Message-Id: <3.0.3.32.19981018190142.007096b0@mail.digital.net>
Date: Sun, 18 Oct 1998 19:01:42 -0400
To: lancair.list@olsusa.com
From: John Cooper <heyduke@digital.net>
Subject: Fuel floats
Mime-Version: 1.0

          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>>
          <<  Lancair Builders' Mail List  >>
          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>> 
>>
Marv and all--

Thanks for the good suggestions.

Right now, I plan to have only a LOW LEVEL sensor and warning light. Since
I haven't come up with an automatic system that I am comfortable with, I
will be going straight manual control of the xfer pumps. I will do some
kind of a test to verify that with both transfer pumps running, I will be
unable to structurally damage the header tank. I am thinking of the
possibility of a pressure switch to sense the increase in pressure when the
xfer pumps are trying to pump into a full tank, but don't know yet. One
thing at a time...

So I have two options with my low level switch: I can make it so the float
hangs down when the tank is dry, in which case I would be presenting an
OPEN set of contacts to whatever warning system I chose. Reverse logic IOW.
This has the advantage of self-testing for two failure modes: an open wire
or a "waterlogged" float. The warning light would be ON with the sightglass
showing gas in the tank for either of the above. OTOH, an additional
failure mode is introduced by the logic circuitry required to turn a light
on with the sensor open. Using this system, a lamp test circuit would be
required to verify the system was working. But maybe not...if I were to run
the tank "dry" (my sensor will be about 2" above the bottom) each time,
then I would have a verification that the system was working at least once
per flight.

Turning the float around so that it would float up (not down) when the tank
was not empty would allow me to connect the light directly to the reed
switch. Simpler. But an open wire would not be detected in advance of
running the tank dry, although a "waterlogged" float would be.

Right now I am kinda partial to the first (reverse-logic) method...but I'm
still pondering. Using some 7400 series logic would allow de-bounce and
set/reset/monostable options that might be beneficial. I dunno...I'm not
really in the mood right now to spend a month designing some kluge <sigh>.
Maybe I'll just take the easy route.

Sometimes I sits and thinks, and sometimes I just sits...