OK ok ok. Spark plugs didn't help at all. My MAF never went above 4.5V and still was cutting out badly in first gear. So I was about to drive the damn thing into a wall, since I am sure I could get more insurance for it than I could sell it for.
Then it dawned on me.
The Long Term Fuel Trim could be applied during open loop. So I don't have good OBDII software since the disk that came with my bluetooth module was snapped in half, but I found some trial software for my brother to watch while I did a pull, and sure enough LTFT was +25 all the way through to redline.
Now I know exactly what is going on.
So AEM saying that the PCM would eventually learn out your O2 modifications is kind of right, but it doesn't lean it out, it simply turns the STFT into LTFT and carries that trim into and through the open loop operation.
So when the PCM reads it's already rich table and adds 25% to it, it hits 100% and that's your fuel cut. Fuel cut is only a byproduct of the MAF, it's the injector duty cycle that is the issue. I kept pulling more and more fuel in open loop to stay lean, which worked just fine, but the PCM doesn't know that I am pulling fuel, all it knows is what it's outputting, and if it sees 100%, it's done.
4.8V may be where you are hitting 100%, but it all depends on your fuel trims, both short and long. There is no short term fuel trim in open loop, but the long term sticks. I am not sure if this is just the last LTFT it reads or it is an average of its LTFT table, but ideally you want your LTFT to be zero. So you'll just need to watch your LTFT's like a hawk and adjust the closed loop fuel to keep it at zero.
This is not something anyone can do on a dyno or even in a single night.
So the good news is I reset the PCM, upped my fuel trim to +20 anywhere I was modifying the O2 (my Analog A), and the car runs like a champ again. I kept an eye on my LTFT today and it's still at zero, which is great, but....
The bad news is that now i need to add 20% in closed loop and remove fuel in open loop (to get rid of our crappy rich tune). So I have to get that transition perfect, and since the RPM is an analog signal, it's never perfect. It's around 4170-4200 rpms with my current calibration, but there is no way to get it right on the dot. This is why I liked letting the PCM fix it's own transition with the STFT, and I was able to get perfectly smooth transitions into open loop, but only for a few drive cycles until that STFT became LTFT and now was being added to my open loop operation as well. That nasty transition comes back, and you'll start running 10:1 AFR's again (not lean like AEM claims in the video).
You could just pull more and more fuel to keep it ideal, but if you hit 100% duty cycle you are screwed, and the more you try to correct the transition with a an abrupt +20 to -20 fuel trim, the worse it is.
So in short, our cars suck and there really isn't a way to get a perfect tune. The car will always eventually mess it up. I thought about making a PCM reset switch and just reset it everytime I start the car, but that is just ridiculous.
I can watch my LTFT over a few days of driving, and keep it at 0, but I don't think it's possible to get rid of the transition permanently.
The SSAFC, the Unichip, etc all will have this problem. The AEM is probably the best tool for the job since you can at least monitor the injector signals and change your trip points, but none of them are going to be able to sample fast enought to catch that transition perfectly.
Why no one makes a piggyback with an integrated OBDII scanner and can tune on the fly by reading the STFT and LTFT I have no idea. Or how about just one that targets an AFR in real time? Would that really be so hard?
Here's a nice picture to demonstrate how hard it is to tune out the transition. This was adding 20% fuel up to 3600 rpm, 6% at 4140rpms and adding 0% at 4200 rpms.
This actually doesn't feel bad at all, just a little slow with the 10:1 AFR (just like stock), but notice my AFR's going northward at higher rpms due to my MAF adjustment (which I did to try to avoid fuel cut). I will need to fix that. As you can see it's only about 81% duty cycle. 1.25*81 = 101 and fuel cut. Doesn't really matter what your MAF reads.
Now that I have a more complete understanding of the behavior, I am more convinced that a stand alone is needed to really do this right, and I just don't think that's going to happen for me.
It's better now that it is stock. I don't have to worry about the 16:1 AFR's in closed loop and the hesitation is gone. The fuel cut riddle is solved and the AEM is perfect for getting rid of fuel cut on higher boost cars since you can actually see the injector duty cycle, but there just is no way to completely defeat this rigged PCM.
