87 911 Carrera 3.2 Hesitation

[johnman001]

I've been scratching my head on this one, so I'm looking for some outside opinions.

I have a 1987 3.2 with stock motronic installed in a 914. The car has run flawlessly for many years. The other morning, I experienced a no start. I felt the air stabilizer and it was buzzing, but during cranking, I could not hear the fuel pump at all.

Keep in mind that this is installed in a 914 so the engine is installed backwards compared with a 911.

I changed the DME relay because I know it controls power to the DME as well as the fuel pump. Car started almost immediately and ran well for the most part, but it started and continued to experience a loss in power (as if it briefly was running out of gas) on hard left hand turns. It did this only when turning left. Not when traveling straight or turning right. This was intermittent and inconsistent, with the exception of only occurring during left turns.

The gauges and tachometer are not affected when the loss of power occurs.

I need to drop the gravel pan and check the fuel pump wiring as well as fuel line routing, but the poor running would clear up immediately so I'm not sure if fuel line routing would really be the culprit since this hasn't been touched in a few years.

I have not researched enough, but if someone knows what controls the fuel pump relay, that information would be beneficial to me. I was considering hard wiring the fuel pump to a switched hot so that I know that it always is running but I am wondering what actually turns on the pump when the engine is running. I was thinking it was tied into the AFM somehow, but I have not verified that. If the fuel pump is controlled from the AFM, that might lead me to think that for whatever reason turning left might have some affect on the AFM, or that I may have a vacuum leak that only opens on left hand turns (and might bypass the AFM).

I know I have a bunch of items to check, but would like to hear other suggestions. Thanks!

Things for me to check:

All connections - grounds, connectors, clips on connector housings
DME Relay
Altitude Sensor
DME harness Plug at DME
AFM
Full Throttle Switch/Throttle Position Switch
Idle Microswitch
Intake Manifold Boot
Fuel pump pressure
Fuel Filter
Fuel pump wiring
Fuel pump fuse (for clean connections)
DME Harness
Fuel Injector harness & Connector
Flywheel and Reference sensors
Cylinder Head Temp sensor
Coil
Plug wires & Distributor Cap

If all this checks out, I'm not certain where else I can look with the exception of cracked/cold solder joints in the DME itself, but I'm uncertain why this would only occur during left hand turns.

[steely]

For the FP relay control, first, the starter power is routed to the DME and thru it to the DME relay.
The DME relay provides power to the rest of DME when ignition is on. (I corrected this)
The DME determines if the engine started by the ref sensor.
If there’s no engine rotation, the DME will not hold the FP portion of the DME relay closed.

So items involved electrically are the DME, relay, fuses, and ref sensor. The AFM isn’t in this path.
Your starter’s ok b/c the DME turned on (your ICV vibrated).
This leaves the relay, the FP and its fuse, wiring, to a certain extent the DME, and the Ref sensor.

I’d swap the relay w/ a new one and inspect the DME plug and the sensor, and clean the fuse #(?)
You could jumper the DME relay to verify the associated wiring & FP work.

The left turn issue is odd, but the above would be my first suggestion, and then on to the other items from list you have. Am sure you'll get more feedback.
HTH, Good luck, and let us know.

[johnman001]

So, I agree with part of your answer.

With ignition on, the DME half of the relay turns on and powers the DME which in turn allows the ICV to operate (vibrate). I believe this to be independent of the starter.

What you are indicating is that the reference sensor pulse (1 pulse per rev) is what the DME is looking for in order to turn on the pump side of the DME relay. That makes sense. Perhaps I was thinking the AFM was similar to the L-Jet where there had to be air flow in order to operate the pump.

Can anyone else verify that the reference sensor pulse is what the DME looks at to control the Fuel Pump section of the DME relay? I can certainly change the reference sensor, but I can't keep an oscilloscope connected and re-create the situation (hard left hand turn). I can also be sure to check the reference sensor cable from sensor to plug. I was also considering installing an indicator light to see if the DME is dropping voltage to the FP side of the DME relay under hard left turns.

If the DME is looking at reference sensor pulses, then any time the engine is running, the fuel pump should be running so hot-wiring the pump should really cause no ill running.

