Alternator question

[ischmitz]

1974 2.7 CIS, alternator with external regulator, red base plate diode assembly on alternator.

Here is the symptom:

At low idle speed (cold start) the alternator light comes on bright red but as I rev to 1500 - 2000 RPM it turns off. Well, almost. It never turns off completely but continues to glow visibly.

I had a guy check the alternator on the running engine and his comment was: The charging current is not as high as in a good alternator but it does charge. The voltage is a little bit on the low side @ 2000 RPM. (He used the inductive current clamp on the battery lead method).

I had the alternator out of the car and checked the brushes. They looked O.K. I am guessing I might have a bad diode. Does that make sense? Is there an easy way to tell without having to take the alternator apart and desolder the diodes?

For example would an oscilloscope connected to B+ and ground on the running engine exhibit one of the three phases missing with funny pattern at 130 Hz and thus confirm an open diode? (1.3 fan ratio, 2000RPM)

Would there be a different effect on alternator performance between one of the six main rectifier diodes going out and one of the three exciter diodes going out?

Ingo

[911pcars]

"...For example would an oscilloscope connected to B+ and ground on the running engine exhibit one of the three phases missing with funny pattern at 130 Hz and thus confirm an open diode? (1.3 fan ratio, 2000RPM).

Would there be a different effect on alternator performance between one of the six main rectifier diodes going out and one of the three exciter diodes going out?"

Dunno.
Got a voltmeter? Measure the output voltage at the battery at around 2000 rpm? It should be at least 13.8 volts indicating an alternator in good shape. If it's not charging until higher revs it could be the regulator.

Hope this helps a little
Sherwood Lee
http://members.rennlist.org/911pcars

[ischmitz]

Originally posted by 911pcars
Dunno.
Got a voltmeter? Measure the output voltage at the battery at around 2000 rpm? It should be at least 13.8 volts indicating an alternator in good shape. If it's not charging until higher revs it could be the regulator.

Sheerwood, my DVM reading is about 12.8V. So I know something is not perfect. I just want to know what it is before I start to rip things apart again and buy parts. That's why I was thinking of using a scope and maybe get some more "insight". I wonder if anyone has ever tried that before.

Ingo

[Early_S_Man]

Ingo,

Yes, a missing phase would be visible on an oscilloscope! But, 12.8 Volts is too low ... that is NO OUTPUT AT ALL!

Did the tech who mentioned the 'low' charging current say it was down by 33% or 66%?

Your glowing warning lamp isn't related to a 'missing phase' problem ... it is due to an increase in resistance in the D+/#61 lead due to oxidation/corrosion on the terminals of the lamp holder or at the two 14-pin connectors ! My suggestion is to clean the terminals (both male and female)and the bulb, and apply Dow Corning 4, 111, or 112 Silicone Grease to the terminals ... and do one additional remedy.

Install a 68 Ohm, 5 Watt resistor directly to the warning lamp bulb holder terminals per the 4th Edition Bosch Automotive Handbook, p. 810. The resistor will lower the 'Zero-Ampere Speed' of the alternator, and may cure the 'glowing lamp at idle problem! Installation is shown in the following bulletin:

[autobonrun]

Similar problem a few weeks back turned out to be a loose solder joint at the diode. So the symptoms for a bad diode would be similar; single bad diode on the B+ circuit could allow the voltage to show ok but reduced current. If you've not tried the tests on the Pelican site, take a look at the Alternator Trouble Shooting Table. Click on Alternator on the left and the link is in the heading.

http://www.pelicanparts.com/catalog/shopcart/911M/por_911M_electr_main.htm

Unplug the alternator cable from the voltage regulator and make the jumper connections discussed in the article. It will isolate a problem with the D circuit from the B circuit.

I could be wrong, but I would think an oscilloscope would show one of the humps missing on the signal. Since the AC is not rectified to pure DC but only inverts the negative portion of the wave, I would think that it would show up as missing using a scope. I've never tried it though. If I get a chance, I'll borrow a scope and see what the signal looks like at the battery. You've got my curiosity up.

