Alternator not charging

Tiny
CSIGONA
  • MEMBER
  • 1991 DODGE DAKOTA
  • 3.9L
  • V6
  • RWD
  • AUTOMATIC
  • 226,000 MILES
Sorry for the long story. I have done quite a bit of testing and analysis so I have a lot to say. I am also sorry for using my own terminology. I just do not know the right words.

The alternator stopped charging the battery. I checked the connectors for corrosion, cleaned them up, and hooked them back up. No change.

I checked the field coil connectors. The side that's supposed to carry battery voltage (I will call it F+) reads close to battery voltage whenever the ignition key is set to run. The voltage on the other connector (I will call it F-), the one that goes to a green wire which ends up at pin 20 of the ECM, reads about the same. When I short F- to ground, the alternator charges, sometimes as high as 18 volts. I presume this means the alternator is good, but I am not certain. I followed the green wire back to pin 20 and measured the voltage drop between F- and pin 20 at effectively 0, so it checks fine for continuity.

Here is the questionable stuff. If I measure the voltage drop between F+ and F- when the block is not mounted to the alternator, it's close to battery voltage. This suggests that it should have told the alternator to charge since F- is close to 0. However, if I put the block back on the alternator, F- equals F+, that is, the voltage drop across F+ F- goes away. F- now also measures around battery voltage, just like F+.
As impossible as it sounds, it's as if the circuitry in the alternator un-grounds F-. Does this not sound screwy to you?

I am running the truck now for short trips by starting up the engine, shorting F- to ground to have the alternator charge for a minute or two, enough to resupply the battery power drained by the starter. Then I disconnect the ground clip and drive away, running on battery alone. When I get to my destination, I leave the engine running, short F- to ground again, charge the battery for another minute or two, then turn off the engine. I am sure this is not good for the electronics or the battery, but this routine has kept me mobile for a few days.

Thanks for putting up with me.
Do you
have the same problem?
Yes
No
Wednesday, July 19th, 2017 AT 4:41 PM

3 Replies

Tiny
CARADIODOC
  • EXPERT
As a former Automotive Electrical instructor, it is obvious you know plenty about electrical theory, but I think you are getting lost by not knowing how this circuit works. Your tests made sense up to, "here is the questionable stuff".

You have full battery voltage on one field terminal. On older models that came through the ignition switch. On newer ones it comes through the automatic shutdown, (ASD) relay. That turns on when the engine is rotating, (cranking or running). The other field terminal goes to ground through the voltage regulator in the Engine Computer. The lower the voltage on that second terminal, the larger the voltage drop across the field winding, and that means a larger electromagnet. The larger electromagnet means more voltage is induced in the stator winding, which means more current for the electrical system. Think of the voltage regulator as a variable resistor.

There is a separate voltage sensing wire at the engine computer to tell the regulator what system voltage is at. Where you got derailed is you performed the definitive test, did not recognize the results, and kept looking. By grounding the green control wire, you proved the alternator's output circuit is okay, its input circuit is okay, and all the wiring is okay, up to the point where you grounded that green wire. You still could have a break in the green wire after the point where you grounded it. Most commonly that would be an open voltage regulator, but it also could be a bad connection between terminal 20 on the computer, and its mating pin in the connector, or a break in the green wire after the point where you grounded it.

The clue to the break in the control side of the circuit is you found the same voltage on both field terminals. That means there's no voltage drop across the field winding, and that means there is no current flow through it.

Here is the fastest way to arrive at your diagnosis. You have battery voltage on one field terminal. The other one will have 0 volts, battery voltage, or something in between. O volts can only be caused by an open brush. Full battery voltage, (exactly the same voltage on both terminals), means no current is flowing, and that can only be caused by a break after the second field terminal. Something in between, (typically 4 - 11 volts), means some voltage is being dropped across the field winding, so field current has to be flowing, meaning the circuit is working properly. If there's a charging problem with 4 - 11 volts on the second field terminal, the cause is in the output circuit of the alternator.

When you ground the green wire, then see system voltage rise, and/or the head lights get brighter, there is no need to do any voltage or continuity testing. Everything is working up to the point where you grounded the green wire. That is called "full-fielding", and is a standard part of our testing. The problem is when you want to do that as a preliminary test so you can give your customer a rough estimate to find the cause of the problem, full-fielding takes all of five to ten seconds, if you know which field terminal to ground. In your story you found the exact same voltage on both terminals, and the green and 12 volt wires go through a black plastic block, so there is no way to know which wire goes to which terminal.