I had been under the impression that the reference sensor was only used at start-up, but then again, I have no idea of a sequence of operation for the DME.

Thanks for your help and please keep the comments rolling in. I have checked most of my connections and will continue troubleshooting as time permits.

[techman1]

Check the state and connection of both (I believe there are two) for the hall effect sensors on the flywheel. I believe the left hand turns may be pulling the wire in such a way as to lose connection.

[steely]

Originally Posted by johnman001

With ignition on, the DME half of the relay turns on and powers the DME which in turn allows the ICV to operate (vibrate). I believe this to be independent of the starter.

True statement above.
But the initial pick of the FP relay (DME pin 20) is done with power from the starter (DME pin 4).
Unless of course it was modified for the 914.

[Quicksilver]

Usually you get one failure at a time or a failure cascades to cause a secondary failure. Looking at this I would try to focus on root causes.

There is nothing about the DME system that should have any lateral G sensitivity. Because of the "left turn" issue I'd expect that to point to something. First thing I'd look at is ground issues. I'd look at the various grounds and make sure they are clean and sound. Other parts of the wiring harness should be checked too. Maybe something is shifting or it is getting pinched.
A DME solder issue could show G related symptoms (but kinda unlikely).

Other than that I'd go through my whole TEST, TEST, TEST mantra. Test fuel pressure while running. Test spark. (Full throttle at higher RPM is the hardest to get the spark to jump.) Get an oscilloscope and test the reference sensor signal. Plus a search will show you how to test almost any part of the system.

[johnman001]

Originally Posted by steely

True statement above.
But the initial pick of the FP relay (DME pin 20) is done with power from the starter (DME pin 4).
Unless of course it was modified for the 914.

I kept all the DME wiring per the factory workshop manuals when installed in 91'. I believe that there is a yellow wire connection so it would see the starter signal while cranking.

Thanks for the replies guys, please keep them coming at least until I figure out what the culprit is. It may take a few days before I can spend more time on it. Scouts and such take up some time in the evenings. They say "an hour a week". I find it is a bit more time consuming than that.

[johnman001]

A little more information on this one:

I was checking through all the items I had considered and cleaned ALL connections to be found for the DME, engine sensors, fuel pump, etc.

I checked fuel pressure while running and all items seem to check out.

The only questionable item I came across was the connector for the cylinder head temperature sensor. The wire clip is still in the plug, but the top ear is missing off of the sensor side of the plug. I ordered a new cylinder head sensor so that I know it will stay plugged in.

Just for grins, I unplugged the cylinder head temperature sensor and it lead to a no-start condition. I started it and wiggle-tested the connection and it appears to gain RPMs and act unusual if unplugged. I am assuming from this that it is leaning out the mixture drastically.

Can anyone confirm how the cylinder head temperature sensor affects the running of a Motronic system? (would infinite resistance (open) richen or lean out the mixture, and conversely, would no resistance (short) richen or lean out the mixture)

This could certainly be a cause for loss of power and could account for only occurring during hard left hand turns on track and under power.

Unfortunately, the silly plug didn't simply fall off so I would know with any certainty that this was the root cause.

Thanks for looking once again.

[johnman001]

Well, I did find this, so unplugging may cause it to be rich and shorted would cause lean condition. Either way, there would be a significant loss of power.

[johnman001]

I changed out the Cylinder Head Temperature Sensor and the car appears to run well. I'm not certain I will be able to prove one way or another that it is fixed until I am able to drive it again on track (in a few weeks).

I did find a couple of things:

• The old sensor was single wire original sensor
• The new sensor is dual wire (provides a direct ground reference to/from DME)
• BOTH locking ears were sheared from the connector
• The wire at the head may have been subject to wire chaffing
• The plug (harness side) is now clean, tight, and locked on with the retaining wire

I am still uncertain why this intermittent loss of power was only occurring during hard left hand turns, but the harness side plug may have been poorly inserted into the sensor connector, or the sensor wire might have been shorting to the engine tin right at the sensor.

I do have my fingers crossed, and will post back after trying it out under track conditions.

