M&K headers install now car is going lean
 

I did a m&k headers and gt3 exhaust install on my 81sc with stock cis, planning on doing another thread on that, but I am now having trouble with the car going lean as the rpms rise but not at idle. I have a wide band o2 sensor with the stock exhaust and it wasn’t exhibiting this behavior before the install. Here is a link to a video, https://youtu.be/3wZfbk-qm5A.

Post install I have:
- adjusted the mixture and idle screw using the lifting the air plate method and with the wideband o2 sensor.
- tested the cis with my pressure gauges and am within spec for a 090 wur, 75psi start, 21psi cold, 50 psi running @ 1:05 to warm.
- verified that the aav is opening and closing properly.
- noticed that removing the oil filler cap didn’t drop rpms so I replaced the cap. Still no drop in rpms.

So I am thinking that I might have developed a vacuum leak and my next step was to do the carb cleaner test. However I wanted to reach out to the board and get any opinions or experience in the arena.

The end game is an itb install from Al and efi/ignition from Rasant but I would like to drive it around a bit before that an enjoy the new exhaust for the next month.

Quick update, tried out the “trying to find vacuum leaks with carb cleaner” trick today. No changes in rpm when sprayed over fuel injectors, vacuum lines, and air box. Any thoughts?

Did you spray on the underside of the airbox as well? I found my (present) leak under the airbox on the 4-5-6 side.

All things being equal, if you make it easier for the engine to breath with a less restrictive header/exhaust set up, wouldn’t more air will be moved in the system creating a lean condition?

Yes but the cis system should be able to compensate. Think back the BAE turbo systems they used the stock cis system and compensated based off of the need though sometimes they added an additional injector.

too lean air/fuel mix
 

I have a wide band air fuel gauge on my 1982 911SC. When I remove the oil fill cap the system goes lean( as it should) - up to a hi 15% do to increased air being introduced into the mixture for a few seconds until the "system" compensates by enriching the mixture to get it back to the Stoichiometric fuel air ratio at around 14.7 at idle or cruise as it is designed to do. The issue is that the o2 control system can only compensate for a too rich or lean mixture within certain limits of around +/- of 20%.
Bottom line is that if you have a significant leak of unmetered air it is probably making your mixture leaner than the system can can react/adjust for.
Just my thoughts.

Dave

Somewhat apples to oranges comparison, but when I put SSI's and a sport muffler on my old '86 Carrera back in '88, the car felt flat in the 3500-4500 range as I recall. Ran it on the old Autothority dyno and the AFRs proved what my seat of the pants said, very lean in that rpm range. Took some custom software tuning to take best advantage of the new freer flowing exhaust, picked up hp and torque by electronically bumping up the fuel delivery in that range. Motronic was compensating but not enough range to get full benefit of the new exhaust until new software was burned.

The vid shows AFR at 3K around 15. Are you saying that's lean ?

What does it do under load?

Thanks for the dialog. I have not had a chance to check for leaks under the air box but will report back when I do.

Wanted to note that I have unplugged the factory o2 sensor, which I have read some people recommend on my year, and installed the wide band in its place.

Under load the car behaves the exact same. It seems to be under bursts of throttle that it will go leaner, sometimes off my gauge, than if I am cruising at a constant rpm. I am not driving the car much and haven’t ventured higher than 4K in the rpm band as I am concerned about damaging the motor.

Another data point is that I can go richer at idle but the same rising lean condition occurs. It seems to be a linear curve though so I could bring idle to 12 and I would see it rise to 16 as I go up in rpms.

I don't see a vid, sorry. And as I said- apples to oranges. My point is that headers will change the AFR, whether the CIS or Motronic can compensate or not is what I was attempting to convey.

OP has a youtube link. See it ?

I was just pointing out that AFR looks fine at the steady 3K RPM. It would be interesting to see a video running under load to 3K.

Here is a link to a video of it running under load, https://youtu.be/J61_HlNbAl4. Put my order in for the Rasant setup...