What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

facts about capcitance fuel level system

bird

bird

Well Known Member
Hi, Im about to start my rv8 wings, I have ordered the capacitance type senders. I am interested to hear from others about the accuracy and compatability of vans senders. I will probably have a mgl or similar glass. I have read some previous threads about it so no need to repeat. I guess my most conern is about the variations in fuel. Does it really make a big difference in fuel types. Keep in mind that I will most likely be using the same fuel such as 100 ll avgas for lycoming. Is this difference in name brand fuel? Or is it like if I went to auto fuel? thanks

bird
 
The output of a capacitance fuel quantity measuring device depends on a property of the fuel called dielectric constant, which is a measure of how much electrical charge the fuel can hold per volt of an applied voltage, compared to air. The dielectric constant is dependent on the chemistry of the fuel. Once you calibrate for 100LL, you it's not likely you'll see much difference between different name brands of 100LL, as all brands must meet a spec which doesn't allow much variation in the chemistry.

I'm not sure how the dielectric constant of non-ethanol mogas compares with 100LL, but I think both are in the vicinity of 2 (that is, fuel holds about twice charge per volt than does air). The dielectric constant of ethanol is 24.3 (per wikipedia) so it greatly affects the dielectric constant of the fuel. The effect of ethanol-laced fuel on a sender calibrated for 100LL would be to show a significantly higher fuel level than is actually present.

I've toyed with the idea of adding a separate small plate to the bottom inboard bay of the fuel tank so that it is (almost) always immersed in fuel. It's function would essentially be to measure the dielectric constant of the fuel and enable compensating the output the measuring plates. The trick would be to come up with the compensation circuitry. My guess is that someone has already done this, but I haven't been able to find much info on it. I've taken a couple of stabs at designing one, but haven't yet come up with anything practical.
 
Do yourself a favor and install the float type. As we phase out 100ll you will be glad you did. Any mixture of mogas (unleaded fuel) and 100LL will result in VERY in accurate readings. Up to 7 gallons per tank. Unacceptable in any plane I have owned / flown with capacitive fuel gages.
 
Last edited:
longranger said:
I've toyed with the idea of adding a separate small plate to the bottom inboard bay of the fuel tank so that it is (almost) always immersed in fuel. It's function would essentially be to measure the dielectric constant of the fuel and enable compensating the output the measuring plates. The trick would be to come up with the compensation circuitry.
Click to expand...

I'm considering this, too, and since I have the wing crates sitting in the garage I need to come to a conclusion in the not-so-distant future. The easiest way of doing the compensation would be to combine the information of the two sensors in software, I think. I started a thread about this question recently, but haven't received any helpful replies yet: http://www.vansairforce.com/community/showthread.php?t=68259

If you come up with something, please let us know!
 
Very accurate

fuel readings with Vans capacitance set-up with the Advanced AF-2500 engine monitor and the Princeton converters. The gauge readings often mirror the fuel computer reading. If 100LL goes away, one could simply recalibrate the tanks with the new fuel.

Bill
RV-7 N151WP
Lees Summit, MO
 
hendrik said:
I'm considering this, too, and since I have the wing crates sitting in the garage I need to come to a conclusion in the not-so-distant future...
Click to expand...
You can add the compensation probes now, in anticipation of a solution for their use being found.

hendrik said:
... The easiest way of doing the compensation would be to combine the information of the two sensors in software, I think. ...
Click to expand...
I've also pretty much come to the conclusion that the software solution would be the most straightforward, especially if one of the EIS boxes would provide a couple of extra inputs and a way to implement some custom software...;). But in the meantime a hardware solution is still ricocheting around in my head...

hendrik said:
If you come up with something, please let us know!
Click to expand...

Will do!
 
longranger said:
I've toyed with the idea of adding a separate small plate to the bottom inboard bay of the fuel tank so that it is (almost) always immersed in fuel. It's function would essentially be to measure the dielectric constant of the fuel and enable compensating the output the measuring plates. The trick would be to come up with the compensation circuitry. My guess is that someone has already done this, but I haven't been able to find much info on it. I've taken a couple of stabs at designing one, but haven't yet come up with anything practical.
Click to expand...

