Calibrating fuel qty gauges or using Optical senors

[Thruster87]

The floats-on-a-swing-arm senders are

 

a legacy product from cars and other

 

vehicles that go back a very long way.

 

 

 

Making these things really accurate in more\

 

than one place is a bit fussy.

 

 

 

Years ago, we crafted an electronic signal

 

conditioning board for the Bonanzas and Barons

 

that allowed dead-on calibration of empty

 

and full. All other readings across the scale

 

simply fell where where the physics of the

 

sender dictates. The physics of these devices

 

are affected mildly by linearity of the wire

 

wound sensor resistor (usually within 5% of

 

true) but a whole lot by trigonometry of

 

the swing arm and tank geometry.

 

 

 

But as you've already recognized, the one

 

level you really want to be accurate is

 

the empty point. For this you can do some

 

things with series calibration resistors

 

and/or bending the float arm on the sender.

 

This CAN be a tedious, trial-by-error activity.

 

 

 

If it were my airplane, I'd probably craft

 

a microprocessor based signal conditioner

 

that would allow me to take readings at 5%

 

increments from empty to full and generate

 

a lookup table that converts as-installed

 

sender (transducer) readings into real

 

numbers. The BEST way to watch full levels

 

is with installation of a "dip stick" style

 

sensor at the low fuel warning level (generally

 

1/4 to 1/3 tank). Consider devices like this:

 

 

 

 

One of these stuck through the tank wall at

 

the warning level will light a lamp on the

 

panel at the desired fuel quantity with

 

no risk for drift of calibration. See:

 

 

 

http://www.gemssensors.com/content.aspx?id 82

 

 

 

This is the no-brainer, dead-nuts accurate

 

low liquid lever sensing method I know of.

 

Capacity fuel gages with processor augmentation

 

are also easy to calibrate . . . but I think I could get by with no active fuel gaging

 

other than a set of optical level detectors

 

cited above. [ATTACH]8555.vB[/ATTACH] by nuckolls.bob(at)aeroelect