avandyke

Hey everyone. I've been having an issue with my 928:

1984 / US 928S / Auto

and would love to get some ideas on how to proceed. This is my first post here on Rennlist, so please forgive formatting issues or whatnot. I'm very excited to join you all!

On to the issue. Over the past two and a half weeks (this is important below) my 928 has occasionally failed to start. It's happened four times in this period of time, and has behaved in a couple of different ways.

The first time it failed to start, I let it sit for about a half hour, tried again, and it ran.

The second time I let it sit for the whole workday, tried again when I got home, and it ran.

The third time was this past Wednesday in the work parking garage, so I left it overnight, tried again the next day and..... nothing. The pump won't even prime when the ignition is turned on, it's just silent back there.

So I replaced and tested a bunch of stuff over the next couple of days after that failure, partially on the advice of my mechanic and partially through following fault guides online:

- New fuel pump relay
- New fuel injection relay
- Tested positively for spark
- Tested pump via jumping pins on relay board

And the last one there is what ended up giving me results. If I jump the pins on the relay board and force the pump to run a bit as I turn the engine over, it starts right up. I can then reinstall the relay, and it'll run. I can hear the pump prime the system when the ignition is turned on, and I can start and stop the engine without issue from there on.

So that was last Friday. I tested the car on Saturday, to see if it'll start, and it started. Primed and ran nicely. I tested it on Sunday, it started. I jump in this morning to go to work, and boom, no prime, no start. I jump the pins again, forced the pump to run for two second or so, threw the relay back in there, and it primed and started right up.

Now, I'm flabbergasted as to what might be causing this very repeatable, but oddly fixable issue. This car is fairly new to me, so I'm still learning its ins and outs, so any ideas on what might be causing this issue is welcome.

Any thoughts or advice on how to troubleshoot from here? Should I look into checking the ECU? Or ignition issues? Is there anything else that might interrupt or otherwise cause the pump signal to not reach the fuel pump relay?


And lastly, because I haven't forgotten that I mentioned the importance of the timeframe, but didn't want it to bias the base issue, I wanted to add that I actually had a large maintenance project done to this car two and a half weeks ago, specifically to replace the timing belt and water pump. Additionally, the mechanic replaced the ignition "green wire" (which he replaced after noticing a non-start scenario after he attempted to get my car off the lift after these repairs), and a oil separator breather hose.

Could any of the maintenance done to the timing belt system be causing these non-start issues? Maybe a change in ignition timing (I understand it runs off this belt) that is now causing the ECU to incorrectly send signals? Even more simply, would a faulty ignition distributor wire cause these sorts of issues?


Thanks in advance for advice and ideas!

dr bob

First of all, Welcome to Rennlist! This is probably the best and certainly most supportive 928 group in the neighborhood.

The fuel pump...

The fuel pump relay gets a "trigger" signal from the ignition system based on the Hall sensor in the distributor. That "green wire" you replaced is considered the most fragile component in the system, so you've headed off at least that common failure piece as an option. That same signal from the distributor generates the tach pulses, so if you see that tach needle jumping some as you crank the starer, the Hall and green wire circuitry is -probably- OK. You can confirm with a timing light on a plug wire -- if you see a spark pulse then the ignition and that pickup are OK.

The relay itself has a couple timing circuits in it that depend on seeing the tach pulses. Initially with the key on but before the starter turns, the relay closes for a few seconds to run the pump and "prime" the system. If you crank the starter within a couple seconds of "key on", the pump will stay running. Otherwise, it waits until another pulse from the tach loop before it closes again and restarts the pump.

The connection between the distributor and the ignition box is a very low-voltage and high-impedance circuit. It is therefore sensitive to connection quality. Look at the connections at both ends of the Green Wire, and clean anything that isn't really bright and clean. A good contact cleaner like DeOxit will help if there's any question.

Lately there's been a rash of users looking at failing/tired fuel pumps and the excess current they draw. There are some easy ammeter checks that can be done to see if the pump is drawing enough current to trash the relay.

Be aware also that the fuel pump relay feeds power to the aux air valve that hangs in the same area where you changed the timing belt. The wire to that valve, if it's shorting to ground, can cause the symptom you described,. It's too small to blow the fuse but may melt its insulation. Check for a hot fuel pump fuse as a telltale of either a short in this wire or a failing pump.

a4sfed928

Have you tried a new relay?

avandyke

dr bob - Thanks for the ideas.

I actually just spoke with my mechanic and he mentioned the aux air valve. When I got the car back from them after the maintenance, I also noticed that the idle speed was about 300 RPM higher than it was than when I brought it in, which concerned me at the time, but we chalked it up to that breather hose being replaced, thus sealing a vacuum leak, perhaps. He's now saying that this valve can also modify idle RPM, is that right?

