Electrical voltage issues and missing

[Earendil]

I know I have one bad component in my system, the battery. But using an analog voltmeter I have found some other behavior that I'd love an opinion on.

First, the battery: After a good long run, and then turning the car off, I get 11.8 volts on the battery (Edit: I'm now seeing 12.3 after after 10 minutes being off). I believe a good battery should be up around 12.6-12.8? This battery is barely a year old (May of last year) and still under warranty. I'll look into getting it replaced, but before I do I want to know if anyone thinks I have a separate issue.

1. On starter engagement system voltage drops to 9v (battery issue).
2. With the car running and at idle, system voltage is 13.8
3. While at idle, putting the turn signal on causes a 0.3v drop and rise in the system with every flash.
4. While at idle, pressing the brake pedal causes a momentary 0.3v drop in the system.
5. Once the car is warm (or after it's run for a while), and while it's at idle, the car will see a drop in voltage of almost 0.8v every 4-8 seconds, which correlates with no external input. Simultaneous with the drop in voltage is engine roughness (sounds like missing) and a change in tone of the fuel pump. It will recover after a half second, and the engine doesn't lose more than maybe 25rpm from a steady 600rpm (a needles width).
6. If I bump the RPM up to 1500 and hold it there, I get steady missing once every couple seconds. This however does not seem to correlate with a change in voltage.
7. Reving the engine up to 2500rpm does not cause the voltage to go up. In fact it consistently drops from 13.84 to 13.77. I thought above 1500 the voltage would rise?

What do you guys think? Two separate issues, battery and an occasional miss? Or is something wrong with the system that would cause a premature battery failure and be giving the ignition system a rough time?

I might replace the voltage regulator just because I don't recall when that was done last. However my experience with a bad regulator is that the system voltage just runs high and causes gremlins. The fuel pump was tested according to the Bentley's procedure for current draw and passed. I didn't do a flow test. Cap, rotor, and spark plugs don't have more than 200 miles on them. I'd rather not replace the battery until I know it won't be immediately KIA.

Thanks guys!

Edit: I'm going to go poke at it some more. I'll update this post as I find more potential oddities.

[lcoleman]

One diagnostic step that you haven't done is to check for a/c voltage in the system "upstream" of the battery (ie directly at alt output, using alt case as the ground reference). The regulator could in theory be dying, but I'd say it's unlikely. What you describe sounds pretty normal for the charging system and the battery except for #5. I suspect whatever causes that voltage drop is your culprit, as .8V is pretty substantial.

[jlevie]

The drop in voltage in #5 is probably from the drop in engine rpm that accompanies the misfires.

Have the battery tested.

[Earendil]

I did a little more testing last night, and discovered the following:

1. Regulator is within spec. Bentley calls for replacing at 5mm left on the contacts, and my contacts are at 9.1 and 7.1mm out of a possible 12mm. I'll order one and have it standing by.
   
2. While pulling the regulator, I noticed the wires running to the alternator were experiencing some hard times. I'm going to pick up some new ring terminals today, cut back the wire, and redo it with some shrink tube.
   
3. I cleaned up a few ground points, but found nothing that would make me suspect they weren't functioning at 100%.
   
4. Based on another thread, I measured the voltage difference between the chassis and the engine. I found a 0.08mv difference without accessories, and a .20mv difference with. This was repeatable. Note that the voltage was higher on the chassis side, specifically I was using the shock tower ground point. So I guess the engine is grounded, and grounded better than the wires to the shock tower? Or perhaps that is too small a quantity to matter, and it's simply caused by normal resistance.
   
5. I tested my coil and measured .6 ohms and 5500 ohms. The bentley only says that it should be .5 ohms and 5000 ohms, without giving any sort of acceptable range. The used spare coil I have measured at .55 ohms and 5650 ohms. I might toss it in and see if I get any behavior differences, since I know that the coil test doesn't prove a coil is good, only that it is bad.
   
6. Since the fluctuation seems to happen more once warm, I checked the CTS resistance at the ECU plug, and found it to be within spec when the engine is warm.
   
7. I checked the harness connection for the injectors, and didn't see any corrosion. I shook the harness with the engine running and observed no change in behavior. I'm not sure how else to check this one.
   
8. I used my amp meter, and with the car completely off I tested the amp draw from the battery. This ended up being zero. So we aren't shorting something out on the always-on list, which admittedly is a small percentage of possible short locations.

One diagnostic step that you haven't done is to check for a/c voltage in the system "upstream" of the battery (ie directly at alt output, using alt case as the ground reference). The regulator could in theory be dying, but I'd say it's unlikely. What you describe sounds pretty normal for the charging system and the battery except for #5. I suspect whatever causes that voltage drop is your culprit, as .8V is pretty substantial.

I can check the Alternator specifically, but if all ground points are solid, it should read exactly the same as the battery, shouldn't it? There is nothing but copper wire between the two, no component that would cause a difference or delay in voltage. At least that has always been my understanding.

0.8v is an estimate, since I can only really see the change with an analog meter. When using my digital one the best I can see is a 0.4v drop, but the refresh is ever half second so it's hard to tell.

The drop in voltage in #5 is probably from the drop in engine rpm that accompanies the misfires.

Have the battery tested.

If the rpm drop is only an RPM needle tick, and starting from 600-650, would that really cause a noticeable drop in voltage? I'll try and get down and have the battery tested today. If I end up with a new battery I'll run through some of these tests again and see which measurements change.

[lcoleman]

If the regulator isn't properly rectifying the a/c voltage from the alternator, some can "bleed" through the system. Since lead-acid batteries work great at absorbing a/c current, you should check it right at the alternator. This is a shot in the dark, but worth the simple effort of checking imo.

I'm not a big believer in parts store battery testers, but I would make it a point to follow jlevie's advice before anyone else's.

[Earendil]

Either I have electrical gremlins that are playing hide and seek, or I've resolved the issue.
I couldn't make it out to get the battery checked last night before they closed, so I spent a little time redoing the wires to the AC, and buffing up the battery terminals and battery ground. For the AC, I trimmed back the wires, put new ring terminals on the positive and negative wires, and replaced the ground wire that goes from the AC to the block.

I took the car out and got it warm, came home and let it idle. I experienced no fluctuations in voltage, and no missing that caused the tach needle to move. I still have a slight miss that I've had for a long time, but nothing as bad as it was. So for now, I guess this is resolved. I'm also getting 13.78v at the battery, AC unit, and Sig lighter.

Conclusion: Having solid but high resistance wiring at the AC unit can cause some weird issues.

Thanks everyone for your advice and letting me talk this one out as I trouble shot it! :)
Cheers!


Parting shot:
After parking, my battery measured a more happy 12.7v. under load it only dropped to 12.55v or so, which means it might be okay. I didn't (and want to) check it under the load of the starter, as last time it dropped to a measly 9v.

While checking voltage at the AC unit, I decided to test the volt difference between the engine block and the chassis on that side. I got the same 0.20mv difference as I did on the passenger side, with the chassis registering as the positive side. This still baffles me but I don't know if it's a symptom of anything.