What's new
Van's Air Force

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

Fuel levels not registering

Flying Canuck

Well Known Member
Patron
This is probably just going to be a forehead slap opportunity but maybe someone knows this one. I started doing my fuel level calibration last night - Dynon Skyview HDX, standard resistive type fuel senders connected to default EMS pins (20 and 21). I started with the left tank and was prompted to do repeating 5L pours. I did 3 pours and the sensor reading wasn't changing, Skyview was complaining about that. I decided to abandon my calibration and check the sensor levels on the debug screen, think that maybe I'd reversed left and right. What I found was the left value was ~3.7V and the right value was ~2.9V. The left one was the same value showing on the calibration screen. I decided to try a pour in the right tank to see if the sensor value changed. It didn't. I don't get any movement on either sensor.

I expect that this is an installation error. I recall checking the senders with a multimeter before installing them and they worked fine. Unless I completely missed the point, there is only one wire connecting the screw on the sender to the EMS wire. I'll climb under my panel tonight to confirm the connections but hoped that maybe someone has an idea of what I did wrong?

Thanks for your help.
 
Make sure the sender plate is well grounded by connecting a ground wire to the plate. I am assuming the sender is the float type with potentiometer and round mounting plate.
 
Didn't get ANY change on the measurement, or just didn't get a LARGE change on the measurement? Skyview with the standard float sensors is known for having very small voltage changes between the fuel pours during calibration - typically complaining between each pour that it didn't change, when in fact it changed SOME but not enough to keep Skyview happy. Rather than using the full 5 volt range on the float potentiometer, it's common that empty-to-full range with the Vans tanks and the standard float pickups you'll see a difference in 1-volt range. Skyview will still do an adequate job of interpreting the fuel levels if you complete the calibration, but you're looking for changes between pours of tenths and hundredths of a volt from the sensor.
 
I had one tank's float stick when I first tried the Skyview calibration pours. I ended up giving the round plate a few raps with a rubber hammer and it unstuck and has read accurately for 600+ hours. Guess it hadn't moved for a year or two while building and just needed a little coaxing.
 
Didn't get ANY change on the measurement, or just didn't get a LARGE change on the measurement? Skyview with the standard float sensors is known for having very small voltage changes between the fuel pours during calibration - typically complaining between each pour that it didn't change, when in fact it changed SOME but not enough to keep Skyview happy. Rather than using the full 5 volt range on the float potentiometer, it's common that empty-to-full range with the Vans tanks and the standard float pickups you'll see a difference in 1-volt range. Skyview will still do an adequate job of interpreting the fuel levels if you complete the calibration, but you're looking for changes between pours of tenths and hundredths of a volt from the sensor.

I had a 0.02 V move after the 2nd pour, none before or after though. The right side didn't move at all but I only put in maybe 8L.
 
I just went back and looked at mine, I have about .35 volts delta between full and empty.
 
Tonight I'll connect a temporary ground to one of the plates to see if that helps. I'll see too if I can get my inspection camera up inside the emptier tank to check the position of the float. Maybe a mirror, a flashlight and an eyeball is safer.

I'll bring a new battery for my multimeter and check the resistance too. I'm not going to let this little problem beat me! Need to get this done so I can do my first start on Saturday and final inspection in 10 days or so.
 
Connecting a ground wire *may* help but probably not. These senders are grounded through their mounting plates and the tanks through the plate nuts. I went through this with my hangar mate when we attempted to calibrate his senders and the best we could achieve was not satisfactory at all. Bottom line is that with the SW senders in these tanks, it is very difficult to get anything approximating a full range, and then adding to that the inability of the Dynon EMS to sense the very small differences in voltage, and you have a situation where you just do the best you can, and then really work on fine tuning your fuel flow sensor K factor. To say I don't trust my fuel level readings is a huge understatement. Sometimes I think they are on target and other times, they are wildly off. The only value they might provide is an indication of a fuel tank leak....enough said, do the best you can and move on.
 
Well that wasn't a very productive night. I checked resistance between my firewall and the sender plate and got tiny numbers, so I can't see this being a ground issue. I pushed through the calibration for both tanks. The right tank stayed at the peak level until my 45L pour (out of 67L) when it jumped a lot. It stayed at the second level right through to full. The left tank never budged from 2.67V.

