sergei wrote:My only question: how fast ECU samples that sensor?
It will not be good of fast sensor if ECU samples slow and filters heavily the input.
Not sure how fast the stock sensor would sample and how much filtering there would be, but given the nature of the stock sensor I'd guess that the sample rate would be low(<10hz), but not much filtering would be applied.
By contrast my Motec M-800 samples at 150hz and the suggested filter level is for intake air temp is 8 samples, i.e. the value used is the average of ~ 0.05secs or around 2-3 combustion cycles @ 6000rpm, which is probably more than accurate enough bearing in mind that the response rate of the sensor is probably <20°C/sec i.e. about 1°C in the same period.
FWIW (ignoring humidty) 10°C change in temp give approx 3% change in density. The following table give the fuel compensation values for Apexi and Motec ECUs and it can be seen that they don't exactly follow density changes, with resulting mixtures being richer at higher temps and leaner at lower temps, with the Apexi being notably richer. Probably in an attempt to control detonation, though to be fair it uses less ignition retard that the standard Motec Settings (5°@80°C and 3° at 70°C vs 6° & 5° repectively).
- Code: Select all
80°C 50°C 30°C 10°C -10°C -30°C
__________________________________________________________________________
Density Change -20.5% -10.2% -3.4% +3.4% +10.2% +17.1%
Apexi Correction -8.2% -5.1% -1.2% +2.7% +7.8% +12.9%
Motec Correction -12.0% -6.0% -2.0% +2.0% +6.0% +10.0%