Joe6pack

Ok, tuning gurus, I am curious. I recently replaced my cat with a cat bypass. I am still running the oxygen sensor. I also purchased an Innovate LM-2 (greatest thing ever). So in playing around with various lambda settings, I noticed that the initial setting does make a difference. I am still trying to figure out exactly why, though. Here's the thing. If you set the ratio to .95 lamda (5% rich) using the LM-2 and then plug the O2 sensor back in, the DME forces the car back to 14.7 (1.0 lamda). Yet the throttle response does seem to change. I can't really figure out exactly why unless it has to do with subtle mechanical differences. I think the DME always tries to maintain 1.0 lambda except under certain conditions (wide open throttle). If it is alway pushing for 1.0 and reads the AFM as the input, why does the inital AFM setting seem to make such a difference.

P.S., those of you who know, what are you running for your initial lamda setting?

theiceman

I think the initial setting is to get the car started .. based on CIS experience car starting is critical and to best accomplish this the intial CO setting is set up so the car will start without backfiring.
even though the motronic is somewhat different i am sure the physics involved in getting the car to start is the same so those other sensors can come into play .

Amber Gramps

I can't think of a single active board member that messes with this kind of stuff. Most of us let the Motronics and a good chip and O2 sensor do their jobs.

Double check yes, manipulate no.

Ed Hughes

Loren would have the best answer, I believe. He may weigh in.

I can deal with vacuum or physical issues, but I kinda concur with Doug. Truthfully, and I say this with all respect to SC owners, the Motronic system is the BIG plus for Carrera years, IMO.

rusnak

they say that the Motronic system has an "easier" (is that the word?) time leaning out than making rich. In other words, it can shorten fuel pulses better than lengthen them (pulsed ground). But you're right. For sure the dme wants .97 to 1.03 lambda. When you drive down the road and apply load, you really see this.

P.S. Joe, do you know offhand how many turns out you're running? Just a data point check, thanks.

Droops83

Joe, are you "setting it to .95 Lambda" by adjusting the idle mixture screw on the airflow meter? Or are you messing with the spring tension on the potentiometer/wiper arm? If you are merely adjusting the idle mixture screw, this mainly only affects the air/fuel mixture at idle and light load. When you turn the screw you are changing the size of the opening of an air passage in the AFM that bypasses the main flapper door. The more load the engine is under, the more the flapper door opens and the idle air bypass has less and less of an impact on the overall A/F ratio. I suppose that changing the idle mixture can make a small difference in part-throttle response, but not much. Nonetheless, even a 5% difference in overall AFR can have a noticeable difference in how an engine runs. Keep in mind that an otherwise engine will only run from an AFR of ~11:1 to ~17:1, and not very well near either of those extremes. 13.9 can be a big difference from 14.7.

If you are changing the spring tension on the flapper, don't unless you know what you are doing; if your engine is stock you are best off to leave it alone.

During normal engine operation, when the oxygen sensor is hot enough to generate a voltage signal, the DME uses this signal to manipulate the fuel injector pulse width however necessary to keep the overall A/F ratio at 14.7. The DME is also programmed to rapidly change the mixture from rich to lean (with an average of 14.7 AFR) as the three-way catalytic converter requires a relatively rich mixture to combat NOx emissions, and alternatively a relatively lean AFR to kill HCs and CO. This is why on a normally running Carrera 3.2 you will hear a very slight surging of the idle when warm with the O2 sensor plugged in.

The oxygen sensor signal is ignored at wide open throttle, which is signaled to the DME via the throttle position sensor. The stock DME is programmed to provide a pretty rich AFR under full load, likely in the 12.0 range. This is because excess fuel provides a cooling effect in the combustion chamber to help prevent detonation and/or preignition. The engine would actually make more peak power with a slightly leaner AFR, but the factory sets the WOT mixture conservatively to save the engine. This is why it is important for you track guys to periodically check your TPS function and adjustment to make sure the DME is seeing the WOT signal!

Joe6pack

Thanks all. I'm actually changing it using the idle mixture screw on the AFM and then adjusting the idle as necessary. There was an article in Panorama recently on how to do this. My understanding of what adjusting this screw does is that it changes the spring tension on the AFM. The Motronic reads the AFM flapper position and uses it as an input for fuel regulation. Since the flapper is passive, it is something that the DME must compensate for. That said, yes, I do think the Motronic always shoots for 14.7, thus my questions on why it makes such a difference.

There must be some sort of non-linearity or hysteresis in the system like Russ said. With the AFR adjusted richer, the engine seems to rev quicker and is smoother off idle and throughout the range. Maybe because the DME is trying to lean rather than richen. With everything set at 1.0 lambda, however, the idle is smoother. Probably because the DME isn't having to do anything. I know some run without an O2 sensor, but I'm not prepared to go there yet.

Thanks for the insight.

Lorenfb

"Or are you messing with the spring tension on the potentiometer/wiper arm?"

