Hi Roland, I did two days of scanning on the car. (PS. had to use VCM Scanner, PCMTEC logger needs a bit of work - for another subject)
I removed the downstream (after cat) O2 sensor and installed a AFR gauge. Disabled the logic in the PCM. I then started looking at the injector data and plotted some stuff in excel. Roland was correct in saying the High Slope and Low Slope are linked and any change to the low slope will impact fueling when using high slope. Here was my starting LTFT.
At idle, cruise way too much fuel being pulled out.
I started by increasing the low slope curve - which makes the system leaner. And found immediate results. However, around the mid section 1800-2200 rpm the car ran very lean. This made good sense from looking at the excel charts. So I pulled back the high Slop to correct - with some good success. However I am not too concerned with high load just yet.
What I have found was under light acceleration the pinging actually got worse, than I had originally. AFR lean about 15.5. No matter what I did with the low and high slope this region was a bit of problem. Then I noticed the commanded injection pulsewidth was very close to the breakpoint. Hence this region needed to be richer. The only option I had was to move the offset as the higher RPM in closed loop cruise was perfect. In the end my settings went from:
The voltage offset was adjusted by 105% to get:
So after about 1/2 hour of driving i have ended up with a LTFT histogram like this:
STFT (ignoring the reds as it went to open loop at a few fangs. There is not enough counts but its a good start. It shows I have gone from super rich to mild lean at idles. This also shows my pressure regulator is doing its job with a walbro 460 up behind it (including an oversized return orifice at the pump - a must do). I am happy with the mid cruise areas for now.
I have plotted some graphs in excel to tune in these settings.
The car feels very snappy at the slightest touch of acceleration. I am working on the lean mixtures around 1800-2200 and boost pressure about zero (101kPA) just as the supercharger gets load on, seems to get lean for a couple of seconds until the ECU catches up. This i guess is where the accelerator pumps come in!
As for the accelerator pump shots:
Went from gain 0.1 to 0.05 as it slugged down at throttle impulse. Again - minor improvement to my lean issue around mid section vac-to-boost area.
The transient seems to work but for a very short time. It still takes 2-3 seconds for the ecu to play catch up and richen the injectors.
I got some data from a 5.4L Cobra S/C file from HP Tuners and had a look at their base fuel lambda table - interesting they run lower lambda much further down .
My lambda table is:
The Mustang GT500 (same engine but different compression ratios) is like this
At 30-40% throttle and 1500-2000 rpm they target 0.95 lambda. So I will be trialling this after work today.