[shadowman]

Oh man, been a while since I posted here.

So keeping it short, here's the car: 1987 Chrysler Shelby Z Daytona, intercooled turbo 2.2. As I understand (from previous owner), the car's been modified with a Mopar Super-60 ECU, "purple" cam, and retrofitted cruise control. (I'm not sure if the power windows or power locks were retrofitted. I believe the A/C is stock.). The ECU is the power-module + separate logic-module type.

...and here's the problem: The alternator is not producing a sufficient charge voltage. After a few minutes of running, the "POWER LOSS" light comes on.


Here's what I have done so far:

I checked the alternator's brushes and field coil (while disconnected from the car's electrical); they are good.

At the workbench, I put a current-limited 12V DC source across the field terminals, and with a DC voltmeter on the ground and [+] output, I gave the alternator a few whirls by hand; each time after spinning it, a small voltage was present, that decreased as the rotor spun down. (This tells me the alternator is good.)

Back at the car, with the battery disconnected, I checked for continuity between the [+] batt cable and the [+] terminal of the field harness; when turning the ignition switch to RUN, the meter read 0.5 ohms. (That tells me the [+] field wiring in the harness is good.)

With the alternator connected and the car running, I bridged the field coil [-] terminal to ground THROUGH A 12V LIGHT BULB (to act as a PTC thermistor -- a crude "regulator" of sorts). Suddenly the system voltage increased! The alternator was producing a charge! But the light was flickering...

...so then I connected a small 12V light bulb between the battery [+] terminal and the field coil [-] terminal, and it too was flickering. (Both this test, and the flickering effect of the previous test, tells me the N-channel transistors in the ECU power module that control the alternator's field current (to regulate the output voltage) are indeed working and pulse-width-modulating the field current. And of course all the wiring between alternator and ECU must now be good.)

So here's the puzzling question: With the alternator functional, and the wiring harnesses still good, and the ECU power module still good, why is the system not increasing the field current pulse width enough for a high enough charge voltage?

When I do the ignition switch on-off-on-off-on thing to flash diagnostic codes on the dashboard, I get a code #12 and #16 -- "memory lost / pwr disconnected" and "loss of battery voltage", respectively, according to the Haynes book. (I believe the #12 is typical after the battery has recently been disconnected from the vehicle. I dont know if #16 means the same thing, or if #16 means the ECU is sensing the voltage never increases to a sufficient level to charge the battery.) Anyhow, there are no other fault codes.


Where could the real fault be?

 

[dana44]

There could actually be a problem with the alternator, as in a diode not putting out the amount of voltage it is capable of producing. YOu have enough to produce the 12 volts, but not enough to run more than a few items in the car, thus the code #16. The other possibility could be the computer has the voltage regulator inside it. I had a friend with a LeBaron convertible that I changed the alternator on and ended up with the same problem, turned out to be the regulator. An electric shop was able to wire one into the system (no idea how), but it did bypass the computer itself and functioned just fine.

 

[shadowman]

I tried different light bulbs as my "PTC thermistors", and with low-resistance (ie. high power) bulbs sinking the field current, the system voltage got as high as 16 V with the engine running at only around 1200 RPM. So the alternator is fully capable of putting out sufficient power.

The computer does indeed have the voltage regulator inside it. The power module contains the necessary N-channel transistor (I'm guessing a MOSFET) that momentarily pulls the field [-] to chassis ground in a sequence of repeated pulses. By changing the on-to-off ratio of each of these pulses (ie. pulse width modulation, aka PWM), the computer adjusts the overall average field current, and therefore average magnetic field strength in the alternator, to control the alternator's output voltage.

I don't like the idea of bypassing the power module's system and slapping an external regulator on it, because as I understand, the power module contains a temperature sensor in the intake air stream to read ambient air temperature, and uses this data not only for optimizing the fuel injection ratio, but also fine-tuning the system voltage. (Something about optimal voltage for maximum battery life depends on the temperature of the battery, or so I read. I'm not an expert in battery chemistry.) Well, maybe if external regulators sense ambient temperature too, it's ok... but it still seems like a "dirty" hack repair.