This is the dilemma I left my students with. Except for my "bugged" cars, that green wire is buried in the harness and is not easy to find. My students were never ever allowed to pierce a wire to take a reading, and just because you might know where the connector is located that makes a convenient test point, it will be different on other models even though the same circuit was used on all Chrysler vehicles. If you ground the right terminal at the alternator, you got lucky and you proved the alternator was okay. If you grounded the wrong terminal, you would blow a fuse, or, in the case of our 1997 Dakota, burn open a copper trace inside the engine computer that yours truly had to repair.

I will share the rest of the story next time. You cannot tell which terminal to ground with a test light, and you cannot tell with a voltmeter, but that is all you have to work with. And you will not find the information or the solution in any service manual. "You have to be as smart as I are"!
Was this
answer
helpful?
Yes
No
+1
Wednesday, July 19th, 2017 AT 8:28 PM
Tiny
CSIGONA
  • MEMBER
Thanks for the comprehensive reply.

After fiddling with this problem for some days now, the alternator started charging again this evening. It now works in the normal manner (without my jumper wire) and the check engine light went out. This suggests a bad connection somewhere, and I'm sure the problem will return if I don't find the bad spot. I am sure the green wire is good for its entire length. I will keep looking for the problem, but until I do fix it, I'll be sure to keep my little jumper wire in the truck with me, just in case I need it to kick the alternator into charging.

Let me try to answer the quiz you put to your students, namely, how can one tell which of the two field connectors is the one you can't ground (because it might blow a copper trace on the ECM board), and which is the one you can ground (the one which goes to pin 20, which regulates the voltage). I would remove the plastic block from the alternator and test each of the connectors on the block for battery voltage. The one which you can't ground is the one with battery voltage; the one which you can ground is the one that's much lower in voltage. You can't test them while they remain connected to the alternator since the field wire, in effect, connects the two and provides little resistance to DC. I would also bet that it makes no difference which of the two field connectors go to which of the two field terminals. That is, if I jumpered battery + to one of the field terminals and grounded the other, the alternator would run in full-field; and if I reversed them, the alternator would still run in full-field. The difference, and the potential for damage, is not on the alternator side, but on the ECM side of the wires.

Let me know my grade, professor.
Was this
answer
helpful?
Yes
No
Thursday, July 20th, 2017 AT 7:05 PM
Tiny
CARADIODOC
  • EXPERT
Your understanding of the system is on par with students who have already learned basic electrical theory, and have had this circuit explained. You can't figure it out by just looking at the wiring diagram.

You are correct about switching the two field terminals. This goes back to when Chrysler switched to an electronic voltage regulator in 1970. The two wires plugged in individually, and could be switched. Instead of the electromagnet developing a rotating north - south - north - south.... Magnetic field, you get a south - north - south - north... Magnetic field, meaning the same thing.

You are also partly correct about identifying the field terminals with the terminal block removed. One will have full battery voltage, either with the ignition switch turned on for older vehicles, or during cranking or running on newer vehicles. There will be 0 volts on the other terminal. The voltage you find on that second terminal comes through the field winding from the 12-volt terminal. The problem with this is it takes way too long. When it's our customers' time and money that's involved, and especially up here in Wisconsin where they throw a pound of salt on an ounce of snow, those tiny terminals get real rusty, and it's common for one or both of them to snap off. Now you have a lot more work to do. Even on a good day, it takes too much time to take that block off relative to the small amount of information you're going to get.

I went one step further with my students and gave them another set of clues. A digital voltmeter is powered by a battery inside it. For all practical purposes it is an open circuit, and draws no current from the circuit. It is measuring electrical pressure, just like a pressure gauge on a compressed air system. No air flows through the pressure gauge for it to do its job. For this sad story, lets say you measure 12.60 volts on one field terminal. You'll measure exactly the same 12.60 volts on the other one too. No current flows through the meter, so there's no current flowing through the field coil, therefore no voltage drop across it. You start with 12.60 volts, drop 0.00 volts, so you end up with 12.60 volts. No way to tell which is which.

As a side note that gets included a few fays later, system voltage is fluctuating thanks to the current being drawn from the battery to run the electrical system. Pulses of current for the injectors, ignition coils, solenoids turning on and off, etc, can actually show up as the feed terminal has 12.60 volts, and the second terminal has something a little higher at the time you took the reading. Wouldn't THAT add a pile of confusion to your already hurting head?

Now consider the test light. It DOES operate by according to the current flowing through it. On the feed terminal it sees 12.60 volts. When on the second terminal, a trickle of current flows through the test light AND through the field winding. That tiny current causes a tiny voltage drop, so you might see 12 50 volts on the second terminal. That.10 volts difference means the light won't be as bright, but can you see the difference.10 volts makes? You have to observe the brightness of the light at one terminal, remove the probe, move it over to the other terminal, then observe the brightness there. Your eyes are still seeing stars from the first test. The only way this might work is if you tied each terminal to a SPDT switch with the test light on the common terminal. That would let you switch back and forth instantly, then you MIGHT see the difference. The next problem is the higher-quality test lights use bulbs that draw very little current, so this method loses even more validity.