The one thing I managed to get was a calibrated dip stick, so at least I can tell how much fuel I have on the ground. I think my next step will have to be emptying the tanks and seeing if I can tell if the floats are free to move either with my inspection camera or by poking some hinge pin or something up through the drain. That's probably going to have to wait until after I do my first engine start on Saturday.
 
Connecting a ground wire *may* help but probably not. These senders are grounded through their mounting plates and the tanks through the plate nuts. I went through this with my hangar mate when we attempted to calibrate his senders and the best we could achieve was not satisfactory at all. Bottom line is that with the SW senders in these tanks, it is very difficult to get anything approximating a full range, and then adding to that the inability of the Dynon EMS to sense the very small differences in voltage, and you have a situation where you just do the best you can, and then really work on fine tuning your fuel flow sensor K factor. To say I don't trust my fuel level readings is a huge understatement. Sometimes I think they are on target and other times, they are wildly off. The only value they might provide is an indication of a fuel tank leak....enough said, do the best you can and move on.

I have to disagree.

The tanks cannot read beyond a certain level as they angled up. On the RV-8, there is no reading beyond 16 gallons or so. On the RV-10 this is 24 gallons. However, from 16 (or 24) gallons down to 2 gallons the fuel reading is within a gallon of actual on the SkyView (level flight). I crosscheck these readings with the fuel flow totalizer on all flights to make sure I?m not leaking fuel.

When you build the tanks it helps to adjust the fuel sender arm range such that it goes all the way to the bottom of the tank. This means you will have most accuracy at low fuel levels (desired).

Do the calibration per the instructions and they work well.

Carl
 
Well that wasn't a very productive night. I checked resistance between my firewall and the sender plate and got tiny numbers, so I can't see this being a ground issue. I pushed through the calibration for both tanks. The right tank stayed at the peak level until my 45L pour (out of 67L) when it jumped a lot. It stayed at the second level right through to full. The left tank never budged from 2.67V.

The one thing I managed to get was a calibrated dip stick, so at least I can tell how much fuel I have on the ground. I think my next step will have to be emptying the tanks and seeing if I can tell if the floats are free to move either with my inspection camera or by poking some hinge pin or something up through the drain. That's probably going to have to wait until after I do my first engine start on Saturday.

When you say tinny numbers can you verify you are getting < than .2 OHMs between your battery negative and the float mounting plate, and make sure the meter is not reading thousands, there should be the OHM symbol and not a (K) or an (M). :rolleyes:
 
On my -6, when I bought it my float senders would only read from about 2/3 of the tank down to empty. It came with two automotive-style needle gauges and when I replaced them with a single MGL fuel gauge found it a little disconcerting that the level stayed at 72L per side for more than an hour's flying before jumping to somewhere just under 50L.

The Red Cube I installed along with the MGL gauge, however, has been bang on. Never more than 1L off when I re-fill.

I had to remove the senders and re-seal the covers due to leaky cork gaskets, when I put them back in I find that the readings are more capricious than they were... I wonder if the grounding resistance has changed enough that i'll need to re-calibrate. Still, the totalizer via the red cube remains spot on.

If I have to remove the covers again, i'll remove the senders permanently, and mount Red Cubes in each wing root. That would be more accurate than the floats, and would read accurately from full to empty.
 
Well that wasn't a very productive night. I checked resistance between my firewall and the sender plate and got tiny numbers, so I can't see this being a ground issue. I pushed through the calibration for both tanks. The right tank stayed at the peak level until my 45L pour (out of 67L) when it jumped a lot. It stayed at the second level right through to full. The left tank never budged from 2.67V.

The one thing I managed to get was a calibrated dip stick, so at least I can tell how much fuel I have on the ground. I think my next step will have to be emptying the tanks and seeing if I can tell if the floats are free to move either with my inspection camera or by poking some hinge pin or something up through the drain. That's probably going to have to wait until after I do my first engine start on Saturday.

This sounds very much like a stuck float, or with some resistance to movement, possibly due to debris from building becoming lodged in the rheostat arm. You might try putting about 4 gallons of fuel in the tank and rocking the wing back and forth to slosh it at the wing root, to force the float up and down quickly a few times and see if that frees it up. Other than that, you'll probably end up pulling the end plates open to get at the sender.
 
