A little history...
I used to run my twin turbo GTO on my self tuned, 2.5 BAR OS, speed density tune. This tune was great. I didn't touch it in nearly 2-1/2 years of daily driving. Dynoed at 554 RWHP/572 FTLB. Since I was running it as SD only, I 'gutted' the MAF. I took the screen out and drilled a hole in the center airfoil for my meth injection nozzle, I like the stealth look . I left the circuitry intact and logged from it still, but the data was not used for any fueling calculations.
Recently, I've changed my engine combination (cam, heads, compression and injectors). I've disconnected the meth injection and reinstalled an untouched, unmolested MAF sensor. I've re-enabled the MAF for fueling calculations to make tuning easier until I get the VE tables tuned.
Now I'm seeing an unexpected lean condition under boost when at part throttle versus full throttle. When at part throttle, with PE enabled, as the boost rises, the MAF doesn't appear to follow along properly.
With the old setup, the MAF I descreened and hacked up did follow the boost, in that, as the boost rose, so did the Hz output of the MAF.
Since my fueling is now based on the MAF, I suspect this is the cause of a lean condition under boost. The lean AFR curve appears to follow the boost curve identically. During this time, the MAF stays pretty flat, relative to the boost curve.
I have a screen shot of two logs. On the left is the new combo, MAF based fueling, with the lean condition. On the right is a log I took two years ago with the SD setup using the hacked up MAF. Both log snapshots are during a part throttle pull.
You can see in the OLD log that the MAF output follows the boost curve. But the NEW one does not.
If I got WOT, this doesn't happen. At WOT, the boost and MAF curves follow one another nicely.
Any ideas what's going on? I would have thought an untouched stock MAF would be more accurate than one which has been gutted.
Guess it's time to get those VE tables dialed in and get away from the MAF.