took the car out for a drive, while on the road constant speed smooth highway, the ammeter starts going nuts, 60+ then about -30 amps. then settle down then starts up again, lasting several seconds.

I heard NO odd noises (but a 70mph prob would not hear over road noise anyway). No other indication of problems (radio fine, lights seemed normal (I turned on the HL to soften the charging down to about 30 amps indicated).

I was not far from home so just disconnected the VR and drove on home, no more intermittent anything, just a constant to be expected discharge rate. Installed a new backup VR that had been tested on a running engine, same thing random wild swings but mostly, like before.

I tried unplugging the connector from the back of the alt, shot it with some deoxit, same with the shunt (actually removed it to inspect the wiring inside, pretty basic the shunt IS the plug no way it can have a solder issue) the plug for the meter is soldered to the shunt, all looked good. I moved the small fuses around inside the shunt that I presume are there to protect the gauge. reinstalled everything, checked over the battery ground (wire from batter post to fire wall) and the engine harness ground (wire from harness that goes to the alt, has a ground lead). All seemed good and clean. checked the wire that goes from the starter positive back to the engine harness, that looks a bit suspect, some rust on that nut, will give that a good wire brush treatment. made sure the E and B connections at the alternator were solid. oh forgot also deoxit the spade connectors of the two fuse links mounted by the VR.

Anyway after the above (but for the last large wire at the starter), took it for a test drive (70 miles, lots of revving, lots of rougher roads to shake it up) I had NO more jumping around +/- like before, and yes the ammeter is working as I will show a drop when loaded momentary until the alt adjust for it back to center.

If it happens again I will take the time to install a lead at the battery terminal to attach a volt meter to. IF it truly was seeing a 60 amp charge I would expect there to be a noticeable increase in battery voltage. Frankly I doubt there was a 60amp charge as I was only turning about 2.5k when I noticed it. I should have thought to push in the clutch and drop the RPM to see what happened, I doubt the alt could make that kind of power to a load at idle, heck it will not do that when testing and the VR is bypassed.

 

The battery is a napa legend 24f less than a year old, test good with a digital tester (not the carbon pile type, the Internal resistance type) resting voltage is about 12.65-12.75. the batter cables looks good, have molded on contacts, no corrosion and are on tight. I really should have gone to that white lead on the positive side of the starter. From what I read, the VR looks at the voltage of the battery (as determined by this connection for the positive) and the body ground (that mid point ground lead I presume for black ground lead of the VR), to establish an overall state of charge, then cycles a zero/med/high field current thru the VRs contact point, one side is high, one side is zero and when off both contacts there is a voltage divider that sets up a med current flow. Pretty neat. But if it "sees" a voltage that is low (presume a poor connection at that starter or ground) then it would assume a low state of charge and attempt to correct by cycling max field current. Still I don't see how it could swing SO much, but that is the nature of intermittent faults. Oh I did a through visual exam of the wire harness looking for any rub thru ect… I am lucky this car is totally unmolested as far as PO work to the electrical I all seems fine and well supported with the correct tie downs etc.

I looked at the plug in connector of the VR to harness, it looked fine, no corrosion was noted, but I still should have hit it with some deoxit. will go back and do that as well, and reinstall the orig VR as a test. I don't the VR was the issue after all.

 

 

 

Edited January 4, 20214 yr by Dave WM

another data point, the OE Nissan green fuse link looks like the plastic heat shrink on the female spade terminals cooked itself off. Ordered some new ones. The wire is still intact but I suspect it got hot...

I have searched on intermittent over follow by under charge then return to normal in google, nothing came up. there was something on shorted windings but it was a blocked site for me.

Edited January 13, 20214 yr by Dave WM

When I get a chance, I'll take another look at the wiring diagram and see if I can draw any conclusions.

I still don't think you were really pushing 60A into anything. That's welding current. 

Those headlights were getting quite bright. I know it was not just a meter twich now.

I would suspect that the voltage regulator was messing up if you were seeing voltage spikes.

The symptoms you described would be along the lines of a loss of sensing (maybe from the VR overheating) causes current injection into the stator of the alternator, kicking up the voltage. However, shortly after that, sensing resumes, the system stops injecting current into the stator, dropping the voltage of the alternator (maybe too much).

Is your charging system stock, or do you have an internally regulated alternator?

