Can't find the problem. What would you do? Replace VR?
 

Ok... I need some of you "Click and Clack" types to help me with this puzzler.

I've been chasing a problem for months and I cannot find the smoking gun.

I'll try to recap the events as best I can.

About 9 months ago (last spring) I noticed the headlights would not come on full power on the way home one night. Brights were fine. Just the dims. Dash lights flickered a bit on the drive home. Put the battery (Optima Red Top) on the battery minder. Problem went away. Also JWest headlight relays are installed. Turn signal stalk was replaced 2 years ago.

Some time later, same problem. Tried swapping relays. No affect.
Battery minder again, problem went away.

Series of little things happened that I didn't think we're related and the time but now appear to be.

- stereo shut itself off
- parking brake light randomly came on (parking brake NOT on)

These things were random over the course of weeks and months. It wasn't until I read another thread on here that it occurred to me they could be related.

And then my CDI box failed.

After I replaced the CDI and coil I backed the car out of the garage at night, shut the garage door and turned on the headlights. The light output varied (surged) with RPMs.

So I'm thinking the VR is shot. Pull the alternator and discover I have a Valeo and not the original Bosch. Great. Didn't look very old and I have no record of when it was replaced. Took the alternator to an old-school rebuilder recommended by a local Pelican and explain what's going on. Seems like a great shop but they can't find any problem with alternator or VR and suggest maybe it's the battery.

Take the battery to the local FLAPS and get it tested. They say it tests clean. I bought a new yellow top because the red was 2008 vintage and I figure it's just time.

So... No smoking gun found. Wiring looks good. Grounds look good. I just can't seem to find anything wrong.

Before I put the alternator back in I'm wondering if I should just replace the VR anyway for peace of mind?

What would you guys do next? I'm stumped...

Have you installed and tested the yellow top?

I had last year a problem with the Varley Red top in a race car. You could charge it, it would hold charge but during the race it would short cicruit. Everyone told me alternator, but I replaced the battery and all was well. The battery seller claimed that the battery was fine and even suggested that buying a new battery was a waste of money....

I have the impression that batteries are not very transparent when it comes to internal problems...I beleive this is one of the reasons why you see so much effort in hybrid and electrical cars put into wiring each bank of batteries and cells independently...

Give it a try.

I would not put too much weigh on having the battery tested at a chain store. When I took my battery to Auto Zone they put it on a charger for a couple of hours and it "tested" OK because it kept the charge.
I told them that the fully charged battery would go flat after a few days. They "tested" again and found the battery would not charge past 52%, no matter how long they kept it on the charger (I had to insist they let it on overnight).
My battery was effectively dead...

Did you check the connection on the starter the red wire that goes to the alternator.

I don't see anywhere in your post that you've monitored the voltage out put.

^ +1 - but I bet it's your VR.

Hook up a DVM to your cigarette lighter jack - or somewhere else convenient that can be monitored from the driver position. I've used an old cell phone plug modified with wires that hand out for easy access.1

zippy,

Roger that. I guess I was just hoping for that smoking gun. Would a bad battery throw an intermittent voltage spike?

pors1968,

Good call. I will double-check.

Pete,

Good catch. I posted some numbers in another thread. I'll go dig them up. Voltage with car off, idle and at 2k RPM was fine. I could never catch a spike in the act but I had many of the same symptoms as a bad VR.

The voltage reg can fail intermittently, mine did. I would rig a temporary volt meter in the interior so you can watch what is happening when you drive.

2nd the cheap voltmeter so you can monitor what's going on. The one below worked well for me. Voltage at the lighter socket may be a little less than at the battery but I found the unit to at least be consistent.

I had the tell tale signs of a spike once. Radio blinked off, seat belt light came on. I bought a gauge and started monitoring voltage output. Took another six months to show a spike. One thing I noticed in hind sight was the old alternator/regulator would show wild swings. 12.5 - 14.0 - 13.0 etc. Constant fluctuations of a volt or more. Once I bit the bullet and replaced the alternator/regulator with a new Bosch unit I found that voltage fluctuations would be in tenths or hundreths of a volt not whole volts. Yes voltage still varies based on load and output but the wild, constant fluctuations stopped. -J


ETA - When they actually have them in stock Amazon sells new(not reman) Bosch alternators for $250-$300. That's about half what they sell for at most other vendors.

INNOVA 3721 Battery and Charging System Monitor : Amazon.com : Automotive

http://ecx.images-amazon.com/images/I/4105fQmT1vL.jpg

Somehow I missed your post earlier. I was reading from my phone so maybe I missed it.

I just picked up the battery today and haven't had a chance to install it yet. I'm going to get a DVM for the cig lighter as others have suggested. I'm just wondering if I should trust the alternator/VR before putting it back in.

Open the alternator and look for burned insulators, stator winding standoff insulators.

LJ851 & JAR0023,

I've got that Innova DVM on order. And good to know on the Bosch from Amazon! I might just do that. The guy at the alternator shop did not speak highly of the Valeos which is consistent with just about everything I've read here.

Any idea how to separate the fan from the alternator?

