Princeton fuel level converters adjustment?

[swatson999]

I have the Van's capacitive fuel level sensors in my tanks, and replaced the stock Dynon fuel level converters with the Princeton 2-set-point converters as I wasn't happy with the readouts from the stock ones (inaccurate and "jumpy" below 15 gallons or so). The new converters are much, much better. I did the full calibration setup procedure, first emptying the tank and setting that to 0V output, then filling the tank and setting it so that would output 5V, then re-emptying the tank and doing the Dynon SV 2 gallons-at-a-time gauge setup. Took for freaking ever to do both tanks, but it's done and they're quite stable.

My only problem is that both tanks show just about 1 gallon "too high" from around 17-18 gallons on down into the single digits (haven't had a tank below 9 gallons or thereabouts so can't say how they do when they're getting failrly empty).

Is there a way to "tweak" the setup on the Dynon so they're more accurate, without having to do the full calibration of the SV again? And a way to save the current settings in case I goober it up?

I jotted down the actual setpoints as read off of the SV and plotted them, and they look reasonable and smooth, but I'd like to just do this minor adjustment to make them all the much better...but I am loath to go through the 2 gallons at a time painful process yet again.

TIA...

Steve

 

[Dynon]

Yep, once you're calibrated you can muck with the voltages and values right on the product. Just go back to the fuel calibration page. DON'T start a new calibration. Press the EDIT button instead to play with the values in the table. Have fun! You might take a picture of the screen or otherwise record your starting values. Or even better, do a settings backup to a USB stick in case you fix it too far

 

[swatson999]

Any suggestions on how much to change them to start?  I suppose I could play with the chart in Excel to get some idea.

What sort of curve fit do you use internally? A 4th order polynomial looks like a close fit to my data, and I can see how I might adjust a few of the points in the middle of the curve by about 0.3V to get it to match, but if you're using a different polynomial or some other curve fitting, then it'll be off...

 

[dynonsupport]

The interpolation is linear between the points. If you need a more complex curve, you need to add more points.