Edit: I forgot to add that you could get something like this to get a voltmeter in your car: https://www.amazon.com/gp/product/B07VZFSY4SInsert other media

Edited January 13, 20214 yr by SteveJ

I discounted the external VR as an issue (stock system) since I got the exact same behavior after switching to a backup new VR. I could accept a bad VR even a bad second VR but for the  identical pattern, "ok, then high charge then low charge then return to normal" exhibited by both.

So since I know its not a meter malfunction (much brighter head lights during the charge event) and the behavior is the same regardless of a different VR, that leaves me with 3 options

Bad battery (somehow influencing the charge/discharge cycle).

Wiring (I checked for good ground and good connections thinking the VR's may be sensing a low voltage) but the recurring pattern seems odd for that.

Alternator somehow confusing the VR's, this was the most likely although hard to figure the exact fault. My assumption is the alt was having an intermittent fail, be it loss of internal ground of the diodes, intermittent open of one or more of the diodes, winding faults of the stator or field. Again not exactly how but it seemed the next logical thing to replace.

After replacing the alt I made the 400 mile return leg with out a recurrence of the "charge/discharge/back to ok" cycle. there remains some very small events I noticed at idle but it was a very different event, no pattern and a much smaller scale (amps not 10's of amps flux), and more of a ammeter needle vibration than a sustained 2/3 seconds long event. this event is not constant either, just every now and then when I am sitting at  stop light the ammeter would vibrate a bit. I suspect something is still not 100% correct, I should have just rebuilt the OE part with some bearings and a new diode pack (I kept the OE alt from years ago when it shorted the diode, and I replace the alt back then, did not give it back as core). So the presumed bad alternator that I replaced on the trip was the autozone one with about 5+ years and about 25k miles on it. Maybe that's about what to expect with a reman...

I did order some new OE Nissan green fuse links, since the one in there looks like it got hot.

and after going out to take a ride around the block to test again, I have a FLAT TIRE, thanks to a well place nail, argh...

 

Edited January 13, 20214 yr by Dave WM

Ah, since you seem to have eliminated the VR from potential causes, I would think it's the stator or something related to the stator. 

clamp on ammeter to test with

 

Cool. So are you going to drive around for a while with that attached?

I still haven't looked closely at the wiring diagram to see how things are really connected. I will do that when I get a chance.

Since you kept talking about it, @Captain Obvious, I decided to take the plunge. 😉 While @Dave WM is trying to see if it's just the meter, the description of the behavior during the trip tells me it is the charging system itself acting up. 

By the way, Dave, I like the current monitoring setup. The ammeter I linked would work the same way. As long as the place you clamp the meter feeds the white wire in the wiring harness, you should be good. You may want to get a voltmeter that plugs into your cigarette lighter, too. The lights getting brighter tells me your voltage is going up during the issue. I'm betting you'll see the voltage go up when the ammeter says you have current flowing to the battery. The ammeter behavior seen in your video looked typical to me. You have the voltage drop from starting the car, so the alternator is replacing the battery's loss. You didn't have to crank long, so it doesn't take long for the battery to get back to a sufficient level, dropping the alternator output.

Let's consider the operation of the VR as laid out in the wiring diagram.

As the neutral of the alternator builds sufficient voltage, it energizes its coil in the VR.
The contact that closes provides a ground for the second coil that is powered by the battery. Note that this second coil is wired in series with a resistor, probably to drop off the voltage and cause the coil to de-energize should battery voltage drop from discharging.
When coil 2 is de-energized, the switched voltage is going straight to the field wire on the alternator (full field voltage).
When coil 2 is energized, the switched voltage goes to two resistors (voltage divider) to reduce the voltage going to the alternator field, dropping the alternator output.
From reviewing http://www.secondchancegarage.com/public/83.cfm it's obvious for the mechanical VR the contacts would constantly open and close. With a solid state VR, you have discrete components handling the same function in a similar manner.

Here's the thing. If the VR is not dropping the voltage going to the field wire, you're going to see increased output from the alternator. Keep in mind that a flaky ground at the VR could present itself in a similar fashion. I am assuming you're running a solid-state VR. Also, a flaky neutral could also cause the problem. It could also be a winding issue in the alternator if the grounded leg has a winding shorting out close to the neutral. (I don't think you would have an intermittent short on a winding, though. I don't have enough practical experience to say one way or the other.)

