Diagnosing Alternator Self-Excitation

[JHZR2]

I had another thread about trying to learn about the operations of the exciter. While related, I think a new thread specific to diagnosis is prudent.

It’s a 1993 Mercedes 300SD. Has lots of modules with issues, but car fundamentally operates properly, and the engine, transmission, and brakes are perfect (primary reason why I bought the car).

The alternator bulb checks out perfect, 9 ohms like all the other cluster warning bulbs. It just never turns on.

When I check the bulb socket, key on, I get voltage on both sides. Around 8-11v on the one side, and 12+V on the other.

I do not get any parasitics through the battery and D+ cable combo per my clamp on meter.

If I disconnect D+ at the alternator and ground it, I don’t get the light to illuminate.

If I start the car, and idle it, I am discharging the battery. Once I take the engine over 2500RPM it self excites and gives me 14.3V.

Never under any circumstance does the bulb illuminate. But again, I see voltage at the terminals.

So what’s the best bet to start?

I was thinking step 1 is to do the following:

Start the car, verify no current flowing on D+ or B+. Verify voltage at B+. Use test light between B+ and D+ stud, verify that this excites the alternator.

What else should I be testing and checking to start to track this down?

Thanks!

 

[JHZR2]

So I ran the test I laid out above today.

With the car off, I see 12.69V at the B+ terminal, 0V at the D+ terminal. 0A on the amp clamp on the main battery wire. I believe this is the expected situation.

With the car on (which means a heavy load due to glow plugs), I see 12.3V at B+, nothing at D+. 0.1A on the amp clamp on the battery wire - I think this was a fluke. I would expect to see 12V at the D+ terminal/wire, to excite the alternator.

With the car cranked and started, I see 11.9V at the B+ (because the battery is supporting other loads and computers), nothing at D+. 0.0-0.1A on the amp clamp.

I put my test light with the ground clamp on the D+ terminal, and touched the B+ terminal. The engine adjusted (kind of like when an AC compressor kicks in), and immediately I saw 80A on the amp clamp. Alternator was putting out around 13.9V, and it crept up shortly to 14.3V, and current started dropping into the 70s. I was surprised somewhat that it put out that much current at cold idle... I guess because the engine does afterglow for a while, and the battery (brand new Interstate) will accept it.

So I guess this confirms that the D+ wire is broken someplace. Or perhaps, since we see voltage on both sides of the bulb socket, the cluster is also shorted somewhere...

Question becomes what is the best step to take? Im almost thinking just putting a switch in the cabin that I can instantaneously pulse 12V to the D+ terminal (does it need to be limited, I can put it through a bulb of some sort) upon startup so the alternator gets working right away, not after I get it over 2500 RPM.

I feel like this may be safer than pulling a cluster and dealing with the issues that could occur with that. Though my interest is in fixing things properly, what I describe, I would think would be a good compromise for now... Thoughts??

 

[mk378]

You have voltage at both sides of the bulb (in particular, the side that goes to the D wire) but nothing at the D terminal-- that means somewhere along the wire between the bulb and the alternator is an open circuit.

If you ground the alternator end of the wire and turn the key on then you can probe various places where that wire connects to another harness and see where the voltage is being lost. While you are doing this look at the bulb to see if it comes on when you poke at the connections.

If you really can't find the break and are of the mood to jury-rig something, just run a new wire all the way from the alternator to the bulb. This is such a simple system you don't need to add more complexity with a switch etc.

 

[wrcsixeight]

............ The engine adjusted (kind of like when an AC compressor kicks in), and immediately I saw 80A on the amp clamp. Alternator was putting out around 13.9V, and it crept up shortly to 14.3V, and current started dropping into the 70s. I was surprised somewhat that it put out that much current at cold idle... I guess because the engine does afterglow for a while, and the battery (brand new Interstate) will accept it.

80 amps at idle is impressive, my one 50/120 rated alternator does about 75 at 900 rpm when the battery is well depleted, the other 82, but when hot and idling at 525rpm, in gear, foot on the brake, they are lucky to make 50, fully fielded.

The self exitation only after 2500 rpm is reached, can be viewed as a simpler way to relieve stress on the cold engine. It is said every 25 amps the alternator has to make, sucks up one engine HP, so an instant 80 amp load on a cold engine requiring it make ~3+ more HP, could certainly add to cold engine start up wear. The glow plugs heavy draw, would seem to drain a healthy battery enough every cold start so that it will max out the alternator, for a period of time.

My externally regulated alternators have customizied adjustable voltage regulators controlling them. The are only powered up after oil pressure builds, so basically start up a half a second after the engine fires up. If my battery has been well depleted, and my VRs set to mid 14's, the belts might squeal badly, and I have to crank down the target voltage to the 13.4 range, to stop the belt squeal. Temperature and humidity play a big role in the amount and duration of squeal, as does battery state of charge.

My system, when I had only one alternator, and two battery banks, I would not have the depleted battery bank feeding on the alternator until I turned a 1/2/both/off switch , to both. When I would switch it to both, the tone of the 5.2 liter v8 engine would change dramatically, with the ~ 50 amps of extra load on the alternator.

I can try to remember to just lower my target voltage before startng or befoer engine shut off, when I know My battery bank will be well depleted, to prevent belt squeal and the extra load on a cold engine, but I am also thinking about adding an~ 25 second delay before the voltage regulators turn on and put the alternators to work. If I go more than 30 seconds without the ECM seeing charging voltages, the check engine light will come on.

the 2500 rpm self exitation could be seen, if one desired to view it so, as an effective way of limiting additional cold engine start up wear. I'd personally prefer 1400 rpm, if I could not control it manually.

