Garth S

Garaged my 928 after a run 3 days prior, all systems OK: started it last PM ( for an evening run to the much cooler Bay of Fundy ) and noted the following bizarre behavior as indicated by the pod icons.

Ignition key turned to the 'run' position, and all icons illuminated save the alternator red rectangle ( exciter current) - PLUS both red exclamation warning icons were flashing on the pod and center consul .... as was the red rectangle at the bottom of the oil pressure gauge ( signaling low level, as the engine was not running).

Knowing the oil level was OK, I turned the key to 'start', and the engine started instantly; however, the voltage gauge remained pegged to the bottom at "12V". Also, both flashing "red !" cancelled as the engine caught.

With the engine running, the Fluke DVM indicated battery voltage at the hot post - no alternator charging! The absence of the voltage gauge red icon was correct in indicating no exciter current. I did not rev the engine >2000rpm to see if the alternator would be stimulated to charge, rather choosing to let the car sit overnight and reflect on the possibilities ( I did repeat that key sequence three times, with the same results demonstrated.

With thoughts of pulling the alternator, chasing continuity on the blue wire from the hot post to the alternator, etc, I tried to visualize the circuitous path the exciter current follows from the battery source - the first stop is the ignition switch.

That aroused a suspicion, as I've had rare situations where the starter did not engage immediately as the key was turned: those were 'written off' as being my inattention to twisting the key ..... but now indicating the key switch as being a problem????

So, early this AM, I repeated the sequence - after first cycling the key rotation 3-4 times> Viola, the car started perfectly at first touch, no alarm icons, and the voltage gauge sprang to normal running value. The Fluke confirmed hot post static voltage @ 12.59V, and running @ 14.29V.

At this point, it appears that I need an ignition switch ..... or is it that simple??

The flashing indicators and absence of an exciter current of the prior PM were real enough .... but I cannot logically connect them with a dirty contact in the ignition switch; however, the results after burnishing the switch contacts were real, logic be damned .

So, your advice please. Am I missing something that would have caused these (temporary) symptoms?

SeanR

Very possible. I can say for sure that if you still have the original switch in place, it's a good bet it's worn out and needs to be changed. Can't hurt right? At least you will get another 30 years out of it by doing so. Be sure to get a good Porsche switch and not one of those junk ones off E-bay. The light bulb in the cluster/connections to it also may be on the way out and that is why you are getting the issue you have.

Mrmerlin

Any idea when the alternator was last overhauled
after you swap the IG switch inspect the regulator brushes

Jason89s4

Based on your very good description (thanks!) of the actions, and my recent experience on two cars --- I'd bet my dog its your ignition switch. And I like my dog.
-Jason

dr bob

On the 1980 model (and likely all...) the excitation starts at the same ignition switch 15 circuit contact as the ignition. That means that if the engine was running, the ignition switch is not the source of the excitation problem. Sorry about your dog, Jason. I know you were quite fond of it.

The wiring diagrams for that car and this circuit start at the ignition switch at rung 4 on sheet III, where the switch supplies power to bus 15 when in positions 2 and 3 (RUN and START).

Current from bus 15 (ignition switch run or start) exits at H7 to the instrument cluster at 1R, passes through the indicator lamp and a resistor in parallel (mounted on the cluster backboard), exits the instrument cluster at 2R to re-enter the CE panel at H8. These are shown at rung 38 on sheet X, near the top.

The circuit continues out of the CE panel at Z6 and returns at Z1 in the same Z connector. Black conductor. The circuit then exits the CE panel at O8, a black conductor to the 14-pin connector on the fenderwall, pin 1. The front of engine harness (KS-2 on later cars...) carries that circuit as a blue conductor to the D+ (excitation) terminal on the alternator itself. These are shown as rungs 1-5 on Sheet I.

Practical steps would start with cleaning the 14-pin connector and treating it with some Deoxit. Inspect (visual) the quality of the harness wires to make sure they haven't deteriorated to the point of failure. You can verify the circuit upstream pretty easily by testing for battery voltage at pin 1 of the 14 pin with the engine side of the connector lifted. DMM should show battery voltage with the key in RUN position. If true, the problem is likely the connector on the engine side, the blue conductor in the front of engine harness, the connection at the alternator, or the alternator itself. If you don't see battery voltage at your DMM connected to pin 1, look upstream via the connections mentioned above for the CE panel and the cluster. You can test the cluster indicator light function by grounding pin 1 in the 14-pin with the key in run position, and should see the same as you move towards the cluster connection at 2R. Do not ground H7 in the CE panel, nor 1R on the cluster as they are connected directly to ignition switch contact 15 and will have battery voltage with key in RUN or START positions.

There are no fuses protecting this circuit or wiring.

Jason89s4

Ha! Not giving up the dog just yet.
I completely agree with your breakdown of the circuitry and analysis. However, Garth's description is that sometimes the ignition switch will NOT start the car and it will not run, but then sometimes it WILL start the car after cycling it a few times and the car will run and drive fine. When tested while running, he has good 14.29v. That is combined with a variety of other intermittent (and non-rational) lights and conditions that sometimes, but not always, occur. So, sometimes he is getting starter excitation from bus 15 at the ignition switch, other times not.
I chased down these very issues on an '82--and after doing all of the things you suggested and more, after frustration and scratching my head, I changed the ignition switch and "viola."
When about two months later an '89 started having intermittent slow-crank, hot-start issues, and having to cycle the switch a few times to get the starter to engage, along with odd warning lights, I strongly suspected some wiring and grounds needed cleaning (particularly at the starter, which was really bad). After all doing all of that and more testing... turns out it was the ignition switch. (At least everything else is clean now!)
The problem with diagnosis of a bad ignition switch is actually five switches inside of it--not just the switch to circuit 15. All of these have very close tolerances inside the switch housing. Take a bad one apart and you'll see how a worn out switch would cause not only intermittent signal to the starter, but also unintended contact with other circuits. (The attached is from an '83, but the same ignition switch is used on all 928s, and you can see all of the circuits coming off of the same switch.)
Garth--when you try to start the car and turn the key from "run" to "start" and nothing happens, when you are "turning back" to try again, does the car just briefly act like it actuated the starter? Just a fraction of a "start" on the way back to trying again?
I would send you an ignition switch to try, but I suspect by the time it arrived to Nova Scotia you will have solved the problem!

