Thank you to folks that have sent me AFMs (or sold them cheap). I have
discovered many interesting things. Here are a few (I'll have a full
report once I get one more piece of test equipment) observations:

1) I have not yet seen one an AFM with a bad IAT. There is about a 5
percent variation, but that is within tolerance.

2) The AFM is set up as a voltage divider. So, it's not absolute
resistance between pins 2&4, it's a ratio. Even AFMs with difference in
absolute resistance, the ratios are almost spot-on.

3) There are several...well, many, corrosion/oxidation points of failure
on an AFM: board contacts, wiper, wiper I, and wiper II, not to mention
the plug contacts. Hence, my quest for something better.

4) There seems to be two manufacturers of the interior "pot" board: Bourns
and Sferfine.

5) The "pot" board screws are a pain to take out, but I have developed
several quick methods to solve that problem.

Currently, all the AFMs that I have are 2.1WBXs, but data I have been sent
from air cooled vanagon AFMs seem to point to the conclusion that they
work the same. The AC AFMs have two extra pins, since an "AFM closed"
funtion is in the unit. This might be some sort of throttle closed or idle
function.

The CO screw simply controls the amount of air bypass around the AFM. In
other words, it will decrease or increase unmeasured (by the AFM) air.
I'll have to "do the math" to figure out that one, since the O2 sensor is
part of the equation.

My opinion is that the digifant is a well designed and simple FI system.
The problem is that is depends on several components (ala AFM) that like to
fail. The 2.1 engine has very few hoses, and junk.

That's it for now.

Kpm