Charging System Design Flaw?, Or is it just my car? Options

[pbanders]

It may be that what I found is specific to my 914, but from what I can tell, this issue should apply to all 914's, unless they have a different engine relay board than mine.

In a thread on the PP board, I report (very long) on recent things I've done while chasing down idle instabilities with my car - the main one being a large drop when I add electrical load (e.g. lights). While tracking this problem, I found what appears to be a signficant design flaw in the charging system. First, some background on the charging system.

See the following URL for an excellent description and troubleshooting guide for VW-based charging systems:

http://www.ratwell.com/technical/ChargingSystem.html

I used this article as my guide when tracking down the problem I describe.

A properly-operating charging system supplies sufficient current from the alternator under all running conditions (except starting) to power the car and to keep the battery charged. The voltage regulator works with the alternator to make this happen. The voltage regulator accomplishes this by measuring the voltage of the battery and controlling the voltage put out by the alternator. This means that when the car is running, the alternator voltage will always be somewhat higher than the fully-charged battery voltage, to keep it charged.

A fully charged auto battery produces about 12.6 to 12.8 V. Even a small decrease in battery voltage indicates that it is not in a fully-charged state. From the article referenced above, a battery at 12.35 V is only about 50% charged, at 12.0 V its at 25% charged, and at 11.8 V it's completely drained. When your charging system is working correctly, your system voltage will be at least at 12.8 V or greater.

For the voltage regulator to work properly, it has to have the correct ground reference. If you look at the following diagram of the 914's relay board:

http://www.pelicanparts.com/914/parts/Elec...lectric_73E.jpg

.... part number 4 is the voltage regulator. Note that the D- connection, the ground reference, is connected to the case of the voltage regulator, which is shown connected to ground.

BUT WHERE IS IT GETTING THE CASE GROUND?

The relay board is made of PLASTIC. The screws which hold the voltage regulator to the relay board do not connect to chassis ground. As a result, the current path to ground for D- has to go through either the alternator or through the voltage regulator itself. I've measured the resistance to ground of D- on my car, and I got a value of 4 ohms. When the car is running, this resistance causes the voltage measured at D- not to be zero volts (as it should be, as per the reference I provided above), but to float to -1.1V. This causes the output of the voltage regulator to never exceed about 12.0 to 12.2 V, preventing the battery from ever charging to more than about 50% of a full charge.

I did some further measurements of the resistance of D- to ground. If I pull the alternator plug from the relay board, the resistance increases to about 100 ohms. If you follow the current path to ground, you can see that it now goes through the voltage regulator, through the D+ connector, and though the alternator warning light circuit - which has a bulb in it that has about 100 ohms of resistance. If you open that circuit, you get no path to ground - the case of the voltage regulator isn't grounded.

I connected a short ground wire between the voltage regulator case and chassis ground. Now, I read 13.0 V or greater under all running conditions, DF is at about 5 to 6 V when idling, and my battery stays charged to 12.8 V. The charging system is now working as it should have.

EDITED: DON'T DO THIS. The cause of my problem is a bad alternator ground. D- is grounded to the case, and the case is strapped to chassis ground. DON'T PUT A WIRE between the voltage regulator and ground if you have a bad alternator ground like I do. Why? Because ALL OF THE CURRENT will now flow through the jumper wire - as much as 20A or more with high beams on. You could start a fire or worse - DON'T DO THIS.

The question here is - is this common to all 914's, or is my car somehow different? If your voltage regulator is as mine is, connected only to the plastic relay board plate, and with no case connection to ground, then I see no way that your system should be any different from mine, and you should be seeing a lower charging voltage than is needed to keep your battery fully charged.

I'd like to see some readings from other 914 owners cars. Here's what I'd like for you to measure, it only takes a few minutes with a simple digital voltmeter to take the readings.

1. With the ignition switched off, measure the resistance from the case of your voltage regulator to a chassis ground. If your charging system is set up correctly, you should read about 0.1 ohms. Anything more than 0.5 ohms is suspect.

2. Remove your relay board rain cover, and start the car and let it idle, headlights off, heater/blower fans off, stereo off. The alternator connector is the three-pin connector on the side of the relay board. Set your DMM to DC volts and measure the voltage between a chassis ground (there's a ground connection on the firewall beside the relay board that's nearby) and each of the pins (you can stick a probe down beside the wire into the connector and contact the pin easily). Here's what you should get:

D+ (pin that's nearest to the front of the car: greater than 12.8 V, no more than 14 V

D- (next pin down, brown wire): no more than 0.1V

DF (next pin down): about 5 V or so (depends on the electrical load of the engine and the charging state of the battery)

If you don't get the results above, then your car has the same problem as mine did. A simple ground jumper between a chassis ground and the case of the voltage regulator fixes the problem. Let me know what you find.

Also, if my analysis here is wrong, tell me where I've gone astray. If there's another way the voltage regulator case is grounded on your car, or your relay board is different, let me know. Thanks!

[SGB]

Thats what I'm gonna do, just crimp a couple of u- shaped ends on a 10 ga wire and ground from the VR attachment screw to the ground pont on the side hidden behind the relay cover...I DID NOT DO THIS!

[pbanders]

Thats what I'm gonna do, just crimp a couple of u- shaped ends on a 10 ga wire and ground from the VR attachment screw to the ground pont on the side hidden behind the relay cover...

It won't hurt anything if you just go ahead and do it, but spend a minute first and check to see if you really have a problem. Check the resistance between the voltage regulator case and chassis ground first - if it's under 0.5 ohm (closer to 0.1 ohm), then the extra ground isn't needed.

If you do find that it's grounded already, let me know. I'd like to figure out where it's getting the chassis ground from.