Let us know the results. Dr. Bob may need to start making room for a female doberman.
-Jason

dr bob

I guess I read the OP's initial post differently. In it,
it seems that there were no issues with engine starting, only the relatively odd behavior at the dash indicators for the excitation (red alternator light not testing well) and then the low voltage on the gauge with engine running. The low voltage condition is confirmed with the DMM at

While there very well be some issues with a worn ignition switch, the OP's symptoms and diagnostics with the DMM don't point to the switch as the problem. Perhaps, as you suggest, the worn layers in the wafer switch are somehow passing current to switch terminal 15 to keep the engine running. But that exact same terminal supplies the same 15 bus in the CE panel that supports excitation. If the engine is running via the switch and the 15 bus for ignition and injection triggers from any source, the excitation circuit has the same available.

-----

Add to my Likely Suspects list the alternator and specifically the regulator & brushes assembly that Stan calls out. Sticking brushes in the holder part of the regulator will leave part of the excitation loop open. There's still a transistor gate, an isolation diode, a few resistors and a reference diode in there too, but worn/sticking brushes will still cause the symptoms described. Alternators are known wear parts, and while it's sometimes possible to buy some extra life by replacing the regulator and brushes assembly, the commutator section and the bearings are typically in similar need of replacement. I installed a Bosch replacement in my '89 at a bit under 100k in mine while chasing some charging voltage symptoms for around $150 IIRC, and refurb'd the original with new brushes and bearings and put it in the spares bin. The commutator was still in VG condition on the original or I wouldn't have bothered.

You can connect the DMM to the D+ connection at the alternator and verify that you see excitation voltage there with key on prior to engine start. The voltage needs vary depending on which alternator you have (Valeo, Paris-Rhone, Bosch) prior to start, but the reading should come up to equal battery voltage once the engine is running and excitation current is supplied internally by the alternator itself. I'll SWAG that key in run but prior-to-start readings will be in the 8-10V range, open to experienced input for better numbers.

-----

While the sometimes-doesn't-crank symptom points to the ignition switch as a common possibility, I'd more inclined to look at the 14-pin and the front harness condition as common elements with the excitation symptoms. Diagnosing those is faster and easier than the diagnosis of two separate ignition switch symptoms. Not saying that the sometimes-erratic starter issue isn't the switch, but it's coincidental with the excitation symptom.



Please keep the dog and accept our blessings. Consider her a gift.

Jason89s4

Interested to hear Garth's findings. He certainly has enough information to work with now!
(Depending on how much of his vehicle is still original, considering its age it may be an "all/most of the above" answer.)
-Jason

Kiln_Red

I had intermittent exciter circuit issues that was a bear to chase down. I had to remove the instrument cluster and make a repair on the PCB that required compromising a portion of the laminate shield to effectively solder. In your case, it could be as simple as cleaning the connections at the 3 instrument cluster connectors and evaluating/cleaning the contacts on the PCB.

worf928

One of the Beantown crew had similar exciter circuit issues. Long story short: nut for exciter lug on alternator was finger tight.

Garth may have two issues (gosh, I know: *two* issues at the same time on a 928. It’s never happened before...)
- ignition switch may need a funeral
- independent issue with exciter circuit

The couple of times I’ve replaced ignition switches were provoked by logic-defying ‘lectrical symptoms.

Garth S

Appreciate all the excellent feedback. Electrical things are not my passion, but they seem to appear in clusters .... I spent the latter part of today replacing a blown MOSFET on the infamous 'sword' ( blower fan speed regulator) of my 535i E34 - no simple resistor pack, the Germans of the era could over engineer a screwdriver .... but I digress as this is all old hat to Porsche people.

Dr Bob, a fine fleshing out of my inadequate "circuitous" characterization of the circuit - I will verify voltages and continuity from pin 1 to the alternator as a start.
Hoping that I had not misled regards 'starting' - the car has never failed to start, but has given the very rare action that caused the key to be released and twisted a second time ... as if switch contact had not been successful. That led to a gentle suspicion that the switch contacts were wearing; Consequently, I accept your comment Sean, that a 39 year old switch has its best years behind it - it will be replaced, with your blessings too Jason Dog lovers have to support each other ... oh, and there was never a false starter contact made when releasing the key in those rare situations.

Stan, my alternator was out for an R&R ~ 2 years ago. It was a fairly fresh Bosch AL325X version of the Paris Rhone that blew a power diode, so I soldered in a new rectifier plate: brushes and ring contacts were as new.

Just started the car - barely touched the key to bring it to life; however, I had previously opened a similar ignition switch when replacing the one on our MDX .... it does not take much loss of the dielectric grease to promote a bit of burning/arcing to cause occasional havoc!!

Regardless, still leaning towards the switch as the primary suspect: it likely has some time left, at least enough to source a replacement. Will update at that time - thanks all!

worf928

Two years (IIRC) is about the time it took the 928 mentioned above to loosen it's not-tight exciter post nut sufficiently to produce symptoms. The owner had replaced his alternator about two years previous to symptoms.

Just sayin'

Sometimes it really is the simple stuff.