AFR error after OL tuning and knock sensor sensitivity
 

04 LS1, heads/cam, Ford red top injectors, NGK AFX, MAF with saxonpc honeycomb in an OTR intake.

I tuned the car in OL, got it pretty close to where I figured it'd be good to switch back to the blended, and noticed my AFR was a bit off (rich). I adjusted a bit based on MAF error, but it still seems a little bit off. Should I be looking at something else, or keep going on MAF error (from WB) until it's closer to correct? I wasn't sure if trying to adjust MAF/VE was recommended here, or trying to adjust the O2 switchpoints a bit. A few spots in the log have some higher outlier STFTs, but I was more looking at the heavier hit sections in the "driving" range. I have another 20 min log prior to the MAF adjustment if that would help as well.

Also, I've noticed this in the past, and today as well, that I get some part throttle "false" knock. Everything seems normal when this occurs, so I assume it's false knock, but I suppose I can't be "sure". You can see this at 3m51s, as well as in a few other places. When I adjusted the MAF, I also adjusted Knock Sensor Global Gain vs MAP vs RPM -15% in the 2000-3200RPM/35-70 kPa range. Is this thinking correct? How sensitive are knock sensors to factors not within the engine?

Thanks.

Do you have headers or have you moved the o2s from stock location? If so and you've done your OL tuning for VE and MAF and have your AFR nailed, then go back to CL and find the STFTs going too far from zero its most likely a combination of the closed loop integrator delay and o2 switch points. For me I got the most effect from the switchpoints. I increased the delay by about 25% and had to raise the switchpoints by about 150. You can log data in OL and and at 14.7 AFR, see what your O2mv, adjust switchpoints up/down to compensate. Delay tells the PCM that the o2 sensors are further downstream(headers) so it knows it will take more time for fueling changes to be seen by the o2s.

Ah, sorry, forgot to include that in the OP, but yes, car does have longtubes. I'll try this out later and see if I can get it sorted. Any ideas on the part-throttle knocking? Basically, I was just curious if there was any better way other than trial and error of lowering the Global Gain vs MAP vs RPM in the affected area to desensitize the sensors a bit. Lowering 15% didn't seem to have a noticeable effect, and I'd rather not just lower the values at random if there's a better way.