Ideally, it would be great to monitor the voltage on the white/black wire with an oscilloscope. You would expect it to fluctuate too much for a typical voltmeter. Monitoring the voltage to ground on the neutral might reveal good data, too.

Okay, now I need to find all of my O-scope probes so I can see if I can configure something for an educational video.

 

13 hours ago, Captain Obvious said:

Cool. So are you going to drive around for a while with that attached?

I still haven't looked closely at the wiring diagram to see how things are really connected. I will do that when I get a chance.

yes that is the plan I just have to make a better hookup to the meter. I was also going to try and simulate the intermittent by poking around the wiring harness, tapping the VR etc to see if I can locate. This setup has a better resolution and I can see it while poking under the hood.

12 hours ago, SteveJ said:

Since you kept talking about it, @Captain Obvious, I decided to take the plunge. 😉 While @Dave WM is trying to see if it's just the meter, the description of the behavior during the trip tells me it is the charging system itself acting up. 

By the way, Dave, I like the current monitoring setup. The ammeter I linked would work the same way. As long as the place you clamp the meter feeds the white wire in the wiring harness, you should be good. You may want to get a voltmeter that plugs into your cigarette lighter, too. The lights getting brighter tells me your voltage is going up during the issue. I'm betting you'll see the voltage go up when the ammeter says you have current flowing to the battery. The ammeter behavior seen in your video looked typical to me. You have the voltage drop from starting the car, so the alternator is replacing the battery's loss. You didn't have to crank long, so it doesn't take long for the battery to get back to a sufficient level, dropping the alternator output.

Let's consider the operation of the VR as laid out in the wiring diagram.

As the neutral of the alternator builds sufficient voltage, it energizes its coil in the VR.
The contact that closes provides a ground for the second coil that is powered by the battery. Note that this second coil is wired in series with a resistor, probably to drop off the voltage and cause the coil to de-energize should battery voltage drop from discharging.
When coil 2 is de-energized, the switched voltage is going straight to the field wire on the alternator (full field voltage).
When coil 2 is energized, the switched voltage goes to two resistors (voltage divider) to reduce the voltage going to the alternator field, dropping the alternator output.
From reviewing http://www.secondchancegarage.com/public/83.cfm it's obvious for the mechanical VR the contacts would constantly open and close. With a solid state VR, you have discrete components handling the same function in a similar manner.

Here's the thing. If the VR is not dropping the voltage going to the field wire, you're going to see increased output from the alternator. Keep in mind that a flaky ground at the VR could present itself in a similar fashion. I am assuming you're running a solid-state VR. Also, a flaky neutral could also cause the problem. It could also be a winding issue in the alternator if the grounded leg has a winding shorting out close to the neutral. (I don't think you would have an intermittent short on a winding, though. I don't have enough practical experience to say one way or the other.)

Ideally, it would be great to monitor the voltage on the white/black wire with an oscilloscope. You would expect it to fluctuate too much for a typical voltmeter. Monitoring the voltage to ground on the neutral might reveal good data, too.

Okay, now I need to find all of my O-scope probes so I can see if I can configure something for an educational video.

 

its a mechanical VR, I have not had a recurrence of the orig problem since replacing the alt. I wish I had kept the core to evaluate. Now the only issue is an occasional flux, happens at idle and after the car has warmed up some and there is a load beyond just the EFI, like the fan or light. No more wild swings that last or flare up of voltages (head lights flare up) since the change.

My SWAG is after warm up my idle is less stable, prob due to leaner mix, I have it set up pretty darn lean (pulling the dip stick can effect the idle, remove the oil filler and it dies in a second). I may richen up the idle circuit just so its a bit smoother sitting at a traffic light. But anyway the theory is, unstable idle means unstable alt speed, add a bit of a load and the VR will have a time switching the field current quickly enough. The ammeter needle does seem to go with the idle changes. I am slowly noticing these things as I generally don't sit and stare at the ammeter, but of course since the orig problem (flaring up lights etc...) I have become much more aware of it.

I suspect the alt was the whole problem, I will be reinstalling the OE hitachi VR now just to confirm it was never the issue.

Edited January 23, 20214 yr by Dave WM