That's a fantastic idea. Since the distance between the plates is known it would be straightforward to solve for the dielectric constant. Assuming that one is developing their own microcontroller circuit for fuel quantity readings. I'd do this but my tanks are closed up and done. :(
 
rocketbob said:
That's a fantastic idea. Since the distance between the plates is known it would be straightforward to solve for the dielectric constant. Assuming that one is developing their own microcontroller circuit for fuel quantity readings. I'd do this but my tanks are closed up and done. :(
Click to expand...

You wouldn't even have to find the dielectric constant numerically. If I've done the math right (NOT a slam dunk;)), the ratio of the outputs of the measuring probe and the compensation probe would be constant for a given fuel level regardless of the dielectric constant. EDIT: This is only true when the tank is full. I need to give this some more thought...
 
Last edited:
longranger said:
You wouldn't even have to find the dielectric constant numerically. If I've done the math right (NOT a slam dunk;)), the ratio of the outputs of the measuring probe and the compensation probe would be constant for a given fuel level regardless of the dielectric constant.
Click to expand...

Correct. You can call it what whatever, ratio, dielectric constant, etc. Its just a number to correct the fuel level reading. So when you calibrate the fuel readings for the first time, the baseline value would be established based on the capacitance value. Then during the normal course of operation when that number is different compared to the calibration number, is when you apply the correction based on the difference.
 
Calibration procedure please

I urgently need the setup/calibration procedure for a Princeton capacitance probe. It is the one with three LEDs and small push switch on the pug.
I will appreciate a document or a text reply.
Thanks.
 
For another data point... I have the MGL FF-1 gauge, and am using the standard Vans capacitive senders in each tank. The gauges read differently depending on the attitude of the aircraft, even when the tank is full. I elected to calibrate them for level flight, and only use a dipstick for checking fuel levels on the ground.

It's not that hard to change the calibration points, if you really needed to you could record the calibration points for 100LL and Mogas and manually switch them if you have to fill up with one or the other due to availability. Maybe MGL could even add this as a future upgrade to the gauges (Ranier, are you following this?).
 
Something some of us were discussing the other day...

Most automobiles now use capacitance fuel level sensors as well. How do they compensate for different fuels? Specifically I'm talking about the the GM "Flex Fuel" cars so common in the Midwest. You can fuel them with pure gasoline, gasoline with 10% ethanol or 85% ethanol. How do they work? Does ethanol not effect the dielectric properties of gasoline?
 
longranger said:
I've toyed with the idea of adding a separate small plate to the bottom inboard bay of the fuel tank so that it is (almost) always immersed in fuel. It's function would essentially be to measure the dielectric constant of the fuel and enable compensating the output the measuring plates. The trick would be to come up with the compensation circuitry. My guess is that someone has already done this, but I haven't been able to find much info on it. I've taken a couple of stabs at designing one, but haven't yet come up with anything practical.
Click to expand...

My intro to electrical engineering course is a looooong way behind me at this point... but I suspect that with some creative wiring, one could create a circuit that would self-compensate and always present a calibrated range of capacitance regardless of the fluid.

I'm tempted to try and hash out the particulars, but I'm not really equipped with the knowledge to do so in anything but a groping, trial-and-error manner.
 
E85 fuel gauge issues

Jamie said:
Something some of us were discussing the other day...

Most automobiles now use capacitance fuel level sensors as well. How do they compensate for different fuels? Specifically I'm talking about the the GM "Flex Fuel" cars so common in the Midwest. You can fuel them with pure gasoline, gasoline with 10% ethanol or 85% ethanol. How do they work? Does ethanol not effect the dielectric properties of gasoline?
Click to expand...

