3.2 no start / rough running

[rob911]

Problem began as no start on christmas eve. Turn the key to start postion - click then clunk, motor does not turn, all electrics go out. Have to disconnect - reconnect battery to reset. Suspect starter / solenoid. Problem cured when starter motor replaced 3 weeks ago. hooray.

Last weekend drove 75 miles to classic car show in london - meet with other classics at Ace Cafe before drive in convoy to show (reserved parking), all with no fuss or drama. after 10 minutes we drive off in convoy - car will not pull more that 1800rpm, feels like no power or low fuel, car getting hot (1st line on temp guage). Struggle to the show, replace alt belt + increase tension, go to show for two hours. Drive home lasts 20 minutes before we coast to a stop and call recovery truck, which brings me home. Car will start and run (just about) enough to crawl into garage.
Mechanic thinks fuel pump - begin the tests, results in next panel for easier reading.

[rob911]

Apologies for length of post - having two cars off the road on a sunny bank holiday has sent me to the edge.
Begin tests again. Have searched forum and read Bentley manual all day.
Have ordered new DME relay (suspect both main and glovebox spare item), and fuel pressure gauge.
Starter will turn engine - will not "fire"
Suspect DME relay, so remove for test
Jumper sockets 30 (12v supply) and 86 (output TO fuel pump I think) - fuel pump runs. Sounds noisy - but this is the first time I've heard it without engine running.
Insert relay, ignition to start position
Test for voltage at 86 = 12v
Test for earth from DME at 87b = not present
Fuel pump does not run, so suspect DME not supplying ground at pin 20 ("engine turning" signal is missing ??).
What else is needed, before ground is sent to DME relay?

Add temporary supply to fuel pump fuse (switched from ignition), start engine. Engine runs badly, speed not stable and backfires a bit. Will not " rev up" easily. Engine will falter when I flex cable to speed reference sensors
Suspecting speed and/or reference sensors, but continue tests

Switch off to test sparks.
Add spare plug (not porsche) to plug lead and rest on cam cover.
Start engine - spark in NOT blue - only yellow.
Could this be a poor test? Should I remove a real plug?
Suspect ignition health poor
Measured the resistance of the distruibutor cap
Centre coil connection (exterior) to carbon brush (interior) = 1114 ohms
Each plug lead to the internal connector = 900 ohms approx
The plug leads and rotor are within specification.

Looks like the distributor cap is way out of spec.
Do i need a new distributor cap?

Seems that I have two faults at least.

Speed / Reference sensors or wiring causing no start ( no fuel pump power)
Distributor Cap causing rough running ( poor spark )
OR could it be failing fuel pump, which allows start after running up for 20 seconds before turning over.

BTW exactly one week before this my Alfa 166 3.0 V6 24v cambelt snapped at 6000rpm. Car is still with mechanic

[911mot]

Dunno if it helps but I've heard the engine earth strap can cause weird problems if not contacting properly.

[rob911]

Have heard the same, cleaned up engine earth strap ( christmas time) during earlier no-start. Cleaned earth on intake number one this morning, nut will not tighten fully (threads stripped on intake ! eeeek!). Still seem to make a good earth though.
How can I be sure ?
Do I need to bolt them to another point in engine bay?

Just to make it even more interesting - sometimes the starter fails to turn - just like before the starter motor was replaced. Grrrrrrr

Really hope someone can help - I cannot afford any mechanic time - and would love to solve this at home. Look forward to some replies, got to go now , just noticed the clock jump forward one hour.

[911mot]

It wouldn't hurt to try, there's plenty of available places.
Does it get wet a lot? Was it raining on the drive to Ace? Check the engine harness multiplug on the left hand side for corrosion. In fact check every multiplug you can. I washed mine once and didn't drive it after. Condensation built up in the AFM multiplug and when I tried to start it a few weeks later.....
Don't go mad changing things until you've swapped in a known good DME though.

[BGCarrera32]

If the fuel pump runs and is whirring when you jumped the connections to give it power, I suspect you *do not* have a fuel problem.

Is your tach doing anything when the engine is cranking and trying to start but not firing? If its not fluttering when your cranking the engine I'd look again at your speed sensor. Make sure your gap from the sensor to the flywheel is set properly (have you had the clutch replaced or trans service lately that could have knocked it out of place?) See Bentley section 280-14 bottom photo... I had something simliar to what you describe happen once and caused all kinds of wierd things. Sometimes the wire to the sensor can go bad from resting on the hot exhaust. If you really can effect engine performance by wiggling those wires you're on to something. They are very reliable coil sensors but can go bad...

Process of elimination here...start with the simple stuff, keep methodically knocking things off the list and keep notes while you do it...best of luck...hang in there.

-BG

[rob911]

911mot
car was not wet - it was the best day of the year so far, sat March 19th.
Will definitely accept your advice, will wait for good DME relay before changing anything else. Both of my DME relays behave the same though.

BG
Yes the tacho was running, following engine speed. This does lead me to suspect the reference snsor or wiring.
Does anyone know if this is used to detect engine rotation by DME computer?

I too assumed that fuel pump must be OK if it runs. I will still check the rail pressure thursday when my adapter arrives. Why do porsche put M12 * 1.5 female connection on the fuel rail? Real problem getting something to fit this.

I hope to help cure the problem by applying a "bandage" to the ref sensor wiring later today.

[ChrisBennet]

rob911,

It seems like you've got a pretty good handle on things. Here are my thoughts:

You said you got 1114 ohms measured between the distributor cap button and outside terminal and deemed it "way out of spec". I just took measured a 964 distributor cap and got 1010 ohms so I think your cap (at least the button) is fine.

I'm with BGcarrera on the sensor issue. I'm not a fan of throwing parts at a problem but if the speed and reference sensors are the originals (20years old) I'd replace them both because they are at the end of their reliable life. Here is a picture of those wires from the last 3.2 I worked on, an '85:



Off the top of my head. the DME needs to "see" these signals before it will enable the fuel pump:
- reference signal during starting (only)
- speed signal during starting and ongoing
- signal from from air flow meter that indicated air is going through motor (I think it ignores this during starting)

While you are in a sensor changing mood, check that you don't have the old head temp sensor. The old ones have only one pin and use the motor for grounding.

Old head temp sensor connector (note single pin):


I would recommend fixing the stripped out ground. A heli coil or (if you can borrow a kit) a Timesert. Bad grounds mess up the flow of the motor's Chi energy.

-Chris

[rob911]

Chris
Thanks for tip on ground point. Great pictures BTW, wil include some of my damaged wiring later.

Agree that dissy cap is probably OK - guess that Bentley book is a little inaccurate in places.
Might have to bite the bullet and buy both ref and speed sensor - hope that my bandage repair will let me get out for a drive at least. It really pains me to replce a component unless I know it has failed. It seems that only an oscilloscope can reliably test these sensors. Anyone got one in my area?
Luckily I do have a newer style twin wire CHT fitted.

[rob911]

A-ha. I have found the problem.
My speed sensor is damaged, the end has been dragging on the flywheel. Will post pictures later.