Calibrating Fuel Level with Capacitive Senders

[NoahF]

I have capacitive fuel sensors in the tanks and Princeton converters which convert to (nominally) 0-5V. The engine sensor harness schematic requires a resistor in line to limit voltage to 0-4V, as that is the required input for the AFS.

However, since the setpoints on the Princeton units can be calibrated to output 0-4V (I think) instead of 0-5V, I would think that the resistor could be avoided. Has anybody done this? Any issues? Wouldn't this be the preferred way to go??

[Trevor Conroy]

If the Princeton adapters can indeed be calibrated to output 0-4V DC, than the resistor is not needed. However, we haven't seen one that can be calibrated to output 0-4V.

[Trevor Conroy]

Noah, you will really want to make sure it outputs less than 5V, as you will blow a diode on the engine board and the unit will have to come back to us if not. I would just install that 1K resistor and be done.

[rshannon]

I have a related question RE capacitive sensor tank calibration and voltage. I have Westach sensors. My initial calibration exercise (every 2 gal., 22 per tank) was done at in flight and ground attitudes, on main battery power. The process took 4 hours, over which time the battery voltage dropped, perhaps 1/2-1 volt or more. I later found the readings had significant error, when the gauges read 6.5 gal., each, but dipstick readings were much less, and draining them entirely yielded only 3.5 gal. each, including unusable. The only thing I could think of which might account for the error was the declining voltage during the 4 hr. calibration process.

1) My question is what effect does a declining voltage during the calibration process have on accuracy, if any?

2) If it does produce error, would the error be seen as a positive error at the low end of capacity, as I experienced?

3) I assume the software compensates for voltage fluctuation in operation (as opposed to during calibration) or the readings would become inaccurate in conditions where they should remain accurate, i.e., alternator failure. Is that correct, or will the readings become inaccurate with changing voltage, even during otherwise normal operations?

Ron
Murphy Rebel / Jabiru 3300
Dual AF-3400EE/EF+AOA
http://n254mr.com

[Trevor Conroy]

Hi Ron,

It shouldn't be too much of a problem as the capacitance senders output 5 volts to the EFIS. As long as the 5V supply remains constant, there shouldn't be an issue. If you are using the EFIS 5V supply, I could see why fluctuations could occur. The screen should be calibrated using main buss power, rather than the internal back-up battery.

[rshannon]

By main battery power, I meant main bus power, not the internal backup battery. The main battery voltage is what dropped during the 4 hr. process.

Several days ago, I redid the calibration process a second time, maintaining a constant main bus voltage. So far, things seem to be about right.

However, if I understand your reply correctly, the key is the 5V output from the EFIS to the gauge. I assume that voltage is independently regulated by the EFIS, and thus not subject to voltage fluctuations of a couple volts on the main bus. If so, my theory of the cause of the error would be... toast. :-)

[Trevor Conroy]

Ron, it depends where you are getting the 5V from for your Westak senders. If that voltage varies at all due to a lower main buss voltage, it will affect your reading. It sounds like you got the issue resolved with a constant buss voltage?

[rshannon]

Aha! Upon further review <g> the Westach senders are powered by fused +12V from the main bus, so the falling main bus voltage over the 4 hr. calibration exercise probably did cause the error in my first calibration attempt. The second time, I kept a precision charger on the main bus, and yes, it does seem to be OK so far, though I haven't done a definitive drain, or other test.

It would be good to mention the need for a constant voltage source for accurate tank calibration in the manual.