Here's another good article that shows the rectified signal waves and describes some internals on an alternator.

http://www.vtr.org/maintain/alternator-overview.html

[ischmitz]

Autobonrun,

maybe I will hook up my scope later tonight and take some pictures of the results to share with the board.

My trouble started after I did a generous engine wash on the cold engine. When I fired it up at first the light would stay on bright red all the time. After some driving (engine drying up) the situation bettered to what it is right now. I was hoping that after complete drying everything would be back to normal. Well, that hope faded as the days went by.

So I am sure I have damaged something in the process. This is why my guess was one of the diodes fried. Maybe the loose soldering could be something to look into, too. It's not unheard of that rapid temperature changes stress solder joints to the point of fracture.

Ingo

[jmohn]

Hi:

Assuming you know the battery is good and all connections clean, you got either a bad regulator or alternator or both (I know you've already alluded to that, just clarifying). Given the age of both, (if original, you have a mechanical regulator the replacment will be solid-state and better) both are due for changing/rebuilding anyway. I just went through this with my early SC last summer. Why mess around? Since you have to remove the alternator to properly test it (and that's not that much fun) just take it to an automotive electrical shop and have it rebuilt (cost me about $65-$70 last summer) and replace the undoubtably marginal mechanical regulator with a new one, which will be electronic and sealed. I got my new regulator from a Pelican competitor, e-mail me direct if you want the name. BTW the message about installing a resistor in line with the warning light is part of upgrading to an internally regulated alternator, it's not necessary or advised with the external regulator.

Good luck,

Jerry M
'78 SC

[304065]

I respectfully disagree with the above advice: you are not supposed to use a voltage regualtor with a different manufacturer than the alternator. E.g., if you have an externally regulated alternator with a red backplate, I think that's a Marchal/Motorola, and so you should use a Marchal/Motorola voltage regulator. If you have a Bosch, use a Bosch. If you go to a different manufacturer you could have unpredictable results and will have introduced another variable which could bug you during diagnosis.

Also, Warren's advice about installing a resistor in parallel with the bulb will have the same result if performed on an externally regulated alternator: by increasing current flow in the D+/61 circuit, "bootstrapping" will occur at a lower RPM than if you only ran the 2W bulb.

Hmm, hearing the description I wonder whether the belt could be slipping from the water? You would probably hear it squealing though.

Good luck! I would love to see those scope pictures!

[ischmitz]

jmohn,

Good points. I was not going to put the resistor in. I am just trying to understand what I could have damaged in the process of hosing down the engine. What bugs me is that the G-light is only glowing faintly and it apparently charges enough. I never had to recharge the battery externally in 9 month.

If it is as simple as an open diode I could get the diode assembly at a local alternator rebuild shop for less than 30 bucks and would be all set.

Eventually, I agree it would be nice to put in either a rebuilt unit with internal regulator or rebuild the one I have and put in a solidstate regulator.

Ingo

[ischmitz]

John,

yes its the Motorola/Marchal type with the red back plate.

Concerning the fan belt that would be an idea but as I mentioned earlier the washing event was several month ago.

I am sure that before the washing the light went out completely and did not come on at even at engine speeds below 800 RPM.

Right after the washing it first stayed on. After driving around for an hour that went away to the condition it is in currently. I am pretty sure I did not hear the fan belt slipping. Also I watched right after I started it up and the fan rotated fine. So I am still convinced that I fried something in the process.

I will get to the scope thing tomorrow and post some screenshots of the scope at various engine speeds.

Ingo

[RoninLB]

did you check for open diode with OHM meter? ..........Ron

[ischmitz]

Originally posted by RoninLB
did you check for open diode with OHM meter? ..........Ron

Ron,
Do I have to pull the alternator for that?

Would I get any meaningful readings by checking at the regulator leads (with the regulator and battery disconnected)? Let me see - I am checking my Haynes manual as I type

Ingo

[ischmitz]

OK,

here is the scope picture. I decided to shake up the neighbors

Engine speed @ 1000 RPM. The Generator light is glowing. The measurement is taken between D+ and ground. Peak to peak is 15.4 Volts and the frequency is 130 Hz.