[904svo]

Thats what I'm gonna do, just crimp a couple of u- shaped ends on a 10 ga wire and ground from the VR attachment screw to the ground pont on the side hidden behind the relay cover...

It won't hurt anything if you just go ahead and do it, but spend a minute first and check to see if you really have a problem. Check the resistance between the voltage regulator case and chassis ground first - if it's under 0.5 ohm (closer to 0.1 ohm), then the extra ground isn't needed.

If you do find that it's grounded already, let me know. I'd like to figure out where it's getting the chassis ground from.

I'm getting my ground from the grounding strap from alternator to engine case
see post #3.

[pbanders]

Thats what I'm gonna do, just crimp a couple of u- shaped ends on a 10 ga wire and ground from the VR attachment screw to the ground pont on the side hidden behind the relay cover...

It won't hurt anything if you just go ahead and do it, but spend a minute first and check to see if you really have a problem. Check the resistance between the voltage regulator case and chassis ground first - if it's under 0.5 ohm (closer to 0.1 ohm), then the extra ground isn't needed.

If you do find that it's grounded already, let me know. I'd like to figure out where it's getting the chassis ground from.

I'm getting my ground from the grounding strap from alternator to engine case
see post #3.

Look at the following wiring diagram:

http://www.pelicanparts.com/914/parts/elec...lectric_71b.jpg

D- isn't connected to any case ground in the alternator.

[Spoke]

D- isn't connected to any case ground in the alternator.

I looked at the alternator on a spare engine that I have and it has 4 wires as shown on the schematic:

1) D+
Reference voltage output for voltage regulator and alternator idiot light. On the relay board, this signal
enters on D+ and connects to the D+ of the VR and to pin 2 of the connector to the vehicle interior.

2) DF
Positive control for the rotor field windings from the voltage regulator. On the relay board, this signal enters on DF and connects only to DF of the VR.

3) D-
Negative control for the rotor field windings from the voltage regulator. On the relay board, this signal enters on D- and connects only to D- of the VR. No other connection to ground on the relay board.

4) B+
Output voltage to vehicle for battery charging, vehicle operation. As shown in the schematic, this wire goes to the starter first, then on to the battery positive terminal.

Note that the ground for the alternator has no wire. The ground connection for the alternator is assumed to be the case.

What is missing on the schematic is the alternator ground connection and is your point. In the schematic, the box drawn around the alternator (3) is the case of the alternator. The anodes of 3 diodes are connected to the box which again is the case of the alternator. The ground (case) of the alternator is not shown connected to vehicle ground. Without the vehicle ground connection through the case or the added wire, the alternator will not work. The D- (negative field winding) is not shown connected to the case. On the spare engine, D- going from the alternator to the voltage regulator measures zero ohms (short) to the engine case.

In the schematics in my Haynes repair manual for 74-76 914, the ground connection to both D- and anode of the 3 diodes is clearly shown to ground. Since there's only 4 wires to the alternator, the ground connection here must be the alternator case. For 70-73 914, the manual uses the same schematic as shown in the previous link and does not show a clear vehicle ground connection for the alternator.

I wonder if in your car, the case ground connection is resistive and loading the alternator causes a voltage drop thus the lower battery voltage. One test that maybe you've done is to measure D- with respect to vehicle ground. If without the wire this wire goes NEGATIVE when the alternator is loaded, then the alternator ground is not good. The value of the negative voltage should be similar to the difference of the before- and after-wire-install battery voltages. Thus if your wire makes a better ground connection, the output voltage will be higher.



Spoke

[pbanders]

<good analysis of the system clipped....>

I wonder if in your car, the case ground connection is resistive and loading the alternator causes a voltage drop thus the lower battery voltage. One test that maybe you've done is to measure D- with respect to vehicle ground. If without the wire this wire goes NEGATIVE when the alternator is loaded, then the alternator ground is not good. The value of the negative voltage should be similar to the difference of the before- and after-wire-install battery voltages. Thus if your wire makes a better ground connection, thus higher output voltage.

Yes, that's exactly what I measured. Without the ground jumper, D- with the car running reads - 1.1 V, indicating the alternator ground is bad. My problem was that I couldn't find a schematic that showed the alternator case ground to D- until 904svo showed the Factory Workshop Manual page.

EDITED: DON'T ADD A GROUND JUMPER FROM THE VR CASE TO GROUND. IF YOU HAVE AN ALTERNATOR GROUND FAULT, ALL THE CURRENT SUPPLIED BY THE ALTERNATOR WILL FLOW THROUGH THE JUMPER AND YOU COULD START A FIRE. DON'T ADD A GROUND JUMPER!

Like I said, this looks like it's a common problem, but it's very easy to test for as you suggest, all you have to do is to monitor D- while the car is running. If it shows a signficant negative voltage, then the alternator ground is bad. It also looks like once you've established that the alternator ground is good, that it wouldn't be a bad idea to tie the VR case to ground, too.

So...

1. There's no design flaw with the charging system - makes sense, hard to believe that all 914's would have this problem without it being solved, I'm sure any good mechanic (not me) who has worked on 914's would have figured this out years ago if there was no alternator case ground tied to D-.

2. Despite what the wiring diagrams for the relay board in the Haynes manual and other places show, there is NO ground connection to the VR case provided by the relay board. It's not needed, as D- is tied to ground through the alternator.

3. Despite what many (but not all) of the wiring diagrams for the 914 show, D- in the alternator is tied to ground through a case ground in the alternator. This is clearly shown in the Factory Workshop Manual page provided by 904svo, as well as in the current flow diagram shown here:

http://www.pelicanparts.com/914/parts/Elec...lectric_74E.jpg

4. A simple measurement of the running voltage on D- can test the case ground of the alternator, in the case of testing for the cause of a low charging voltage situation.

Thanks again to all who helped me figure out the root cause of this problem!