 

[Bob Lincoln]

Have you carefully cleaned all of the connections? Battery posts and clamps, grounds to the block and fenderwell, and all of the ECM connections?

 

[dana44]

Well, the next option would be, after Bob's suggestion, would be a new PCM. Unless there is a dirty connection you have overlooked (which I doubt, given the detailed and technical troubleshooting you have reported), I have no other possibilities to add.

 

[1991spirites]

Unfortunately I have to do the external voltage regulator on my 89 Lebaron GTS Turbo II 5 speed car. The voltage regulator is going out only charging 12.9 volts. I am running a 89 2.5 Turbo I automatic computer in it right now. It's charging 13.8 volts, but I don't the how the radiator fan cycles with the 2.5 vs. 2.2 computer. I have tried all the parts stores and the company that makes the reman computer doesn't make one for the Lebaron anymore, so I can't order it, and the dealers can't get a reman computer anymore. I'm going to open up my original computer to look at the pins in it to make sure they look good. I already purchased the external regulator and plug. I also have a spare alternator harness that I will use to hook it up. I don't want to hack up the original harness.

 

[Bob Lincoln]

Check out any unusual current draw before wiring external regulators. It's entirely possible that a malfunction can draw enough current that the alternator can't hold proper voltage, trying to supply enough current into a partial short. Pulling one fuse at a time for accessories until the voltage jumps to normal, is a good way to troubleshoot.

 

[1991spirites]

In my case the car started out charging at 13.4 with all the accessories off with 13.2 with everything on. Now it's at 12.9 with everything off, and 12.7 with everything on over a period of 3 months. One draw back for the external regulator is that it will throw a check engine like, I don't like that. I could order a performance computer, but right now $300 is a lot of money for me to spend on the car, since I would not send a core back.

 

[Bob Lincoln]

Regulators usually die a hard death, instead of the voltage gradually diminishing. I'd look for a fault elsewhere. I had a bad fan harness once with an intermittent short, which would not blow a fuse, but caused a downward spike to 10.5V. Sounds like something is degrading other than the charging system itself, and loading down the alternator. Or corroding connections.

 

[dana44]

A faulty diode in the alternator is the only other thing I can think of that would do this, too, other than what Bob notes and faulty regulator. The thing is, unless you get the new computer, an external regulator is the only other solution at this point. Unless there is a short, dirty connections (usually don't end up as a draw unless it is part of the charging system itself), or a ground that is deteriorating, what other choice do you have?

 

[Bob Lincoln]

A bad diode that deteriorates can act this way, but usually it wouldn't last 3 days. A good test is to let the car sit overnight and see how much the battery voltage drops. One bad diode will suck it down overnight so that it won't crank. Regardless, assume the regulator is the least likely thing to fail, and check ALL else first.

 

[1991spirites]

With my Lebaron the battery voltage after sitting for about a week was at 12.7, started the car and was charging 12.9. I put a new alternator, no change. I swapped my 89 2.2 Turbo II 5speed SMEC for a 2.2 89 Turbo I automatic SMEC and it started charging 13.8. I then swapped the 89 2.2 Turbo II SMEC back in just incase I wiggled a wire just right, nope 12.9 volts. So for the last month I have been running the automatic SMEC in the Lebaron, no drops in the the charging. So external here I go.

 

[shadowman]

Does anyone here know what the typical PWM frequency should be for the alternator field current?

An EE friend of mine said it's strange that the PWM frequency would be so low that an incandescent bulb slapped onto the field circuit still has visible flickering. So we were postulating if maybe the computer is sensing an over-current fault (ie. field coil partially shorted; has too low an impedance), and every 200 ms or so, the power module keeps retrying to switch on the field. (So the flickering I'm seeing is not actually the normal PWM, but repeated retry attempts.)

Does anyone know if the Dodge/Chrysler K-car computers do this kind of self-protection check?

Does anyone know what the DC resistance should typically be for the alternator field coil?