Just check the battery posts with a multimeter. You should have around 13v. I'd also check and clean your battery connections, the starter motor connections, grounds, etc. If other electrical accessories are going strange, (windows dont work, car won't shut off) then you can have a bad ignition switch.

Separating fan from alternator. I took a tip I found in a thread here. There is a circle of 5 or 6mm holes around the hub of the fan. Tap two opposite each other and use screws to press the fan off the alternator shaft. IIRC the size of the holes is such you get a half depth thread. Very easy to tap and still strong enough to press the fan off. -J

Tap gently around the edge with rubber mallet and should drop down enough to get the key out of the shaft.


Sent from my iPhone using Tapatalk

The old battery checked out fine with a multimeter.

The ignition switch and the turn signal stalk were the first things I fixed when I bought the car in 2011. Also added an in-line fuse to the instrument panel circuit.

Multimeter will not detect bad cell in battery, only the voltage of the good cells. Need to charge battery fully and put on a load bank battery tester.

I just replaced my VR with whatever is spec'ed for an 87 Carrera. I'm still getting 1/4 to 1/3 volt voltage swings. Is the new VR crap?

Did you check all the grounds? Particularly the braided engine to body one underneath?

Wire is pretty cheap, I'd be tempted to replace the one I mentioned and disconnect and celiac all the others. If the shop says the alt is ok, and the batt is holding a charge, I'd re think connections and wires.

Kinsley,

I am going through the wiring. While I have the alternator out I replaced the original ceramic fuse panel in the rear with one of Fred Cook's 4-fuse ATO panels so I have taken the opportunity to clean up the wire ends as I crimped new connectors.

What I can't wrap my head around is if it is the wiring, what would be common between the CDI, parking brake light, radio and headlights? And would something like a bad ground cause a random voltage spike? That just seems more like battery/alternator/VR related to me.

If I get it all back together with a new battery and still have the same problem I'm going to be really frustrated...

One data point, when I did my MS2 upgrade to MS and EDIS, I was seeing my voltage jump up and down by a full volt, and fairly randomly. Tunerstudio allows you to see voltage real time and log it. I changed the Negative terminal to body and the engine to body grounds, cleaned up the starter connections and replaced the starter to alternator wire and the alternator ground. Not sure which one was the culprit, but my voltage is rock steady now. For sure the alternator to starter and alternator ground wires were pretty stiff and suspect, also the braided strap between the body and engine was in pretty rough shape.

If you want to rule out your alternator, I have a newly rebuilt one - internal regulator - I'm not using (I'm in Alpharetta).

Kinsley,

That makes sense. If the shop I took it to can't find a problem with the alternator itself then the difference has to be the wiring. I will replace the starter to alternator wire and the ground. I'll add replacing the chassis ground wire to my engine drop to-do list.

Thanks for the offer on the alternator. Let me change the wiring and see what happens. I'm going to open up the alternator and check the insulation and windings as wwest suggested before I put it back in.

I think Dennis (timmy2) makes a new engine harness. I might look into that too... you know... "while I'm in there"...

Check the ground scheme per factory,..rework if necc'y. ALWAYS change out VR's with Alts,.....

I'd keep special eye on the battery.

Connection (all) are important.

Maybe you can ohm out some those harnesses?

What a pain! Best to you!

Doyle

I recently experienced an intermittent short in a 2 year old battery. My 964 started missing a bit at lower engine speeds, and then it would just randomly die while idling. Battery voltage was in the high 13s so I checked all the normal tune up items to no avail until the battery suddenly went dead and would not accept a charge from any of my three battery chargers. The parts store warranted the battery, and amazingly enough all of the poor running went away. The car idles dead smooth and pulls smoothly and evenly at all speeds!

Finally got back home from a business trip and got back to working on the car. I'm going through the wiring -- cleaning connectors. Only thing I've found so far is it looks like someone added an extra ground wire from the alternator directly to the engine case behind the alternator vane shroud. This is in addition to the bundle of ground wires coming through the harness that passes through the rubber grommet.

I don't think this is a problem but I wanted to confirm. Is this extra ground wire supposed to be there?

Yes, that extra wire is normal and won't cause problems.

I don't know whether this relates to your VR problem, Tim. On my '82 Landcruiser the generator with an attached VR has problems charging when the outside temperatures are cold. I can tell from the voltage meter on the dashboard that the voltage does not reach the normal 13.8V when the engine is running, especially when the headlights are on at low outside temperatures, or the blinkers are operating. Cold temperature voltage could be as low as 12.0 Volts with the engine running. The voltmeter always shows a steady 13.8 Volts under any conditions when the temperatures are warm. Cold temperatures here in Southern California are high 40's. I do blame the diodes in the voltage regulator for the problem. But I also had similar problems when the alternator brushes were used up. That was not temperature dependent though. Replacements fixed that problem.

That engine to body braided ground can cause all kind of ghostly faults. It could be on rock solid but somehow it loose ground . Have to take it off clean up the strap and sand down both ground points..

