
Originally Posted by
cobaltssoverbooster
That's what I'm saying.
If you add 5% and the trims report back -4.5% then you IFR value is high by about 8-10%. On the other side if it reports -5.5% then your IFR value is low by 8-10%.
If you want ot be more accurate you can use 10% instead of 5%. 10% error would be -9 and -11 when using +10% commanded.
I suggest using a histogram filter RPM>1,200 so you exclude all of the idle IFR multiplers. This allows you to leave the short pulse modifiers active and have a better idle while performing testing. It should go without saying that the IFR table is linear. If you make changes based on this test method, make sure you interpolate the data to form a straight line. The IFR multipliers take care of the non-linear flow rate variabes.