A/F ratio swings back and forth
 

My car has always had a strange issue, the A/F ratio swings back and forth from rich to lean, making the swing about one a second. It starts doing this as soon as it gets warmed up and does it all the time except at full throttle. It did this way before I installed the turbo. It's keeping me from cranking the boost up, I don't want to get aggressive until i feel it is running perfectly.

I've changed the O2 sensor and made all kinds of mixture adjustments, I've replaced the injectors, o-rings and sleeves, I am pretty confident I don't have any vacuum leaks.
Any suggestions?
The only time it bugs me is at part throttle steady cruising speed, I can feel a slight surging.
I figure I'll prolly have to buy a CIS test pressure kit and maybe an exhaust gas teater to find the problem unless someone has a magic trick that is cheaper and easier. Tanks
SmileWavy

Usually happens when the idle co% is not set to spec (which also corresponds to richer/leaner mixtures across the RPM range to an extent) and the lambda brain is trying compensate for that incorrect mix. NA guys usually either set the idle co% to spec to try and correct the issue, or unplug the O2 sensor and richen the idle CO% to a more appropriate setting (2-3%). Don't know if unplugging the O2 sensor in your case is an option.

Have tried driving around with the O2 sensor disconnected?

Hi,

This is perfectly normal. A normal (narrow-band) O2 sensor switches between a low voltage (lean) and high voltage (rich) right at the stoichiometric point of 14.7 AFR. At 14.7 AFR (Lambda 1.0) the voltage is ~0.45V. The lambda controller richens up when lean (O2 voltage < 0.45V) until the O2 sensor reads above 0.45V (rich) back and forth, normally 2-3 times/second. This is just how it's supposed to work. By adjusting the idle mixture correctly you can minimize the swings. At WOT the O2 sensor is ignored by the controller because the desired AFR is richer than 14.7 and the O2 sensor can't give any useful information away from 14.7 AFR.

Regards,
Klaus

OK, it sounds like I just need to get the mixture closer to what the lambda brain wants it to be and that will make the swings less noticeable and severe? Makes sense.
I think what I'll do is to disconnect the O2 sensor and adjust it to what I think it should be based on my A/F meter, then reconnect it and see how it reacts. I may buy a gunneson gas tester anyway, it sounds like that's the best way to get it right.
I have it rigged up so that at 3 psi positive manifold pressure the lambda wire gets grounded so that it goes richer to compensate for boost, that works sweet. In fact it goes rich enough to allow for higher boost levels, but not untill I get this other issue taken care of.

Hi,

If you ground the O2 with boost you might run too rich. Instead ground the WOT switch wire. This is also connected to the cold enrichment thermal switch on the right chain housing. The lambda controller uses that switch to determine when to ignore the O2.
By shorting the O2 at part throttle before the WOT switch kicks in you make it too rich for the turbo to spool up right. When the WOT switch then kicks in the lambda controller ignores the O2 (even a shorted one) and goes to open loop fueling anyway.

Regards,
Klaus

Hi,

Look in this thread, 2nd page bottom for another way to control AFR in closed loop, even under boost:

http://forums.pelicanparts.com/showthread.php?s=&threadid=152519&perpage=20&highl ight=lm1&pagenumber=2

The lambda wire gets grounded at 3 psi rising, which will never be obtained unless the throttle is wide open. My car doesn't build more than 1 psi boost unless the throttle is WOT.
I rigged it up that way to add waaaay more fuel under boost so I could run higher boost levels without my combustion temps going high. It works well, although it could be a little too rich as you suggest. That just means I can run more boost ;) just kidding.
Thanks for your help, I'll go to that thread and study.