[Diy_efi] Speed-density vs. MAF/MAP...

[Klaus Allmendinger]

Hi,

I don't know if I'm correct but from this discussion I came to the
following conclusions:

1. MAF sensors would be the best way to measure mass air flow. The
disadvantage is that MAF sensors are flow-direction insensitive (the output
signal is essentially proportional to the full-wave rectified flow).
Therefore a low-pass filter (plenum) has to be inserted between the AC
input signal (oscillating air mass in intake stream) and the sensor to
filter out the AC component of the signal and allow to measure the DC
component (the actual flow). Like all filters it creates a delay which is a
problem for fast flow changes.

2. MAP measurement is flow indifferent. The actual mass-flow is calculated
from experimentally determined VE/RPM tables and MAP, RPM, BP and IAT. To
make even sense from the MAP data it must be sampled at a specific point in
the engine cycle. This point may not be fixed ove the RPM range but may
change because of the phase difference between the resonant pressure and
flow of the intake tract and engine cycle. 

3. TPS/RPM inferres MAP from experimentally determined tables, then
proceeds as above.

The three methods increase in workload for the tuner because each method
requires more experimentation than the one before.

I think it would be best to have a MAF sensor that is flow direction
sensitive. I don't know if such an animal exists. I could imagine one built
of two small pitot tubes (used to measure airspeed in airplanes) in the
intake stream, one pointing upstream, one pointing downstream. The
differential signal between the two would create a flow sensitive air-speed
signal. Taking the actual pressure and temperature at the pitot location
into account the mass-airflow can then simply be determined by integrating
the signal over one intake period without the filter delay. Does anybody
know if this has been done?


Regards,
Klaus




_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi