3.2 Suddenly Running Lean after warm-up
 

Looking for suggestions as to what could cause a 3.2 to start running a few points leaner than normal after warming up.

This happened for the first time two weeks ago, and went away after pulling over and shutting the engine off for ~10 minutes. Ran normally for two weeks, and then happened again last night and this morning.

Thanks,
Graham

I would check for a vacuum leak.

How lean? and how do you know it's lean?
Few thoughts:
- Bad O2 sensor? unplug the O2 does the issue go away? Sometimes the O2 sensors fail and put out 1vdc all the time, if this happens it sends a false rich condition to the DME and DME tries to fix the problem by leaning the mixture.

- Could be faulty injector(s)?

- Fuel pressure out of wack?

Idles at 13-14 during warmup, then 15.5 or so when warm.
When driving (cruising at low load) it will be 17+, and spike down to 13 for a fraction of a second when i apply some throttle, but then back up immediately.

O2 Sensor is not plugged in. I have one of the AEM wideband A/F meters in it's place. So the DME is in open loop mode. It's actually one of your DMEs, Sal, with the '89 chip.

Injectors were new last year. Could still be a problem though... will look at the left bank (since that's all the meter is measuring)

Fuel pressure regulator was new also... will try to test, but i think my meter is broken.

New parts during engine rebuild last year:
Injectors
CHT sensor
Speed and reference senders
Fuel pressure regulator (not damper, though)

Checked gas tank, didn't seem to have a vacuum. No change after filling tank this morning.

Vacuum leak was my first thought as well, though I can't seem to find any so far.

OK, what's the AFR at idle? if it's also very lean (above 15.0) I suggest you do a cyl balance check on the LHS that has the WBO2.

With it at idle monitor the AFR reading and unplug each injector (1-3) one at a time. Basically this kills that cyl and the AFRs must lean out significantly but most important is that each of the 3 are within 5% of each other. If you have a dead injector you'll see it right away because when you unplug it nothing changes on the AFR gauge.

Also listen to each injector at idle with a long screwdriver, place the blade (tip) onto the top of the injector and then put your ear on the handle. Be sure they all sound the same.

A dead cyl that does not fuel and/or have ignition becomes a O2 pump and causes false lean reading on the WBO2 gauge.

Alright, I'll go try that.
Forgot to mention that the injector harness was new also.

Interesting! I have headers and one side has always run more lean than the other but then the O2 sensor is only on one set of headers, hmmm

So run my suggested test on that side that has the WBO2
Then swap injectors LHS with RHS and repeat, poor mans injector flow test at small pulse widths.

Or put a WBO2 in the tail pipe and you'll see all 6 cyls, then do the test for all 6.

You don't know how many times I've heard 'but that part is new' do not assume new or prior tested means it's still good. Approach all diagnostic procedures without assumptions, of course it's fair to identify that the part is new vs old but just be careful.

Here's a recent story about parts that burned me and wasted 2 weeks! I bought brand new Bosch injectors for some testing I'm doing, they came in the Bosch box even with the special caps on each end, they looked 100% new! I went to try and model them in my system and things were really off, they seemed to not be at the specs they should be. Flowing way to much fuel, about 20% to much! After working on the issue for hours assuming the injectors were good I decided to simply buy another set, Problem solved! But what the heck was wrong with that first set? I decided to inspect them closely under a magnifying glass and discovered they had been DRILLED! Yup, someone decide it would be a good idea to drill the 4 holes that control fuel delivery! I returned them and the seller claimed he had no idea.

You never know what gets returned and put back on the shelf.

Well, it's not the injectors or the harness... will check fuel pressure next and see if i can't find a vacuum leak.

Hey OP...did you ever figure out your issue? I have the exact same symptoms so was wondering what ended up being your problem.

Thanks!

Sent from my SM-S911U using Tapatalk

Have you checked for induction leak? I’m not familiar with Porsche power plants but have had similar problems on Lycoming aircraft engines. Throttle body to manifold gasket failures can be intermittent due to temperature-caused expansions and contractions. Further, cracked housings, loose fasteners or worn throttle plate bushings sometimes will allow the source of air leaks to open and close with the movements of rods, bellcranks and push pull cables.

With the engine running, the simple trick of flowing propane around the induction system components might help you locate an air leak. Shake linkages and components during the test. Be careful, the engine bay/pressure cowl could allow gas build up. It shouldn’t take much propane to note an increase in rpm caused by the suction of an induction leak.

Wear safety glasses and some sort of fireproof mask if you have to stick your head into the engine bay. I usually wear safety glasses and a full face helmet (with the visor closed) when flash or shrapnel may occur.

Again, the rpm increase will be almost instantaneous when a small volume of propane is ingested. No need to fill the engine compartment (or the helmet) with propane.

I didn’t find a single smoking gun for this, but I corrected a few vacuum leaks and replaced my injectors with some modern ones from five-o Motorsports, and haven’t had the issue since.
I tried the propane method above without luck. I’d suggest buying or borrowing a smoke tester (I built one from a paint can, and it is a little finicky).
I suspect there was an electrical component to the problem as well - I keep finding wiring harnesses with insulation worn through.

Thanks for the response. I'm planning to smoke test for leaks. I guess I could proactively replace injectors and o2 sensor too and see what happens.

Sent from my SM-S911U using Tapatalk

I'd start with the smoke test, and then maybe go through the sensor tests in the Bentley manual. The O2 sensor doesn't do much... you could just disconnect it to test.

I have a supercharger set up with a custom ecu which takes o2 into consideration for AfR gauge...have maybe a hunch that it's actually running rich and not lean and the sensor is bad in exhaust but definitely will start with leak test. Thanks

Sent from my SM-S911U using Tapatalk

Like said above , check manifolds for leaks. First tighten all manifold bolts to 18 foot pounds. They can come loose over time. If you still have the problem must do tests and find the cylinder that has the leak. I just had your problem and more. It was a single manifold gasket installed upside down. Idle and everything went back to normal after repair.

Had a similar issue. Did a few smoke tests and the brake booster line was loose.. fixed it. drove fine for a few days and then went super lean again.

Did a smoke test and found my oil cap O-Ring left the chat.


<iframe width="467" height="830" src="https://www.youtube.com/embed/uZYtqu36L7U" title="1987 Porsche 911 - new Oil Cap smoke leak" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>

Cheap and good smoke machine (You will need a 12V source, I just have a small motorcycle battery)
http://tinyurl.com/jm438bxr

Smoke juice
http://tinyurl.com/3jcjsvwp

Ordered a non-fitting cap, ended up just flipping the o-ring and it fixed it.

Good luck

-Ian

Thanks for the links, Ian. That would have worked a lot better than the paint can contraption i built. Might pick one up and check my engine again.