My single V belt squealing on cold engine starts with depleted batteries and a VR seeking absorption voltage, is only really an issue for the first minute or so. Once it warms up I can still max out the alternator(s) and the belt will usually not squeal. I have Dual v pelt pulleys, Just not gotten around to matched V belts, and am only running one v belt.

Sorry, I have no helpful input on your Light function.

 

[JHZR2]

So, I went out and did some more.

I dug in under the car and got to the harness that has the D+ pin on it...

Supposedly the pinout is 5 wires from X27 connector:
Pin 1 – Br/Bk- 0.75 mm squared – Oil Level Switch
Pin 2 – Br/Gn - 0.75 mm squared – Oil Pressure sensor pin 1
Pin 3 – Vi/Wt – 2.5 mm squared – Starter Terminal 50
Pin 4 – Bu - 0.75 mm squared – Alternator D+
Pin 5 – Br/Yl - 0.75 mm squared - Oil Pressure sensor pin 2

I put a meter between pin 4 and the D+ ring at the alternator. Open Circuit.

I put a voltmeter on the pin 4 female end (that goes into the car), and when the key was off, I saw 0V, key on, 12V.

So I think the issue is in the D+ wire exterior/below the car.

Now I need to find where on the wire it is broken. Thing is, its insulated and wrapped.

So, I think I just need to cut into the harness and probe... But any recommendations for the best way to do that the least intrusive, since Im probing into a harness that is under the car, along side the engine, and will be exposed to the elements...

Thanks!

 

[JHZR2]

I tried some new things tonight. I picked up a tonal sensor to try to see if I could discern where the break was. I put the tone unit on both ends. Both times it was too hard to see where the signal died off. I think the tone generator broadcasts itself, so if the sensor is in the same horizontal plane as the generator, it tends to pick up, especially when in close proximity under a car.

I did get two small wins.

1) I found that the bulb that tested good, actually isn’t. Specifically, one of the little pads on the bulb is broken. I had been using an ohm meter and comparing. But if the bulb is physically broken, it can test ok and not work.

2) I jumpered from the female connector, via an ammeter, to the D+ terminal on the alternator, and got the battery light to illuminate for the first time.




I can’t find the spot where the D+ wire is bad though. I’m worried I’m going to end up hacking up the harness to find the D+ wire. I have to keep the OE terminal, so I’m trying to get to it under the vehicle. In the cold on my back. Fun times…

 

[JHZR2]

So, back on this thread, since I fixed my D+ wire…

My bulb now works. It’s the correct 1.5W rating.

The alternator self excites much faster than before. When it took me to revving the unloaded alternator above 2500 RPM with the afterglow off.

Now it excites at idle shortly after startup, which it wouldn’t do before, but it doesn’t do so instantly…

I don’t see this on other vehicles, including vehicles like my 91 350SD and my Ram a Cummins that both afterglow/grid heat with a significant draw (200A or so).

Is there a reason why the battery light would stay on, but then go off in maybe 3 seconds, and then the alternator self-excite after that?

Again, I have other vehicles that support hundreds of amps of draw to perform heating cycles after the engine is started, without the battery light staying on.

And the battery is brand new, fully charged, good cables and connections….

Any ideas?

 

[LvR]

Without knowing the exact engineering details of the regulator, speculation I think is the order of the day.

It will irritate me too if it was my vehicle (as president of the ANAL club), but I reckon just accept it as a characteristic of the regulator (not alternator) and find a real problem to worry about

 

[JHZR2]

Without knowing the exact engineering details of the regulator, speculation I think is the order of the day.

It will irritate me too if it was my vehicle (as president of the ANAL club), but I reckon just accept it as a characteristic of the regulator (not alternator) and find a real problem to worry about

Well it is a standard issue Daimler-Benz Bosch 90A alternator with a Bosch 14.4 or 14.5V regulator.

I found the issue. Besides the broken D+ wire, the base idle on the diesel injection pump was set way too low. With the electric idle control disconnected, the engine was really rough and the tach didnt register anything unless I moved the throttle. I adjusted the mechanical base idle to around 500, and the electric idle control adjusts it to around 550 or so, temperature dependent, and the light goes right out.

The electric idle control masked the intrinsic low idle condition enough, and the engine is smooth enough and starts perfectly in frigid conditions, so I didnt know there was another issue. It took troubleshooting from the FSM and seeing that all the things flagged checked out, to make me think something else was wrong.

Oh the fun of troubleshooting at night in the snow and cold!

 

[LvR]

.

 

[wrcsixeight]

I'm often pushing my alternator(s) to max output at hot idle with healthy well depleted batteries and 2 awg cabling to and from.

Hot idle, in drive, foot on the brake, and the one alternator can make 32 amps. Field voltage from regulator to field terminal on alternator is only slightly below battery voltage, indicating it is fully fielded.

75 more engine rpm, in park, when hot, and field voltage drops to the 6v range, alternator making 64 amps with battery accepting about 53 amps, rising towards 14.4v.

Huge difference in potential max alternator output at 525 rpm vs 600.

The thermocouple on my voltage regulator's heatsink indicates very fast temperature rise when fully fielded at low rpm, and that temp drops quickly with more rpm as much less field current is required.