3.2 Carrera Hot Start & Surging Issue

[polizei]

Hey Guys,

I've slowly been chasing down an issue with my '87 Carrera. Thanks to those of you who have helped on other threads.

Issue Summary

The car runs beautifully right up until oil temps reach 185-195F. At this point, the engine just cuts out. Normally this happens when I've come off the throttle when approaching a turn or coming to a stop. Then the engine won't restart. If I crank for a few seconds, I can occasionally see smoke black smoke coming out the tail pipe, but it won't start.

After letting the engine cool down for a few minutes, I can normally get it to restart. When I then accelerate for maybe 5 seconds, the RPMs surge and then cut out, surge and then cut out, etc. After a few seconds of this, the engine completely comes out again. During this surge/cut-out stint, some black smoke is visible.

The engine really needs to cool for about an hour before the car is drivable for a decent amount of time.

Diagnosis Summary

After breaking down initially last night, my wife cranked the engine while I confirmed that I had spark directly from the coil by connecting a spare spark plug to the wire leading to the ignition distributor. I did not check for spark on any of the wires leading to the spark plugs.

Also, a fellow local Pelican with a 3.2 Carrera installed my ECU in his car. He drove his car until the engine got hot, but he didn't encounter any hot start issue.

Parts Replaced since onset of issue

New Battery, Negative Battery Cable, Transmission-to-body ground strap and fuel pump check valve. I realize that it's not efficient to replace parts by guessing. That's why I'm here

I'm looking for guidance on what to diagnose next. I appreciate your help!

[Techno Duck]

I would check the cylinder head temp sensor.

[polizei]

Originally Posted by Techno Duck

I would check the cylinder head temp sensor.

Do you have any guidance on how to test it? Thanks

[pmax]

Crankshaft sensors.

[polizei]

Originally Posted by pmax

Crankshaft sensors.

Is there a way to test this before replacing? Thanks!

[pmax]

Originally Posted by polizei

Is there a way to test this before replacing? Thanks!

You can measure them from the DME pins.

84 911 Carrerra 3,2 no start

Or find a way to cool them and see if the car starts.
However, on reading your post again, you have spark so I don't think these sensors have failed.

Your symptoms sound more like the CHT.

Good luck !

quote:[Lorenfb]
911 carrera 3.2 No Spark
"If you have all of the previously mentioned, check the resistance on your CHT sensor"

1. The temp sensor has NO effect on the lack of spark.
2. The engine will always start with a bad temp sensor, i.e.;
a. if open and a cold engine, the engine will start and eventually
make black smoke running rich,
b. if open and a hot engine, the engine will have very hard starting
and immediately make black smoke and eventually die,
c. if shorted and cold, the engine will have hard starting and possibly
backfire when hard accelerated,
d. if shorted and a hot engine, the engine will start normally.

Bottom line: Don't waste time with the temp sensor for definite
NO-start condition.

[stlrj]

Originally Posted by Techno Duck

I would check the cylinder head temp sensor.

That's probably what it is and testing it is so easy. Just disconnect the CHT where it plugs in next to the blower and you will see how you can duplicate the symptoms...hard start and black smoke if it catches.


Cheers,

Joe

[gomezoneill]

Reference sensors on the flywheel. When they get hot they fail the heat causes expansion and the circuit opens.

[polizei]

Originally Posted by stlrj

That's probably what it is and testing it is so easy. Just disconnect the CHT where it plugs in next to the blower and you will see how you can duplicate the symptoms...hard start and black smoke if it catches.


Cheers,

Joe

Just to clarify, are you saying that it's OK to run the engine with the CHT sensor unplugged? If so, I'm assuming I should drive my p-car until I begin experiencing the symptoms. Then unplug the CHT sensor and try restarting the car. Is that right? That sounds too easy!

[Tippy]

Yes, no problem. You can unplug too. I believe when you unplug, the computer thinks it's below 0F outside. It should run really pig rich.

If you plug back in and it runs fine, CHT is fine.

[polizei]

Update

I got the car out last night and got the engine hot with the CHT sensor plugged in. As usual, it surged and then stalled around a slow turn and wouldn't restart. I then unplugged the CHT sensor and tried to restart the engine. The engine cranked and cranked but wouldn't start - just like when the CHT sensor is plugged in. I got a little bit of black smoke followed by even more white/grey smoke.

Afterwards, I read on another thread that you're supposed to use a paperclip or something similar to jumper the two exposed pins at the connector for the CHT sensor in the engine bay. I did not jumper the pins. Should that matter? Does this rule out the CHT sensor as being the culprit or not?

A few more observations

When I initially took the car out last night, I drove around and got the oil temp up to 170F and then returned home to let it idle in the driveway. My hope was to have it "break down" in the driveway rather than on the side the road. Oil temps rose to 200F, but the engine wouldn't stall out. Historically, the engine stalls out and then refuses to restart with oil temps between 180 and 200F. I tried driving it up and down my driveway (<10 mph) but still couldn't get it to stall.

Then I took it out on the road, and within about 3 minutes, the engine surged and then stalled. Specifically, I got the revs up to about 5K in 1st and 2nd gear. Then I did a slow U-turn, and when I started to accelerate coming out of the turn, the engine surged and stalled.

The key thing I'm piecing together is that in order to trigger the issue, the engine needs to have been under load recently; however, it only begins to 'act up' on initial acceleration (<2K RPM) coming from a dead stop or a slow rolling stop. The issue has never reared its head under heavy acceleration or driving at speed. I hope this helps.

I appreciate any guidance on next steps!