To exaggerate this for clarity, you might connect a head light bulb instead of a test light. A full-fielded field winding draws close to three amps. A low-beam 9004 bulb draws about five amps. Without even resorting to Ohm's Law, that shows the bulb has less resistance than the field winding, and it will cause enough current to flow to cause a substantial voltage drop across the field winding. You might also be sharp enough to observe that when doing this to the correct terminal, the current flowing through the field winding in enough to make the alternator charge enough that you'd see it as higher system voltage or the head lights on the truck got brighter. Can't get any better than that, ... If you have a spare head light bulb in your back pocket and the clip leads to use with it.

But there's still a faster way. You have an extremely accurate voltmeter that will show you a change of a few hundredths of a volt, but it won't cause that voltage drop that you're looking for. You have a test light that will cause a voltage drop, but our eyes are not accurate enough to see the difference between the two terminals. At this point in the class discussion, a long and thoughtful pause was in order. One by one the "ahh"s would show up as the light bulbs over their heads began to shine. Use both of them. Put the voltmeter probe on one terminal. You'll see 12.60 volts. Leave it there, then put the test light on it at the same time. If the voltmeter stays at 12.60, it's the feed terminal. If you're on the control terminal, the voltage will drop when you touch the test light to it.

See? I'm not as dumb as I look!

My technical college was one of three remote class sites Chrysler used for their training, and in gratitude, they donated us a lot of cars. Some were used to test new assembly lines and were never meant to be sold to the public. Some, like our '97 Dakota, were used by the instructors to travel to our school, and to use for training. They were poked and prodded on so much that they also were not to be sold, but they could be donated for training. We also had a number of Dodge Shadows that had the same wiring as your truck. That's why I know it as well as I do. I spent years designing and building bugs for my kids to diagnose, and they all mirrored real-life problems that I had found previously. The common one had a "customer's complaint" of "battery goes dead while driving. Problem is intermittent". Gee; does that sound familiar?

In this case the green wire going to the voltage regulator in the Engine Computer ran through a connector between the engine and the body, and that was a perfect place to find corroded terminals. For my "bug", it was a stretched female terminal that wasn't making contact. Hidden in the harness were two wires running to the switch I used to activate the defect or to bypass it. Every car had at least five charging system defects, and the kids found out that by the time they were searching for the third one, they didn't have to look at the wiring diagram any more. They had the circuit memorized, and that knowledge would transfer to dozens of other models and years.

My concern with your truck is I want to know exactly where the defect is. I'm not satisfied to simply make it work, as you have already eluded to too. To that end, what I would pursue is to connect a wire to the computer side of the connector, for that green wire, then run it to the voltmeter and place that where I could watch it while driving. I've done that with the meter clipped under a wiper arm. Depending where the intermittent break in that wire is, when the problem occurs, voltage is going to go to 0.00 volts or to 12.6 volts. That will tell you if the break is before or after that point.

As for your grade, you dun good, but it's just a start. My grading policy consisted of six pages of tiny type listing a huge pile of tasks you might do in the shop, "minor life-altering events", (quizzes), major life-altering events", (final exam), and a number of other things to factor in to your grade. Everything earned points with 1000 points earning an "A". Everyone could earn the grade they wanted if they put in the effort, and they didn't have to do a strict list of requirements, just like you never know from day to day what you'll be doing on the job. To get to your level of understanding, you would have completed enough text book reading questions and shop activities to earn perhaps 200 - 250 points at the end of week two in my eight-week-long course. While that is indeed on track to earn an "A", the final exam was worth over 200 points, and "Core Abilities", meaning things like "works cooperatively", works productively", etc, was another easy 100 points. Those didn't show up until the last day of class. Every year a few students earned in excess of 1300 points. That would probably be you.

Electronics is the downfall of most mechanics because we have to visualize it. We learn best by taking things apart, manipulating them, and seeing how the parts interact. We can't do that with electrical theory, but we have all seen it with water. I can compare anything electrical to water flowing in a hose or river. That's how I achieved success with my kids. With good electrical skills, a mechanic can go anywhere and get a job. There isn't a single system in any car that doesn't have a computer or some electronics involved.

See how far this gets you, then let me know of your progress.
Was this
answer
helpful?
Yes
No
Friday, July 21st, 2017 AT 12:21 AM

Please login or register to post a reply.

Sponsored links