were operated without load or battery the voltage would rise linearly with the speed of the alternator. Posibly rising to as high as 140 v. Paraphraised from my Bosch book.
Oct. 22, 2011 6:10 p.m. kb58 HalfDork

I'm an electrical engineer and sw engineer, plus I work on cars a lot. With newer cars there is no "regulator" - the ECU performs that function without needing a separate box. Because of that integration, the ECU's highly accurate voltage measuring ability and its precise ability to dial up any field voltage it deems necessary to maintain 13.8 V, "should" give it good control - battery or not. That said, only the manufacturer/ECU sw engineers know whether they did or didn't count on the battery as an "energy dump" and buffer. So there's my wishy-washy answer.
Oct. 22, 2011 9:23 p.m. curtis73 Dork

TRoglodyte wrote: Old Chevy's would and this was a quick and dirty test of the alternator.

Old setups used externally loaded regulators, and the only thing they ran were lights and spark. They could stand voltage spikes between 9-18v without anything going bad. These days with rectified diodes, highly sensitive sensors, ECUs, and solenoids, things can get damaged very easily.
Oct. 23, 2011 8:11 a.m. foxtrapper SuperDork

I understand how having the alternator unable to see the voltage it produces can result in the unit charging its brains out, trying to get to 12 volts. But all the circuits I've seen have it seeing that voltage with the battery disconnected. The sensing lead is connected to hot, and will be seeing the alternator output.

The battery would certainly stabilize voltage to a great degree, especially at idle, where the alternator frequently can't produce enough charge.

I could see a minor spike in the systems voltage when disconnecting a battery. And this could cause some problems to components like an ECU, but I don't see it being a problem for the alternator itself.

I can see an alternator killing itself trying to charge a bad battery. Where from the alternators perspective the battery is a bucket without a hole. So the alternator pumps all the current it can into the battery, heating itself up and dying from it. But a battery removed or disconnected wouldn't produce this effect, it would eliminate it. Short the positive lead to ground, and the alternator now would try to pump like mad.

I'm not willing to go disconnect alternators on my various cars just to test all this. I'm just really suspecting this is an old wives tale. Based on something, but overblowing it and missing the real problem. Taking it back in time to generators and external regulators, maybe it's in there. Or back into early alternators with external regulators. But I can't think of one that wouldn't see line voltage with the battery disconnected.
Oct. 23, 2011 11:05 a.m. carguy123 SuperDork

I've never heard of disconnecting a battery killing an alternator. That doesn't make it not so, and in this day of computer controls and solid state stuff I guess anything is possible, but this also doesn't seem logical from an engineering standpoint.

There are any number of situations that it's possible for the battery to not be in the circuit (or appear not to be in the circuit) or not be accepting a charge and that would mean people would be losing alternators all day long.

That would mean old and corroded cables could cause the alternator to pop.

Like I said, not logical from the guy who designs the car perspective.
Oct. 23, 2011 6:02 p.m. iceracer SuperDork

Just one more thing to add. "why take a chance when failure can get expensive". I just bought a new VOM for less than $ 10. Another thing, disconnecting the battery cable can cause sparks, not good around a battery. I never have and never will disconnect the battery with the engine running when checking is so easy.
Oct. 23, 2011 6:25 p.m. CLynn85 Reader

Wow, never heard this one before, might have made me think twice before the stunt we pulled last weekend.

Left key on with parking lights/trailer lights on, load up race car, go to start truck, dead batt, can't get it to jump, pull battery out of race car, put in truck, start, quickly swap batteries back while running, drive home. Driven it all week/weekend and no ill effects yet, guess I'll know who to blame if the alt dies in a week or so though...
Oct. 23, 2011 7:23 p.m. Curmudgeon SuperDork

I've heard the same thing. It was more pointed toward damaging ignition modules, etc.

Back in teh 80's, the Probes (Mazda engine) could develop an alternator problem which would blow every bulb in the car and boil the battery dry. Saw it more than once. So yes alternators can put out stupid high outputs for a short time.