[198491132]

Andy, have you tested the sensors and AFM functions with an ohm/voltage meter? Simple tests, if you have not done so. CHT, speed sens, ref sens, AFM butterfly can all be tested in a half hour. Grab the Bentley manual and try these few simple tests. Cap good? Rotor?

Good exercise, even if you don't find the issue just yet.

Dave

[stlrj]

" I did not jumper the pins. Should that matter? Does this rule out the CHT sensor as being the culprit or not?"

Had you jumpered the pins while it was acting up would have told you if the CHT was the culprit only if it started to run right at that point. Otherwise you have not ruled anything out since your test was incomplete.

Joe

[A horse with no name]

Originally Posted by 198491132

Andy, have you tested the sensors and AFM functions with an ohm/voltage meter? Simple tests, if you have not done so. CHT, speed sens, ref sens, AFM butterfly can all be tested in a half hour. Grab the Bentley manual and try these few simple tests. Cap good? Rotor?

Good exercise, even if you don't find the issue just yet.

Dave

+1

The speed and reference sensors themselves are identical. Try reversing them at the bellhousing, and of course, at the other end where they plug in. 'IF' it is the speed sensor that is failing, the car will not start.

By the way, purchase the BMW ones as they are far less expensive to purchase, and they are the identical part; their wires are a tad bit longer, which is a plus. As a preventive maintenance issue, I recently replaced mine with them. They were/are used on a variety of BMW's in the mid '80's.

Bosch BMW reference sensor # 0 261 210 002

[cabmandone]

In the first post you say you checked and had spark. That to me would rule out the speed and reference sensors but I'd also check to see if you're getting fuel. It sounds like you are since you mention seeing smoke when cranking. Again that would rule out the speed and reference sensors to me as well as the DME relay. Unplug your CHT and try to start the car when it's cold. My guess is you'll see the same symptoms you're seeing when the car gets warm. It will crank and crank but will not fire.

[gomezoneill]

Originally Posted by cabmando

In the first post you say you checked and had spark. That to me would rule out the speed and reference sensors but I'd also check to see if you're getting fuel. It sounds like you are since you mention seeing smoke when cranking. Again that would rule out the speed and reference sensors to me as well as the DME relay. Unplug your CHT and try to start the car when it's cold. My guess is you'll see the same symptoms you're seeing when the car gets warm. It will crank and crank but will not fire.

You can't rule them out they are almost always intermittent.

[ischmitz]

It's your CHT. At operating temp it needs to read 90 ohms. If open or unplugged it richens up the mixture. Put the paper clip into the harness connector and it'll start when hot as the paper clip simulates engine at operating temperature.

Ingo

[cabmandone]

Originally Posted by gomezoneill

You can't rule them out they are almost always intermittent.

You can rule them out when he mentioned he had spark when the car was not starting. IF you have spark, your Speed and Reference sensors are fine. Ingo appears to be pointing to the CHT and that's where I'd go too since it's a temp related issue.

[ischmitz]

This one is the exact same issue. It starts, runs rough and eventually dies when it get to operating temperature. It has spark but the CHT is open (measured). Stick that paper clip into the connector going to the CHT and it'll run fine when warm.

No start for spring

Ingo

[polizei]

Originally Posted by 198491132

Andy, have you tested the sensors and AFM functions with an ohm/voltage meter? Simple tests, if you have not done so. CHT, speed sens, ref sens, AFM butterfly can all be tested in a half hour. Grab the Bentley manual and try these few simple tests. Cap good? Rotor?

Good exercise, even if you don't find the issue just yet.

Dave

Dave,

No, I haven't, but thank you for the direction. I'm a bit intimidated by electrical diagnostics, as I haven't done anything like this before. Thankfully, my wife is a generalist engineer, and she was able to provide me some guidance.

It took me a little more than 30 minutes (I won't say how long ), but here's what I found. All readings were taken on a cold engine with ambient temperatures between 40 and 55F:

1. Speed and References Sensors - 945 Ohms between Pins 1 & 2 (in tolerance per Bentley). Ohmeter (Innova 3320) read "Open Loop" when measuring between pins 2 & 3 and 1 & 3. I believe all of this is OK, as Bently calls for >100,000 Ohms. There were no visible issues with any of the wiring adjacent to the bell housing or in the engine bay.
2. CHT Sensor - 4,170 Ohms (in tolerance per Bentley)
3. Distributor Cap - All contact pairs measured either .95 Ohms or "Open Loop" on my Ohmeter. Bentley says to replace the cap if you get readings other than 0 Ohms, so I've already placed n order with Pelican.
4. Ignitor Rotor - I forgot to write down the Ohm measurement, but I confirmed it was in tolerance per Bentley.
5. AFM Sensor - Pin 3 connector from engine harness reads 4.33 Volts. Bentley calls for "approximately 5 Volts". Any concerns? Also, I measured resistance while moving the AFM flap and observed that resistance continuously changed under movement (correct per Bentley).
6. Ignition Coil - I forgot to writ down the Ohm measurements, but the readings between Points 1 & 15 and Points 1 & 4 were in tolerance per Bentley.

Originally Posted by stlrj

" I did not jumper the pins. Should that matter? Does this rule out the CHT sensor as being the culprit or not?"

Had you jumpered the pins while it was acting up would have told you if the CHT was the culprit only if it started to run right at that point. Otherwise you have not ruled anything out since your test was incomplete.

Joe

Thanks for confirming Joe! Once I get the new distributor caps installed next weekend, I'll taker her out again and run the test per your description. I'll also measure resistance at the CHT sensor a) when the engine's hot but before it acts up and b) when it is acting up.