<VV> Universal Sender Unit

N. Joseph Potts pottsf@msn.com
Tue, 24 Feb 2004 20:07:17 -0500


As Charlie mentions, the adapted sender will not provide accurate signals to
the gauge at positions between full and empty. I got interested in how MUCH
the error would be, and did some calculations (with the assistance of Excel
and MAPLE) and discovered that, for the values in question (90-ohm variable
with 47-ohm fixed resistors substituting for 30-ohm variable resistor), the
error can be quite substantial, coming very close to 50% at its maximum.
     In order to do the calculations, I made one assumption that may be
incorrect, but if it is incorrect, the corrected assumption would yield the
same magnitudes of errors, just at different places. I assumed that the
resistance seen by the gauge at Full is zero, and at Empty, 30 ohms (with
the original sender). It might be the other way around - for the reasons
stated, I don't care much.
     The maximum error for the resistance-error function (a quadratic
formula that I wouldn't have been able to differentiate by hand) occurs when
the tank is 61.8% (about 5/8) full. At that point, the substitute resistors
would cause an indication of 33.8% (about 1/3) full.
     Obviously, on a trip through an area where (open) gas stations might be
sparse, this could lead to a serious misestimation of remaining range.
     Anyone interested in the formula can get in touch with me directly;
I'll send you the spreadsheet that contains and evaluates it.

Joe Potts
Miami, Florida USA
1966 Corsa coupe 140hp 4-speed with A/C

-----Original Message-----
From: virtualvairs-admin@corvair.org
[mailto:virtualvairs-admin@corvair.org]On Behalf Of Charles Morgan
Sent: Tuesday, February 24, 2004 3:03 PM
To: virtualvairs@corvair.org
Subject: Re: <VV> Universal Sender Unit


The method Mark recommends is probably the simplest method but it won't be
accurate except at the two extremes (empty and full).  In between the gauge
will read too high.

Charlie Morgan
Beaver State Corvairs