 

[bamman]

With a code 16, the computer will set the alternator duty cycle to 25% in an attempt to not overcharge the battery. Here is a link that may help you diagnose this problem: http://minimopar.net/fault/code16.html

 

[deder_2]

With the alternator connected and the car running, I bridged the field coil [-] terminal to ground THROUGH A 12V LIGHT BULB (to act as a PTC thermistor -- a crude "regulator" of sorts). Suddenly the system voltage increased! The alternator was producing a charge! But the light was flickering..

I dont it. How can you go from a neg side of the field coil to ground and have a 12v light flicker..negitive side to ground should not cause a light to flicker, unless there is a potential for voltage difference? But even then that shouldnt make a light turn on.. Someone let me know what Im missing..Thanks

 

[shadowman]

I dont it. How can you go from a neg side of the field coil to ground and have a 12v light flicker..negitive side to ground should not cause a light to flicker, unless there is a potential for voltage difference? But even then that shouldnt make a light turn on.. Someone let me know what Im missing..Thanks

The field [+] is connected to the ENGINE circuit (which is "on", ie. +12V, when the ignition switch is in the RUN or START positions). So essentially whenever the car is running, or even just the key is "on" but the engine is stalled, the alternator's [+] terminal for the field coil is always at +12V. The field coil's [-] terminal is connected to a transistor (inside the computer, which does the alternator regulating). This transistor cycles from leaving the terminal "floating", to "pulling it down" to chassis ground. (Essentially the transistor is acting as a switch, but on the negative path of the field coil circuit, not the positive half.)

When the field coil [-] terminal is "floating" (ie. the regulating transistor is "off"), a light bulb connected between this terminal and chassis ground will glow because the field coil and light bulb filament form a series circuit. (The light bulb is not glowing at full intensity, because the field coil adds some resistance.) When the regulating transistor is "on", the light bulb goes dark, because now the field [-] (and light bulb [+]) are pulled to ground, so the field coil is across the full 12 volts and the light bulb has 0 volts across it.

This convention of leaving the [+] permanently connected and switching the [-] is quite typical for electronic circuits switched by transistors.

 

[68RT]

Put the light on the + side and ground the out side. should have same effect on either side. If flickering (which it should do if it flickers on the - side), then put a test light on the + side to ground while the inline "regulator" light is connected and see if the test light flickers. If so, you have problems before the alternator field light. if not, you probably have brush to slip ring issues inside the alternator.

 

[shadowman]

PROBLEM SOLVED!! :w00t: :yahoo:

Thank you to everyone who replied with your suggestions and tips. And thanks to whoever made that document describing the fixed 25% duty cycle field condition, which, turns out my car was doing exactly that.

After building an LED test rig (3 LEDs: Field supply status, and two opposite-phase LEDs for field status / modulating transistor status), I discovered the problem was in the logic module.

Not wanting to lose my Mopar Super-60 logic module, I took it apart and started diagnosing the circuit board. The problem was a burned copper track on the "battery voltage sense" electrical node. What a pain in the rear all these days to find it! But, I did finally find the culprit and repaired it. System works perfectly now! The "flickering" is at a very fast (and erratic) rate, like it should be. And system voltage was 14.12 V tonight. Sounds about perfect!

I can't remember if I mentioned earlier, during my diagnostics, I found there was 0V on both the "battery voltage sense" and "+12V power" lines on the plugs for the logic module. Seeing as how the harness branches 1 wire into these two wires, I figured the fault must be upstream of that inline junction. I thought I had problem pretty much tackled at that point, but then later it "fixed itself" ...yet the alternator still didn't charge (obvious now).

Turns out the harness had an intermittent connection where it plugs into the power module (pin #6 on the 12-pin connector). So I wonder (suspect?) if maybe that intermittent connection caused electrical loads on the "+12V power" branch of the harness to now draw current in a reverse direction, sourced from the logic module, up through the "battery sense" branch of the harness! That could explain why the hair-thin copper track on the logic module's circuit board blew open.

Anyhow, it's all speculation. I can only advise that everyone with a two-piece logic module + power module ECU, make sure your power module pin #6 (12-pin conn.) is firmly making contact to the harness, and that the harness is reliable!

 

[dana44]

Congratulations on finding, and ultimately, fixing the problem.

 

[Bob Lincoln]

Thanks for the detailed reply, so that it can help others.