Along with cleaning and checking all the other wire ground points, have you tested the blue wire to the alternator for continuity at the 14 pin plug and from there to the back of the alternator light in the gauge?
Should be very low ohms. One thing to try is connect a meter set on ohms to the gauge blue wire and the blue wire lead at the alternator and then with it plugged in, wiggle the 14 pin connector at the fuse panel. If the ohm value moves around you may have found the culprit.
If the D+(blue) wire is seeing a current change due to ohmic variation it will affect the voltage output of the alternator. You should clean the 14 pin connector pins and sockets anyways, as well as spread the pins on the panel side with a razor blade.
All assuming your VR is good.... :)

Roger that.

Is there a reliable way to test the quality of the ground strap? Would a simple resistance measurement using a DVM really tell you that much? If there was a ton of corrosion where the resistance was exceptionally high I'm sure it would but I'm thinking there is probably a range where it might not be completely obvious until it crosses the magic threshold.

Timmy 2 may have your solution. Check the blue wire. I had similar problems due to inconsistent charging. Replaced VR. Probably needed a new one but did not fix the problem. Finally traced to the blue wire in the 14 pin connector. Blue wire controls the exciter voltage to the alternator. Without a good connection you will not charge, then the battery loses voltage though you can run for a while with just the battery. At about 10.5 volts the CDI cuts out. Also stereo will shut off at low voltage. I had cracks in the 14 pin connector housing so I finally replaced the connector and all is well.

I was about to respond to Dennis (timmy2) and something else came up. The blue wire check is a good idea. I will definitely check it. Even if it's fine then at least I've eliminated one more source of doubt.

I did a cursory cleaning of the 14-pin connector when I was hooking up the O2 sensor recently. I used an Emory board on the male connectors to lightly sand them. I'd feel better if I could file the female connectors a bit to remove any surface oxidation. Anyone have any tips on how to clean them?

maybe i missed it, but how do you know you had/have a voltage "spike"?
i dont think optima batteries are what they use to be. i think they are made in mexico.
i think i would also take up the offer on another alt. i have a new alt with a valeo VR that i dont need anymore, but, i broke one of the brushes which you could pull off yours.

one at a time, pull the wires out of the fuse block. clean the wire ends along with the end of the screw.
you have a very tough problem. sometimes its just eliminating even the stupid stuff

Documented in my first post of this thread. I have never been able to catch it in the act but I have had several weird things that are consistent with a voltage spike. Radio shuts off (5-6 blinks and then powers off), headlights not at full power at times and then brightness varying with RPMs, CDI failure, parking brake light randomly illuminating. These seem to all be things people report as being consistent with an intermittently failing voltage regulator. The infrequency makes this maddening to find. Jiggle something or clean something to think you might have fixed it only to have it come back later. And then there is the possibility that there may be in fact be more than one problem. I just want to find that one thing or the part that clearly tests bad -- the frayed wire or the loose ground -- that makes this one go away!

I've got to get the fan off the alternator and take a peek at the innards. I may take you up on your offer!

Your logic correctly points to your VR being bad. It's common problem with that
alternator. You should consider ordering one from Pelican Parts, but make sure you
request the OEM Valeo (Paris Rhone) one. If it's not in their stock, they can easily
order you one. Again, always ask a sales person if you don't find it listed in the catalog.

Hi Tim,
You can order parts and replace with possibly good results. Or, you can test and make sure you find the problem. (I'm naturally cheap at heart and starting to sound like Tony...):)

To test your engine ground strap for resistance just connect the DVM to the end it is attached to on the Alternator and the other probe to the case of the engine near where it is bolted on. You should scratch the probe into the metal a little to get past any gunk and oxidation of the engine case. Should be very low reading basically the same as when the 2 probes are touched together, maybe 1 ohm at most.

To clean the 14 pin female connector sockets, a bit of vinegar on a Q-tip should remove any oxidation. Rinse clean with water. Try spreading the male contact pins with a razor blade. They have an X pattern that can be slightly (gently) spread out. Apply a little no-ox to the posts and the connection should be good for a while.

Then check the blue excitation wire to eliminate it as a problem.

The problem is that those regulators are typically intermittent, which is not easily troubleshot.

Can't find the problem. What would you do? Replace VR?
 

Maybe, maybe not, I can't tell if it is, or isn't the problem while sitting hundreds of miles away at my computer...

Methodical, analytical troubleshooting to eliminate all possible causes to a ...

Edit:
Never mind, just saw the connection to Lorenfb in another post...

I've recommended your company to others for your services, play nice...

i dont know if i am being to critical of your description, but it does not sound like a "spike". your problem makes me think of a bad solder joint. maybe something internal to the VR or a soldered wire or diode (dont know if one diode would do this) that may be dropping the voltage perhaps due to vibration. thats why i would swap out the entire alt/VR. (my VR has a broken brush.)

testing resistance with a meter does not really give you a definitive result. the meter does not provide enough current to really test the connection. i have seen this problem on pelican several times:
someone has a light that does not light. they check it with a MM and all is good. they tare the car apart trying to find the answer. turns out to be a connection that is good enough for the meter to show as good but not good enough to pass the current required to light the light bulb.
(this is why i use a test light for electrical work).