[steely]

Could it be a "g-force ignition fault"?3.2 stumbles only when turning right

[johnman001]

I'm not ruling anything out, but I do know my REF and SPEED sensors are in good condition (checked both of them with my oscilloscope and the insulation is good).

I do remember that thread from 5 years ago and I did check all 6 spark plugs to make sure the little barrels are tightly screwed onto the tips of the spark plugs and I checked for arcing/sparking by running it at night, but found none. I do have another spare DME relay I can swap in, to make sure there isn't a problem with the new one I just put in at the last track event, but I'm doubting this is the culprit due to the problem remaining even with the fuel pump hot-wired. Since the problem remained, I returned the wiring to stock (DME controlled fuel pump).

Thanks for any and all suggestions. Intermittent problems are sometimes the toughest to solve.

[johnman001]

I'm not ruling anything out, but I do know my REF and SPEED sensors are in good condition (checked both of them with my oscilloscope and the insulation is good).

I do remember that thread from 5 years ago and I did check all 6 spark plugs to make sure the little barrels are tightly screwed onto the tips of the spark plugs and I checked for arcing/sparking by running it at night, but found none. I do have another spare DME relay I can swap in, to make sure there isn't a problem with the new one I just put in at the last track event, but I'm doubting this is the culprit due to the problem remaining even with the fuel pump hot-wired. Since the problem remained, I returned the wiring to stock (DME controlled fuel pump).

Thanks for any and all suggestions. Intermittent problems are sometimes the toughest to solve.

[steely]

That situation would bug me - keep at it and good luck, sounds like you're on the right track!

[uwanna]

John,
I know you said you checked all the connections, but have you done
this? There's a ton of DME connections at the ECU.
Here's a post I wrote in a previous thread:

I developed a problem with my 964 3.6 motor. Had trouble on cold start,
but ran Aok when warmed up. Turns out to be a bad connection to the Cyl temp sensor in the large cammed connector block to the ECU. Sprayed DeOxit D5 on the the pins in the connector and ECU, problem fixed. Also had a bit of unstable idle and a bit of hesitation on part throttle. Sprayed all the connector pins and ECU pins and seated the cammed connector several times, all problems gone! I guess after all the years the pin connections had developed some oxidation/resistance. I know you have a 3.2, but same type connections. Give it a try, can't hurt.
DeOxit D5 is awesome stuff, removes any oxidation and corrosion and protects from
further corrosion. Used to work in the electronics biz, and it was the go-to contact cleaner.

[johnman001]

Thanks for the thoughts.

Contact cleaner was my first attempt.

I did check the DME plug at the unit itself and went through it sensor by sensor and also checked the connector on the DME. I checked that mainly to make sure I had solid grounds, but while I was in there, I went ahead and checked it all.

NOTE: I call harness side Plugs and unit side, Connectors.

If I still have problems, I think I may have narrowed it down to something internal to the DME such as cracked solder joint(s), or similar issue inside the new DME Relay.

If I have time, I may go ahead and check those out prior to the next track day. I am hopeful, though, that the problem was the cylinder head temp sensor.

[johnman001]

After numerous attempts to narrow this issue down, including one-at-a-time swap out of parts, my issue was determined to be the AFM. When I swapped in a known good AFM, all the problems disappeared, and the car responds as it should.

Once I knew what part was at fault, I went about to correct the issue. Upon disassembly of the AFM, it was found that the wiper had worn through the conductive material on the ceramic substrate. Following along with many of the articles written about this, I bent the wiper arm in order for it to have new wear surface for the contacts to follow.

Several of the write-ups show testing this device with a 9v battery and an analog volt meter. I purchased a plug and contacts from Digikey and now made up a tool that can plug into an AFM for quick testing.

I have no idea why this symptom manifested itself around certain direction corners, but it did. This perticular AFM was purchased back when they could still be purchased new about 12 years ago due to the same problem, but with different symptoms. I kept that old AFM and have modified that one as well so that it too has new material for the wiper arm to follow.

I thought I would post a follow up since this problem was at long last solved.

[Trakrat]

" was found that the wiper had worn through the conductive material on the ceramic substrate."

Yep... That's what I would have guessed.
I have a 3.2 and it had the same exact issue where the engine seemed like it would hesitate at certain rpm range.