That's the answer! The responsiveness can be changed that way and still maintain
the cruising (steady-state) mixture. That's one of the pluses of still having an AFM
versus an MAF as used in 993 and later Porsches.

"I noticed that the initial setting does make a difference."

Changing the idle mixture has very very little effect on the responsiveness.

"There was an article in Panorama recently on how to do this. My understanding of what adjusting this screw does is that it changes the spring tension on the AFM."

You can't believe everything that's written there!

Droops83

Joe,

Re-read the first paragraph of my post. Adjusting the mixture screw only changes the size of the opening of the idle air bypass in the AFM. The only way to change the spring tension is to remove the black plastic cap that covers the potentiometer and do so manually. I would seriously hope that a publication such as Panorama is not spreading misinformation. Which issue was it? I have some laying around the shop and I'll check.

rusnak

Ischmitz has posted his conclusions about the DME feedback loop, and among other things, he seems to say that there is a continous feedback, only lacking when the O2 sensor is not hot enough. The dme does not know this (no fail check mode) , and may be erroneously checking mixture adjustment. I suspect that a richer adjustment at idle only makes a difference when the O2 sensor is cold. Once warmed up, the O2 sensor takes over, even at idle.

The feedback is continuous, not necessarilly linear unless you mean when plotted on a grid with a time axis.

Lorenfb

"he seems to say that there is a continous feedback, only lacking when the O2 sensor is not hot enough"

Not correct! Yes feedback, but not always in closed-loop, e.g. Especially during initial acceleration.

"suspect that a richer adjustment at idle only makes a difference when the O2 sensor is cold."

Only if the O2 sensor system can correct the mixture and the AFR is not outside the O2's
range of control, e.g. fuel pressure too high or bad temp sensor.

"not necessarilly linear unless you mean when plotted on a grid with a time axis."

Do what? Some more mis-info from Panorama?

Joe6pack

Ok, let me start over. My car just turned 25 which makes it emissions exempt in GA. So, I decided to go for a cat bypass since it is a simple change and is easily reversible. When I removed the cat, and replaced it with the bypass, the car ran great, but it was so rich it made your eyes water. Rather than take it to someone or randomly make adjustments blindly, I decided to invest in the aforementioned Innovate LM-2.

When I hooked up the LM-2, I found that I was actually running .85 lambda at idle. This is with the IAC bypassed and the O2 sensor unplugged. I think this is equivalent to 12.5 AFR which would be considered tuned for performance on a carbureted engine. The Panorama article I was referring to gave directions on setting the ideal factory idle on an '84-'89 Carrera. This is done by adjusting the screw on the AFM to get to 1.0 lambda (14.7 AFR) and then setting the idle to 880 rpms (with the IAC bypassed). I think I mis-spoke earlier when I referred to tightening the screw as tightening a spring. Adjusting the AFM is equivalent to adjusting the throttle plate on a carburetor. If you close the throttle plate, the engine idles richer. If you open it, the engine idles leaner. The difference is that the DME compensates by adjusting the fuel to get back to 14.7 AFR no matter what. Carbureted vehicles without O2 sensors don't have this luxury.

What I am talking about is initial settings less than 1.0 lambda. What I noticed is that when I went from .85 lambda to 1.0 lambda the overall character of the engine changed throughout the rpm range. I then set it to .98 lambda from 1.0 lambda and again, I noticed a difference in throttle response. I am somewhat hesitant to say better. What I am trying to understand is why. I think Loren may have addressed it somewhat above.

Loren - What do you mean by "Not correct! Yes feedback, but not always in closed loop, e.g. Especially during initial acceleration.". Under load is when I notice the biggest difference - off the line and part-throttle acceleration. It honestly seems to accelerate better and is more responsive and smoother with the richer initial settings. At higher rpms with the leaner AFR, the car seems want something when you give it part throttle. Does/can the DME only compensate at steady state? It definitely does not startup or idle nearly as well. Also, the exhaust note changes. The car overall runs much more quietly with at the factory 14.7 AFR throughout the rpm range. It seems to have more of growl with the richer settings.

Lorenfb

"If you close the throttle plate, the engine idles richer. If you open it, the engine idles leaner."

No, the throttle plate just allows more air and then it's the AFM that compensates
for the added air by signaling the DME ECM to increase the fuel quantity,
basically the AFR remains unchanged (within a small range).

Again, changing the idle mixture screw will not affect the responsiveness
off idle. Just look at the size of the air hole by the mixture screw in the AFM
compared to the size of the butterfy opening in the throttle body. Once the
butterfly opens, the hole in the AFM is inconsequential.

rusnak

Joe, you really need a second test bung so that you can run the O2 sensor, AND the LM2 sensor at the same time. It does sound to me as if you either have a bad sensor or a bad sensor wire. And I would NOT run without the O2 sensor.

Ed Hughes

You certainly take a step backward without a functioning O2 sensor in place. I'd imagine that no matter how well you tune it, your eyes will water to some extent, without a cat in place.