just curious - air flow sensor

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

william_b_noble - 17 Jul 2005 05:26 GMT

well, I just fixed a rough running problem on my daughter's 944S - wear on
the carbon track of the air flow sensor - I gotta believe this is a pretty
common problem, and replacement sensors are a bunch of hundreds of $$ - so,
it seems to me that a better solution is needed - a shaft encoder, a ROM,
and a D/A should be able to produce any darned voltage curve the computer
needs

how many folks would be interested in such a solution if I have the time to
work out the details - cost won't be under $100, but probably won't be over
$200 -if you do the installation of the new parts into your own air flow
sensor assembly. maybe another $100 to have it done for you.

anyone got a better idea? some kind of non-contact solution seems to me to
be the best idea - if not a shaft encoder, then maybe some kind of analog
approach involving a vane blocking light to a photoresistor, or an inductive
or capacitive solution is also a possibility.

Signature

Bill

to Email me, repair this address and use it:
william_ b_ noble at msn dot com

also check out http://www.wbnoble.com

Reply to this Message

Jack D. Russell, Sr. - 17 Jul 2005 13:17 GMT

===============================================
* Reply by Jack D. Russell, Sr. <jackru$$ell2@notmail.com>
* Newsgroup: alt.autos.porsche.944
* Reply to: All; "william_b_noble" <nobody@nowhere.com>
* Date:Sun, 17 Jul 2005 07:13:08 -0500
* Subj: just curious - air flow sensor
=====================================================

wb>well, I just fixed a rough running problem on my daughter's 944S
wb>- wear on the carbon track of the air flow sensor - I gotta
wb>believe this is a pretty common problem, and replacement
wb>sensors are a bunch of hundreds of $$ - so, it seems to me that
wb>a better solution is needed - a shaft encoder, a ROM, and a D/A
wb>should be able to produce any darned voltage curve the computer
wb>needs

wb>how many folks would be interested in such a solution if I have
wb>the time to work out the details - cost won't be under $100,
wb>but probably won't be over $200 -if you do the installation of
wb>the new parts into your own air flow sensor assembly. maybe
wb>another $100 to have it done for you.

wb>anyone got a better idea? some kind of non-contact solution
wb>seems to me to be the best idea - if not a shaft encoder, then
wb>maybe some kind of analog approach involving a vane blocking
wb>light to a photoresistor, or an inductive or capacitive
wb>solution is also a possibility.

Don't have a better idea, Bill, and undoubtedly you've already seen F.R. Wilk's excellent article concerning the infamous 944 AFM. I'll include the link http://frwilk.com/944dme/afm.htm for those that haven't. Maybe it'll give someone else some ideas. Good luck.

Signature

Jack

Reply to this Message

william_b_noble - 17 Jul 2005 15:22 GMT

no, I had not seen that article - it is a good and comprehensive review -
but the design is wear prone - well, I guess after 15 to 25 years one can't
complain, but it is a critical wear part that is expensive - having the
resistance curves helps.

> ===============================================

> Don't have a better idea, Bill, and undoubtedly you've already seen F.R.
> Wilk's excellent article concerning the infamous 944 AFM. I'll include the
> link http://frwilk.com/944dme/afm.htm for those that haven't. Maybe it'll
> give someone else some ideas. Good luck.

Reply to this Message

Walter Spector - 17 Jul 2005 20:00 GMT

> ... it is a critical wear part that is expensive - having the
> resistance curves helps.

There have been kits available to replace the vane-type AFM with
the more modern MAF design for years. But the kits that I've seen
are expensive, and don't seem to be particularly well sorted.

Walt - 86 944 NA

Reply to this Message

william_b_noble - 20 Jul 2005 06:12 GMT

And, having contacted a vendor or two, the ones I could find are for the
turbo only.
after looking at the curves, it looks to me like the later one is just a
linear resistor - maybe the easiest is to find a mil spec position pot and
use it, but some kind of zero wear thing would be more aesthetic

>> ... it is a critical wear part that is expensive - having the
>> resistance curves helps.
[quoted text clipped - 4 lines]
>
> Walt - 86 944 NA

Reply to this Message