It's come up twice now, so it's time to have a look at it, and the fuses you mentioned as well for heat.

As for the Hall sensor and green wire, there is a spark going to the plugs, I tested it with an inline spark light and confirmed that. Additionally, I am getting a jump on the tachometer when cranking the engine. I will check the connections on that wire, however, and ensure they're clean. Same for the grounds nearby.

Thanks for the advice, and for the concise explanation of how the system works. It helps.


a4sfed928- Yup, and it was actually replaced as little as a couple years ago as well, so now I have two good quality working KAE fuel pump relays. I'll save the second for a rainy day, I suppose.

Alan

There have been quite a few who note their relays no longer do a priming pulse on ignition first turn-on. This should not cause no start - only a delayed start (more cranking) . Several folks have said new relays fail to prime. The circuitry for this is inside the relay so it means poor quality (possibly 'aftermarket' type) relays... You may want to consider a relay from a well known source instead.

In any case this fail to prime certainly won't stop the car from running (eventually).

So for your case its more likely something to do with an intermittently missing rotation signal, or there is also an intermittent fault in the circuit that detects this rotation inside the relay - it looks for pulses at > certain rate to know the engine is running (or at least cranking). So if the relay is flaky in one mode...

Alan

Landseer

Have you ever cleaned the engine/computer grounds?
There are two ground points, both on the passenger cam cover.
one is completely hidden by the air pump valve (which must be removed for access).

Cleaning the body to block ground cable might also be appropriate if not already done.
Its located under the car, passenger side, near the header pipe.

John Speake

Welcome !

jpitman2

You dont state exactly what happens with the no start - does it crank over, but not start, or not even crank? I had a no crank case a while ago, and thought it was the ignition switch, so replaced it. Then it failed to crank again....starter relay was the real culprit.
jp 83 Euro S AT 57k

avandyke

Alan- This path of inspection sounds like the right one in this case. Any advice on what to check/replace/fix right away for a missing rotation signal?

Landseer- Going through all the grounds and cleaning them is high on my to do list now. Thanks for the locations, I'll check them specifically.

Cambridge John- Thanks! The community here is legendary.

jpitman2- Whoops, sorry if I wasn't clear. The engine cranks perfectly fine, I can sit there and crank it until the battery dies. So the ignition switch seems to be alright, and then the starter relay seems to be alright. The issue is that occasionally when I turn the key to the ignition on position, the fuel pump will not run to "prime" the system. It will continue to not pump, hah, when cranking the engine. To fix it, I can jump the pins on the relay board, force the pump to run for a second or two, reinstall my fuel pump relay, turn the key again, and the engine (and pump) will start and run.


Thanks everyone for some good ideas. I'm going to clean up all the grounds, contacts on the aux air valve, and ECU contacts, see if it's simply a ground issue, then update the thread! I'm also going to check whether my fuel pump relay is getting a signal from the ECU next time I see a failure, see what's happening there.

ECU Doctors

It's possible that the ECU which controls the fuel delivery might become intermittent.

That behavior can be easily bench tested here if you cannot find another buddy with a similar car.

Landseer

Do you mean the injectors would randomly drop out?

ECU Doctors

Yes, that and many other symptoms.

ECU's have different components that may work intermittently.

An example of that are capacitors which "lose" with time their "capacity" to "hold" all the charge that it was designed for.

See example attached.

So it doesn't mean that the ECU is "completely" dead, it's just that it's showing signs of wear and tear.

dr bob

Unfortunately the "ECU" in question is the transistor ignition module, a relatively stupid device. It generates the tach pulses used for the console indicator/gauge as well as the fuel pump relay. The OP's symptoms of the tach needle jumping on crank, and the ability to "prime" the fuel system with a jumper, then let the relay run the pump while starting, points to a problem away from the ignition module. Remember that the early-car systems are the same as those used in other period Porsche (and other Bosch-equipped) cars. Distributor, pickup, "green wire" and connectors, fuel pump and relay problems are no different from what 911 and 924/944 owners (plus MB, VW, Saab, Volvo and more) have enjoyed.

Jay G

I have had a similar problem on my 88 with engine cranking just fine, but fuel pump relay not kicking in. Once it was a low engine oil problem (added a half quart and it started right up), and once it was "confused" door locks. Once I got them synchronized to both lock and unlock at the same time, it seemed to clear the problem. Perhaps there are other electrical lockouts, but I can't read the schematic for a hill of beans.

Alan

Neither of these had any effect on the fuel pump, but it is possible that in fixing these you moved some unrelated connection that did help with the fuel pump issue.

Alan