Air Flow Measurement

[tom cloud]

Thank you .... I feel like I've heard the clear ring of reason in the
comments below.  I admit I'm a spectator, not a player, at this time,
but there's much that doesn't add up in this diy_efi discussion.
Among them:
        .. too much reliance on digital.  Ask a guy who works in a power
           plant about SSR's (that's solid state relays) and he'll tell
           you they're wonderful ... some time, but that a *REAL*
           relay is hard to beat at others.  Ask anyone who works with
           transmitters, etc, and tuned circuits about DVM's (that's
           digital voltmeters) and he'll ask for a *REAL* meter
           (i.e. analog VOM).  I love the *control* digital gives me,
           but it has it's limits, and I believe the "wise" designer
           uses each technology to its best advantage.  [If you don't
           understand analog -- LEARN IT!]  Flame if you want -- I'm
           not (necessarily) advocating doing away with micro-pressers,
           but, my analog/LSI digital (LSI means large scale integration,
           implying no CPU, etc) EFI system from Holley seems to only
           need slight improvements to be an excellent system (Now,
           "slight" might be a real witch to implement!).

        .. too much voodo magic (means I don't understand it!).  I'm
           a techno-freek.  I wanna see a data sheet, read some specs!
           Where's the hard facts?  Why do I have to pay $25 to read
           some incredibly dull, meandering, mostly useless SAE paper
           to find out what I want to know?  I mean, it's not like this
           is a new technology or the volume is so low there's little
           activity.  This is as bad as trying to get the Coke formula!
           There oughta be a 'cookbook' on efi, complete with suggested
           hardware (i.e. injectors, manifolds, etc.), controller and
           software implementations.  (Dream on!!)


You know -- if someone would market an "open architecture" (a la S-100
bus, IBM-PC) efi, I believe he could make a whale of a lot of money.
Something that majored in simplicity and gave the customer source code
(probably in BASIC) so he could build on what he got.  If analog
and digital techniques were mixed, the control algorithm could be
simple and processor timing (if there was one) would not be critical.

>>> What's wrong with speed-density?  And what exactly do you do to it? 
>>>I thought it was just found by Air=K*MAP*RPM (where K is a magic 
>>>constant), but I'm naive.  Am I going to get into trouble using 
>>>this?
>>
>>There is nothing wrong with speed-density, especially if you are not 
>>worried about emissions certification. 
>
>I get upset whenever somebody maligns the humble MAP sensor and speed 
>density calculation! :)  Let it be known that the ULEV Honda Accord 
>uses a MAP sensor and speed-density airflow calculation.
>
>After trying to get a MAF to work with a reasonable error over the 
>speed-load map, I'd gladly toss it out and use a MAP sensor except 
>that you can't MAP and EGR at the same time.
>
>I think if most people who swear by MAF had a reference airflow 
>measurement to compare it to, they'd be horrified!

***************************

Tom Cloud <cloud at peaches.ph.utexas.edu>