A 3 phase AC getting rectified would produce 6 peaks per engine revolution. Mine looks like one of the six peaks is missing: 1000RPM*1.3*6 -> 130Hz. Now is this a bad diode?? Any educated guesses as to which one (exciter or rectifier)?

Can someone maybe replicate this with a scope and tell me what his is looking like?

[jmohn]

Perhaps I should have specified "replace with a CORRECT or PROPER regulator". I've never heard that it has to be the same "brand" as the original.

Jerry M

[RoninLB]

I've never heard/read that either..but it is a logical routine.........Ron

[Early_S_Man]

Ingo,

For your next test ... can you connect the second channel of the scope to the B+ or a hot fuse location? Having both signals displayed simultaneously may help to detect the action of the regulator, and responses on the output Voltage waveform.

[ischmitz]

Originally posted by Early_S_Man
Ingo,

For your next test ... can you connect the second channel of the scope to the B+ or a hot fuse location? Having both signals displayed simultaneously may help to detect the action of the regulator, and responses on the output Voltage waveform.

Will do tonight.

When I measured B+ alone I did not see this big of a swing. With all the ignition noise in the signal it was hard to trigger to the alternator. This time I will use AC coupling for B+, DC coupling for D+ and trigger to D+

The more I think about this I suspect one of the exciter diodes to be open. It would cause the net exciter flux to be reduced by about 1/6 and thus at higher engine speeds the regulator would still be able to put enough flux through the rotor to get the output to 14.4V. At low engine speeds the regulator just maxes out and the light comes on.

[ischmitz]

A guy at the local alternator rebuild shop got me totally confused today. They claimed there are Marchal/Motorola alternators for 74 Porsches with a total of eight diodes around. They said the red heatsink plate in the back only carries two diodes (isolation diodes).

I thought we are talking about exactly 9 diodes here (3 ground, 3 D+, 3 B+)

Heck, I am taking the next measurement tonight and then rip the thing out to get to the diodes. 8, 9, or whatever, I will figure it. Does someone have pictures of the alternator and it taken apart by any chance.

Ingo

[autobonrun]

Your scope picture showing D+ looks good. I'd try to look at the B+ by connecting the scope at the battery.

I performed the two tests mentioned in the Alternator troubleshooting table and it worked. Namely:

1) Disconnect alternator cable from the VR.
1) Connect 12v from the battery to the DF terminal of the alternator cable. This should drive the voltage at the battery up to its maximum (16volts +). The light should go out; otherwise you have a bad diode or connection in the B+ circuit.
2) Short the DF terminal of cable to the D+ terminal. Again, the light should go out.

The first test checks the ability of the battery to excite the field and drive the output voltage (B+). The second tests the ability of the alternator to drive the voltage.

If the alternator light gets brighter as the rpm's increase, this indicates that the alternator output (D+) is getting larger than the battery voltage, indicating either a bad battery or failure of the B+ circuit to increase the charging current/voltage to the battery.

I'm fairly sure you'll find the problem is in the alternator. The water may have created a short damaging one of the diodes; similar to the same thing you would get if a screwdriver accidentally shorted the + to ground. Now where that short occurred is another issue. You'll have to find that to make sure the next water treatment doesn't cause the same problem.

[Early_S_Man]

Ingo,

I think your alternator shop guy is confused about the year and style of Marchal alternator! The original 1968 490 Watt 911T Marchal's had a SINGLE isolation diode (for a total of seven diodes) mounted on an external heat sink ... it is illustrated in Volume II of the factory service manuals.

Later Marchals had no isolation diodes or external heat sink, but had a small circuit board internally with three exciter diodes similar to 1N4006, and can be replaced fairly easily after disassembly. There is no maintenance info on the later style Marchal alternator in Volume VI of the factory service manual for '72 thru '83 models!

Considering that the diodes are at the back of the alternator, and completely shielded from water by the air shroud ... I doubt that your water theory is what went wrong in this situation! I would suggest having the alternator tested before deciding to disassemble it. The Voltage regulator may have died, or corrosion on the terminal may be causing the problem!