I am very confused (so what else is new???). I just read Reted's reply on a recent post that stated that the AFM determines fuel injected into the engine. the Boost sensor seems to only control ignition curves. So, I went out and set up to monitor the AFM output while driving the car. At idle, I got a reading of just under 3 volts. Cruising around, it fluctuated between 1 and 2 volts. So far, so good. I then floored the car at 3000 rpm in 3rd gear. Things got interesting in a hurry.

AFM voltage dropped to a little over 1 volt, then commenced to drop linearly to about .42 volts at a little less than 5000 rpm. After that, the voltage dropped to .39 volts by 6000 rpm, and changed no further to 8000 rpm. So, from less than 5000 rpm to 8000 rpm, there was .03 volts change! Throughout the entire run, the A/F diode remained steadily rich. This also happened in second, but was much harder to see (couldn't watch the road, the tach, and the meter all at the same time).

Conclusion: The AFM is definitely not much use above approx. 4500 rpm. So, what does the ECU use to compute the fuel curves above that rpm WOT? I would think the ECU has to use more than just RPM, otherwise how could you run over 10 psi boost with the stock ECU and a chip? Or, am I missing something obvious here?

Irv, Keith's dad

 

Check the other thread that's similar in content...

You have to remember, as the airflow meter gets close to fully open, it takes a LOT more air to move it further - this is more a logrithmic function of airflow versus flapper door position.&nbsp This could explain why you see very small voltage changes at the extreme end...




-Ted

 

Thanks for the reply, Ted. Watching the AFM signal, it reminded me more of a transistor input voltage/output voltage curve than a logrithmic curve. That is, within its operating range, it is linear. However, once the input voltage reaches a certain threshold, there is almost no change in the output voltage whatsoever. This is the saturation region, and is simply referred to as the "non-linear" portion of the curve. Equating this to the AFM, the response to airflow is extremely linear up until we get to a maximum air flow - very close to the .4 volt level - and then there is almost no change whatsoever from that point. I think that actual airflow to the engine changes quite a bit from 4500 to 8000 rpm, and the ECU could not possibly use the AFM as an input to determine fuel requirements as this airflow is way out of the useful range of the AFM (.03 volts is basically noise, different AFM's would probably have smaller/greater voltage fluctuations in this region. Even my AFM may read differently on a hot day/cold day (different expansion/contraction of the metal parts and spring, it is a mechanical device after all). You stated that the Boost sensor output has no effect on the fuel curve. OK, the AFM output is fixed, the Boost sensor does not affect the fuel curve, then what does? Or does the ECU simply inject a set amount of fuel when the AFM reaches maximum, regardless of boost?? If it does, then all our theories about engines blowing up because of lean fuel conditions when an FCD is used need re-evaluation! I don't know, does anyone know????

Irv, Keith's dad

 

this is why the ecu uses pre-programmed maps. the flapper door can only open so far before it is useless. at WOT i dont even think it looks at the sensor

 

well the other side to this is that you would have the same boost at 4500 then at 7000. so if it used the boost sensor to add fuel there would be no change. just like you said there was no change in the AFM.... which would make either way worthless...

 

Hmmm...I guess I worded it a bit vague.

Yes, the output is linear.

When I mentioned "logrithmic", I was talking about actually airflow.&nbsp The first 1/4" if movement takes very little airflow, as the flapper door is basically perpendicular to the airflow.&nbsp At total open, it takes a LOT more airflow to move the door the last 1/4" due to the flapper door being almost parallel to the airflow direction.&nbsp This doesn't even go to the spring rate change at extreme cases...



-Ted

 

if you are trying to tweak for some extra fuel, i believe someone on the FC3S board had come up with a way to modify the coolant temp signal to make the engine run rich thinking its colder.

 

I have had problem with my boost sensor system, so I been talking to a lot of people about it.
I have an N/A, but the same applies to a TII.
According to Six Rotors, who I beleive is actually a mazda tech, says that input from the boost sensor can increase fuel delivery by up to 15% under full throttle conditions.
He has seen posts saying it only affects ignition curves- but disagrees.
Hapy Holidays