Mass airflow sensors.

[Andrew Brownsword]

Thanks for the additional input... I want to get as much information on the
tradeoffs as possible before making the leap.  I can always switch if there
are problems, but that is going to be costly!

The response time isn't an issue for driver sensitivity (although I know
from experience that I'm sensitive to a much shorter interval than 0.1
sec!), it is more important to avoid leaning out.  The existing ECU,
however, has acceleration enrichment that is triggered by an increase in
RPM.  I'd have to double-check but I think it also has a TPS enchriment
table.  It couldn't cover a 200 ms response time, but it could cover a
couple of milliseconds with no problem.

Another VAF->MAF converter doesn't seem to think that running the MAF under
pressure (only 6-7 psi of boost) will be an issue... and his empircal
experience suggests that the MAF is, if anything, more responsive than the
original VAF.  There is also the issue of crud & oil deposits from the turbo
accumulating on the MAF sensor, but he doesn't think cleaning it
periodically is a major issue.  I'm really interested in hearing other
opinions on these issues.

Advantages of the MAF approach are:

- Considerable amount of output data is available
- Cheaper & wider part selection; easier to replace if it fails
- Generates output across the full 0 - 5 volt range, taking maximum
advantage of the input 10-bit DAC

Advantages of the Link AFM:

- Programmable
- Not vulnerable to vacuum leaks
- Certain to work in a boosted application (without cleaning)


Now I'm off to search the archives for info on this topic.

Thanks,
  Andrew


> Just some numbers to mull over while your thinking this thru.
> Joe Average driver thinks .1sec is an instanteous response.  The best of F1,
> can detect .01 (I think it was Senna, that set this "record").   This is
> just in reference to engine response.
> The 200 msec., might be to fully stabilize, on the hot style sensors but a
> good part of that is smoothing to a level you'll never notice/see.  The
> transistional delay to airflow is several msec., as I've heard reported.
>   There is no AE (or little, to none), on the gm MAFs, yet on the MAPs there
> is MAP AE, Coolant temp AE and TPS AE.  (AE=Accleration Enrichment).
>   If your system is looking for a V difference, how does the Ford MAFs line
> up?.
> Can you get a Ford sensor and put it in parrarell with your door sytle
> sensor and compare them, and then use your existing code (with changes) to
> "read" the other sensor?
>
>> Hmmm... this is a strong argument for converting to speed/density.  The
>> pressure change should be instantaneous -- or at least as fast as the VAF
>> signal change.
>>
>> > How about adding a MAP sensor, and looking for a sudden change in that?.
>> >
>> >> I'm also working on converting my AFM to MAF but I've ran into a
>> >> couple problems that I don't see mentioned here.  The response
>> >> time of the AFM to changes in air flow is almost instantaneous.
>> >>
>> >> Any thoughts?  BTW the car is an 88 BMW 535i if that helps.
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org