Guy Prevost

Guy Prevost

Well Known Member
Hi,

Does anybody know what the typical capacitance range is for Van's capacitive senders? I'm getting ready to build my own converter to measure the capacitance at the plates and output 0-5V for a Dynon or whatever.

For any microcontroller nerds out there, my general plan is to build an R/C circuit with the R being constant and the capacitance varying with fuel level. I'll use a microcontroller to measure the time constant of the circuit and output a variable (0-5V) voltage using PWM. There will be a calibration routine to go through, just like you have to do with Princeton's units. Feel free to scoop me on this idea--I won't get my feelings hurt.

No promises that I'll make this work, but I don't think it will be too difficult. (Of course, the devil is always in the details). If it goes well, I'll post the schematics and code here for all to use. Parts for both channels should run <$65.00 depending on where you get them and only require basic soldering skills to assemble.
Anyway, if someone knows the approximate capacitance range of the tanks I can start working on some of the design. I'll work on this in ernest in the 2 months I'll be waiting for the fuselage after my wings are done.

TIA,
Guy
 
Capacitive converter

Hello,

There is actually a post from the tech support at the Dynon forum that states that the converters that Van's sells is actually a capacitive to frequency (PWM?) converter and that they will not work with the Dynon units. Van's sells them for $45.

Vans site says they are capacitive to voltage but the Dynon support says they are capacitive to frequency.

Just food for thought. I am also very interested in a low cost solution to using capacitive plate type senders with the Dynon units.

Were you going to use the PWM signal from the microcontroller to drive an output stage that would be linear 0-5v? Or just output the PWM direct?
 
Seems like you can get a VOM that will also measure capacitance? If so, then you would need to get the empty and full of fuel capacitances (you probably aready know this) and then build for that...

I would call Dynon as I think they may make the transducers and will be able (and hopefully willing) to share their knowledge with you.
 
R and C

I would think Vans converter is a PWM (Pulse Width Modulation) since that would directly drive a mechanical meter correctly.

If it was a straight Frequency output, with the same duty cycle on/off, it wouldn't change a regular meter.

If it's PWM, then something as simple as a resistor and fairly large capacitor would do the trick and drive the 0-5 volt input of the Dynon. This would work since the variable you are measuring (fuel level) does not change rapidly.

This R/C combination would just act electrically to average out the pulses in the same way as the mechanical meter...

gil in Tucson
 
Guy,

My plan was to do pretty much the same thing. This should allow me to handle the different types of fuel that all of the Subaru guys complained about when using capacitive senders. Then if the user says the tanks are full and the capacitance is way off, either the tank is not full or the fuel density is different. You should be able to do a percentage conversion to give the right output.

One thought I had was to maybe use a digital resistor as the output. I haven't looked into this, yet.

Also, here's the link to Jim Weir's articles on capacitive senders.

http://www.rst-engr.com/kitplanes/

Good luck, and keep us informed.

Tracy.
 
Gil Alexander has it right. One thing to remember is that PWM is fixed frequency, variable duration--it's not a frequency converter.

I was planning on using a fairly high frequency pwm output filtered through an R/C filter to provide a variable voltage. The dynon unit will have no idea it originated as a PWM signal.

The capacitance values of the senders are definitely going to be non-linear. There's only one plate with any change when the tank is almost full, and only the other plate has any change when the tank is almost empty. I think the easiest way to handle this is to do a look-up table in the software. The values in the table will come from adding fuel to the tanks in 2 gallon increments and recording the number clock counts required for the R/C circuit (previously brought high) to decay to a digital low.
 
Last edited:
Guy,

I agree that you need a lookup table to correlate fuel level to capacitance. However, the non-linearity is due to the shape of the tank and the shape and position of the senders. However, if the capacitance of the full fuel condition does not correlate to the calibrated value, then you should be able to apply a variance factor to the lookup value to get the actual level for different fuel types.

If after filling the tanks the capacitance is 75 percent of the calibrated value, then either the tanks are not full, or you have filled the tanks with mogas (assuming you calibrated with avgas). Therefore, you should be able to solve for the current problem with capacitive senders where they do not read correctly when you have to use avgas for a normally mogas powered engine.

Of course, that is what I intend to try to do.

Tracy.
 
Capacitive plate value in pF

The Cap plates in my RV6 wings measured 180pF with no fuel.
This was measured at the BNC fitting at the wing root.
Borrowed some test gear from work.

Rolf
 
Capacitive Senders

I have the Princeton Senders in my -8A connected to 1-1/4" UMA gauges. I couldn't get them to calibrate so contacted Todd at GRT/ Princeton. He said the senders were designed for 200-470 pF.

Turns out for whatever reason my tanks are below the lower value. Testing the senders with a 330 pf capacitor showed that senders & gauges were OK.

I still have to "tune" the senders by changing a resistor in the circuit.

FWIW

John Moody
 
moodyjr said:
I have the Princeton Senders in my -8A connected to 1-1/4" UMA gauges. I couldn't get them to calibrate so contacted Todd at GRT/ Princeton. He said the senders were designed for 200-470 pF.

Turns out for whatever reason my tanks are below the lower value. Testing the senders with a 330 pf capacitor showed that senders & gauges were OK.

I still have to "tune" the senders by changing a resistor in the circuit.

FWIW

John Moody
Click to expand...
John,
I just finished my cap plates in my tank (RV9a) and they measured ~250pF.
The outboard one by itself measured ~170pF and I expected roughly double that amount with both hooked up but only got the ~80pF bump.
I would expect the value to be roughly double (~500pF) when full of fuel since the dielectric constant of avgas is ~2 and that of air is ~ 1.

-mike
 
You must log in or register to reply here.