COS#5 Car runs bad

[minytrker]

I have used COS#5 on probably close to a 100 fbodys with no problems until now. I had one that just wouldnt run very well with the pcm using the O2 sensors. I fought and and fought this car, very simple car to tune, just basic bolt ons and a small cam. As soon as I would let the stft turn on the car would run like crap and drive horrible. I tied new O2 sensors, even a another wide band and got the exact same results. I then just took the same exact tune and converted it back to the stock OS (12212156) and the car ran and drove perfect. It will not run with a custom OS in. Has anyone every came across this?

[WeathermanShawn]

Seems like I recall a thread that the COS uses a more aggressive STFT correction than when running a regular OS.

I have ran COS3 occasionally, but went back to the regular OS as I actually prefer LTFT corrections. Perhaps this thread will jog someone's memory concerning the 'type' of STFT correction(s).

[minytrker]

The STFT's were within 2% when I went to the stock OS so they should have been the same with the COS since I used the same tables but with the COS it did not run very good or have very throttle response. It just doesnt make any sense how with the same data in the tune it won't run with the COS but will with the stock OS. I tired everything over about a six month period before going back to the stock OS. I use COS all the time and am very familiar with them. I know I make mistakes so I even started over from scratch several times, and ended up with the same end result everytime. I haven't run LTFT in years so I don't know what they would have have been.

[nevinsb]

COS5 on my 12212156 S10 was working great after I fixed the ECT table that I had backwards. Mine is using STFT. I was using LTFT on my COS3, but the table came out smoother going off of just 1 bank instead of averaging the values. I now have all 6 injectors off of bank 1 since 1 O2 sensor is before the turbo. Long story short, I love COS5.