Date:         Sat, 6 Mar 2010 08:14:33 -0500
Reply-To:     David Beierl <dbeierl@ATTGLOBAL.NET>
Sender:       Vanagon Mailing List <vanagon@gerry.vanagon.com>
From:         David Beierl <dbeierl@ATTGLOBAL.NET>
Subject:      Re: AIR FLOW METER
Comments: To: Arkady Mirvis <arkadymirvis@GMAIL.COM>
In-Reply-To:  <6CB16347182A4D9F99896D7950463080@Guenther>
Content-Type: text/plain; charset="us-ascii"; format=flowed

At 02:57 AM 3/6/2010, Arkady Mirvis wrote:
>Yours is one describing the AFM eventual failure in simple terms. Thanks. I
>never looked inside the part and would think that the manufacturer had an
>obligation to have the part ( wiper ) which wears down to be a replaceable
>part. I understand that no two variable resistors can be absolutely
>identical, but I was involved in sophisticated production and dealt with
>resistors with no more than 5% difference. The design was allowing perfect
>adjustments. Hope that some heads on the list will eventually find the best
>and economical solution.

Ark, the part that wears is actually the carbon material
itself.  It's a carbon film on a rectangular ceramic plate perhaps
4x6 cm, made by Bourne.  Each one is laser-trimmed, so I suspect the
accuracy is 1% or better.  It's possible to move the plate very
slightly (having due care for the screws which are very tight), if
necessary elongating the mounting holes a bit.  This will allow the
wiper to track along an unworn path perhaps half a millimeter away
from the original and seems to work acceptably.  The device functions
as a potentiometer, but the construction is more complex, with the
track being fed by multiple laser-trimmed resistors along its
length.  I strongly suspect that the plates are interchangeable
amongst different AFM units of the same type, but I don't know for
sure.  There are two internal adjustments in the AFM -- wiper
position and spring tension, corresponding to zero and span; but
Bosch clearly did not contemplate adjustments in the field.

The early AFMs maintained contact to the wiper by means of a
spring-loaded carbon button; later ones still have the button but
actual contact is made by a loop of flexible copper braid attached to
both parts -- apparently they had problems with the original setup.

Yours,
David