MAF bypass loop

[George M. Dailey]

At 08:45 PM 4/20/96 -0600, you wrote:
>Hello gang!
>
>I was wondering if anyone has seen/tried this type of MAF setup...
>
>
>-----------------------------------------\
>from filter	     || MAF		  ---\
>--->	------------------------\	      ---\
>--\	\________________________\		  -----------------
>   \		bypass line			to engine ------>
>    \_______________________________________________________________
>
>(sorry about the poor quality ASCII art, obviously I'm no expert!)
>
>The MAF is represented by the two vertical bars in the top tube.  My
>thought is this:  the MAF is usually the most restrictive part of the
>induction system, and who says it needs to be the only path to the engine.
>I figure if you have a bypass tube sized so most of the flow doesn't see
>the MAF at all, you will drastically reduce delta P trough the intake
>tract.  However, this is not without problems.  I am not sure of the
>Reynold's number range an MAF is designed to operate (anybody?).  In
>addition, I could see this type of setup reducing transient response of
>the overall system. Whadda ya think? does this have potential?
>
>Getting back to the Reynolds # range;  no doubt an MAF is fairly accurate
>within a certain range and under specific conditions, how about below the
>lower limit and above the upper limit?  does the sensor response become
>non-linear? is there even a response? will the sensor ouput curve flatten
>off? What are the conditions an MAF is expected to operate? laminar
>smooth, no turbulence? is there any rule of thumb re how far upstream you
>must keep a straight shot to the sensor?  Any help you pros could give me
>would be greatly appreciated...
>
>Thanks gang.
>
>Todd-
===========================================================================
Todd, the Japanese company Hitachi has been there and done that. Here is
what I know about it. Some, not all, of GMs port fuel injected V6s used an
Hitachi made MAF sensor like you described (sort of). The difference is,
they placed the MAF element in the smaller BYPASS tube. My source gave two
reasons for this.

1) As you mentioned, very low delta P accross the sensor.

2) Reduced trash fouling of the heated MAF element. Simply put, less total
air passing by the sensor element. This reportedly eliminated the burn cycle
that is so very needed on the Bosh hot wire MAF sensor. I discovered just
how sensitive the Bosh unit was to trash when my TPI started surging with
the air cleaner off.

These same engines were also equiped with an Delco unit (most of the time I
think). The Delco MAFS used a hot mylar film instead of the platnum wire and
had no burn off cycle!
I don't know why GM switched, unless the Delco was giving problems. I'd like
to hear from an GM person on that.

"... and that's all I have to say about that"

GMD