Jamie,
Not sure of other vehicles [I know that German cars have used capacitance senders for decades] but my 1999 Ford Ranger [flex fuel] has a float style sender. It DOES have a "flexible fuel" sensor located in the fuel line [under the truck, roughly below the driver's seat]. This sensor allows the computer to know which fuel is being used. It needs to know that, so that it can adjust fuel flow, injector dwell time, ignition timing and emissions calibrations as needed, depending on whether no lead, E10 or E85 is being used.
I will see if I can find out what BMW is using now.
Charlie Kuss
 
If you have QB wings with tanks fitted already, can you retro fit capacitance gauges ?

Our 7 has float gauges which only measure the last 12 gals of fuel due to their location in the tank :confused:
 
Jamie said:
Something some of us were discussing the other day...

Most automobiles now use capacitance fuel level sensors as well. How do they compensate for different fuels? Specifically I'm talking about the the GM "Flex Fuel" cars so common in the Midwest. You can fuel them with pure gasoline, gasoline with 10% ethanol or 85% ethanol. How do they work? Does ethanol not effect the dielectric properties of gasoline?
Click to expand...

Auto fuel gauges are notoriously inaccurate. But here is my assumption. Note that the gauge can be calibrated for the empty condition and it will be accurate regardless of fuel type since, when empty, there is no fuel, only air. For the Full condition, it is calibrated so it will read full for the fuel with whatever type of fuel will yield the least capacitance (lowest dielectric). If you use a fuel with a higher dielectric, the gauge will max out, reading full for a longer period of time but, as it approaches empty, will become more accurate.
 
Software solution

It's probably possible to have a system that automatically compensates for the varying dielectric constants of different fuels with hardware, but possibly the easiest solution is software. Any of the modern slew of digital engine monitors (AFS, Dynon, GRT, etc) could easily have multiple calibration tables for different fuels. If you start using 10% ethanol at home station, for example, you could select the table for that blend until you go cross-country, then select the table for 100LL. It's not going to be 100% accurate for mixtures, but it's good enough to keep you safe - and your fuel totalizer will fill in the gaps in accuracy.
 
A few years ago, when I followed the Subaru forum, there was a few posts about the inaccuracy of the capacitive fuel senders when using auto fuel. A few guys had flown back from Sun & Fun, and reported that the capacitive fuel gauges were way off every time they filled up. They reported something like 300 percent variance, and warned everyone against using the capacitive fuel senders if they were going to use both avfuel and auto fuel.

Since I had already installed capacitive fuel senders in my tanks, I decided to solve this problem and create my own fuel gauges. The software solution is actually quite simple. Once the tanks are calibrated, as long as the software knows how much fuel is initially in the tanks after a reset, it can simply calculate the ratio of the calibrated value to the current value. After that is known, the software simply adjusts the displayed value by that ratio. This should work, no matter what kind of fuel the tanks are filled with.

The problem with this solution is that it requires the pilot to either confirm full tanks, or enter the amount of fuel after filling up. This is not too big of a problem, since my engine monitor already requires this, but it would be better if the conversion was automatic. As someone mentioned earlier, this could be accomplished by installing a second smaller plate in the bottom of the tank to do the calculation. If the calibration sender was installed against the bottom of the tank there would be no problems until the tank was almost empty, depending upon the size of the calibration plate and the gap between the bottom of the tank and the plate.

I should note, that I did the software for the gauges, but I never tested it using real fuel, since I am still a long ways from flying. Also, another thought was to create just the senders, which could be used with any gauges or engine monitor, but other priorities got in the way.

Tracy.
 
The only problem I see is that the capacitive response full to empty is not linear.....this may create an error unless you account for it somehow.
 
No - As far as I know The Vans analog gauges will only work on the standard resistive probe.
 
BeLite has some fuel gauges that are based on pressure, that pressure measuring the depth of fuel in the tank. Don't know much about those gauges, but of the five homebuilts I've had, I don't think I trusted the fuel gauges in any of them.

Ed
 
You must log in or register to reply here.
Back
Top