Fluctuating headlight brightness

[JHZR2]

As many know, I have a 1998 Chevy pick up truck that doesn't get a lot of use. It has gone through two batteries in recent memory and so there may be a very slight short-circuit someplace. Regardless the alternator voltages are very good and put out strong enough to always keep the battery charged so it can go for a long time just sitting. Operationally everything is fine. However I was working on it the other day and have noticed then as well as when I have operated the last few times, that the headlight brightness seems to fluctuate a little bit when it's sitting at idle. Before when operating it I thought I noticed this but couldn't really confirm. When I was working on it recently the headlights were beaming directly against a dark garage door and it was very obvious.

In addition when I was working tonight I could also tell if there was a slight fluctuation in the brightness in the gauge cluster. I certainly had never noticed that before.

I know that during operation the voltage is very steady. Oddly I took a voltage reading from both the accessory 12 V plug as well as a reading on the OBD II port. These values are not the same but I'm not sure if it's due to calibration or some other issue.

I would imagine that the auto parts stores have an alternator test rig that I could use. That said I really don't trust that considering that two weeks ago my parents had alternated I was failing and the store tested it as good. So what should I look for I assume I should search for AC ripple characteristics on the alternator output. Of course I can check the battery both load and impedance characteristics, but I'm not sure that that's going to do me much good. If the battery impedances to high simply won't charge without a huge overvoltage, I just won't take very much current which means that the alternator should actually be working less. Of course if it's dumping into a short circuit that would be an issue I do have a DCM clamp but because I don't have a baseline reading I don't know if I can make heads or tails out of it.

Any ideas what typically causes headlight brightness fluctuation? Suggestions on what tests I should run?
Thanks!

 

[ZeeOSix]

Originally Posted By: JHZR2
Any ideas what typically causes headlight brightness fluctuation? Suggestions on what tests I should run?
Thanks!


If you are seeing other lights such as the dash lights dimming at the same time, then it's obviously the whole electrical system's voltage level fluctuating. Probably the voltage regulator. Have any way to hook up a voltage tester that logs data at a decent rate (like 10 Hz or more)?

Another thing you could check is all the electrical system's major ground cables where they attach to the frame. Also, make sure the alternator belt is adjusted tight enough and not slipping any.

 

[JHZR2]

It's only a fairly slight fluctuation, in reality, but just enough to notice if you're looking for it hard.

You're right that perhaps my meter is taking an RMS value, or just doesn't sample at as rapid enough rate. I don't have a scope but that's why I was thinking perhaps to be a way to use the AC setting on my meter.

The belt is fine and the tensioner is as well, at least as far as I can tell by observing.

Many many years ago I thought that the bearings were going bad on the alternator because they were a little bit noisy with the mechanics stethoscope. I may have been wrong I may have been right I can't really say, but I didn't put any refurbished alternator at one point. Probably stupid... But like I said my parents had alternator that failed not 50 miles after they had a test done at advance auto parts that said it was all good, so I'm interested in doing some sleuthing myself.

 

[spackard]

One or two diodes in your rectifier are blown open.
There are six, two for each of three windings. When one or two blow then you get good voltage for 2/3 rotation of the alternator, then no voltage for 1/3.
(IME, the diodes short then overheat and melt open, will be visual damage.)

 

[Gito]

Had the exact symptoms on my 91 Honda Civic 3dr Si. Replaced the belt and problem went away. Would not of discovered an issue if it were not for driving at night and early morning. Alternator belt probably wasn't grabbing on to the pulley.

 

[MolaKule]

Originally Posted By: spackard
One or two diodes in your rectifier are blown open.
There are six, two for each of three windings. When one or two blow then you get good voltage for 2/3 rotation of the alternator, then no voltage for 1/3.
(IME, the diodes short then overheat and melt open, will be visual damage.)


Just to add to spackards comment, a damaged or "leaky" diode in the alternator can cause a slow battery discharge. A leaky diode can exhibit resistance in both directions, verses a good diode, which is supposed to have almost zero resistance in one direction, and a very high resistance in the other.

A diode is supposed to conduct current in only one direction, but if it is leaky, it can allow current to flow backward from the battery to ground.

Quote:
You're right that perhaps my meter is taking an RMS value, or just doesn't sample at as rapid enough rate. I don't have a scope but that's why I was thinking perhaps to be a way to use the AC setting on my meter.


You should be reading your meter with the DC setting of 20 volts or so, unless it is an autoranging meter.

 

[oldhp]

On that old of a truck make sure your battery to ground, engine to ground and body to ground are tight and not corroded. Clean all connecting points and use star washers at body grounds. A bad ground will do exactly what you are describing, as the VR gets confused as current pulses because of bad grounding. Look up the Big Three's when adding a big stereo, it works!!!

 

[JHZR2]

Originally Posted By: MolaKule


You should be reading your meter with the DC setting of 20 volts or so, unless it is an autoranging meter.


Yes but small signal transients on the rectified dc are what I'm interested in, ie if there is some semblance of an AC waveform due to bad/busted rectification.

 

[eljefino]

Do you know if a Body Control module runs your alt or if it's old school completely internally regulated?

BCMs are funky in what they want, and what they think.

I figure it's a battery or cable issue though.

If you like nerdy stuff, hit the alt with some cold air and see if the temp compensation is flaking out. Don't use the computer stuff as it's flammable-- usually 134a!

 

[SteveSRT8]

Likely a ground, that's what we run into here as fleet vehicles age. Easy to fix if that's it...

 

[Win]

I guess you could scope the output.

But, a semi WAG is your alternator is due for replacement.

I replaced the one on our '99 S-10 a few years ago and got rid of a lot of electrical gremlins.

It seems like I also had to replace the one on our '97 S-10 Blazer.

I don't think these alternators were one of the General's better efforts.

 

[JHZR2]

Who rebuilds a good Delphi alt for these applications? Bosch? Remy?

 

[Win]

I got both of mine from big boxes - either both came from Advance, or one was from Advance and one was from AZ.

The core charge was pretty low - they didn't really care if they got a core back or not, as best I recall.

We used to have a guy here that would rebuild starters and alternators - a little embarrassed to say I don't know if he's still in business.

 

[artificialist]

I never understood why if a starter or alternator was a weakness in a car, I would see no parts company invent something better.

Just think of all the aftermarket parts that were invented to replace weak OEM parts in other parts of the car.

 

[Kiwi_ME]

Those are both fundamental electric machines, ideal for the job. Not much more to invent unless you've got a flux capacitor in mind.