On my -6, when I bought it my float senders would only read from about 2/3 of the tank down to empty. It came with two automotive-style needle gauges and when I replaced them with a single MGL fuel gauge found it a little disconcerting that the level stayed at 72L per side for more than an hour's flying before jumping to somewhere just under 50L.

The Red Cube I installed along with the MGL gauge, however, has been bang on. Never more than 1L off when I re-fill.i

I had to remove the senders and re-seal the covers due to leaky cork gaskets, when I put them back in I find that the readings are more capricious than they were... I wonder if the grounding resistance has changed enough that i'll need to re-calibrate. Still, the totalizer via the red cube remains spot on.

If I have to remove the covers again, i'll remove the senders permanently, and mount Red Cubes in each wing root. That would be more accurate than the floats, and would read accurately from full to empty.

My red cube is mounted just below my fuel distribution manifold so it sees exactly what goes into the engine. Almost every fill up, I compare what was pumped to what is predicted by the fuel flow and since I?ve calibrated everything carefully, the result is within 3/10 gal consistently. I do look at my fuel level displays but only as a secondary source of info.
 
SNIP

If I have to remove the covers again, i'll remove the senders permanently, and mount Red Cubes in each wing root. That would be more accurate than the floats, and would read accurately from full to empty.

Good plan assuming you don?t have something like a tank sump valve leaking or when the line guy (if you don?t fuel the tanks yourself) shorts you.

The float senders are dirt simple and reliable.

Carl
 
When you say tinny numbers can you verify you are getting < than .2 OHMs between your battery negative and the float mounting plate, and make sure the meter is not reading thousands, there should be the OHM symbol and not a (K) or an (M). :rolleyes:

If I recall correctly, I was seeing something close to .2 ohms - I was definitely on the ohms setting. The left tank might not have been that good, my multimeter was acting up and I was getting some fluctuations.
 
On my -6, when I bought it my float senders would only read from about 2/3 of the tank down to empty.

Pretty standard for all the RV's that have wing tanks. Because the senders are located in the most inboard bay of the tank, and with the dihedral angle there is fuel in the outboard portion of the tank that is higher than the highest point of the float, it will always require some fuel to be burned off before the float starts to lower.
Never been an issue because it is at the bottom of the gauge range that accuracy is important.

A poor ground connection of the sender flange to airframe ground is a very common cause of a fuel gauge not working properly (either during initial calibration of after some time in service). It is easy to check with an ohm meter (no point in adding ground wires, etc., without even checking to see if it is the problem).

A common cause of calibration error is not accounting for friction in the sender. This varies from one sender to the next when they are new so it sometimes doesn't cause a problem, but often times it does.

Think about the fact that in flight an airplane has vibration induced into it by the engine/prop (regardless of how smooth your engine runs). This vibration helps overcome any friction in the float senders. This vibration isn't present while doing a fuel gauge calibration. (Side note - When the military first started flying jet aircraft they discovered that many of the analog instruments were no longer accurate because the airplane was free of vibration. The had to artificially vibrate the panels)

When we do fuel calibrations in our shop, we do the fuel add, and then tap on the fuel tank skin with an open hand until the voltage wont change anymore. Add more fuel, and repeat. The value will often change quit a bit from what it is if you just pour in the fuel.
 
I had one tank's float stick when I first tried the Skyview calibration pours. I ended up giving the round plate a few raps with a rubber hammer and it unstuck and has read accurately for 600+ hours. Guess it hadn't moved for a year or two while building and just needed a little coaxing.

What he said :). I try to check the easy, fast and cheap things first.
 
I tested both my new senders with an ohm meter before I installed them in my modified ER quick build tanks. One was not at all happy and I was getting a small range of float arm swing with no reading at all. Close examination of the resistor part I found a stripe of some kind of black substance/coating on the part of winding where there was no electrical contact unless I pressed on the contact spring arm. I was able to clean it off with some acetone and and a Q-Tip and it tested fine after that. I have no idea what it was though and they were factory new.
 
Decided to play hooky from work this afternoon and go and try some of your suggestions. Glad I did because I arrived to find a very large blue ring on the floor and an overwhelming familiar smell. Probably should have tightened my fuel lines at the bottom of the valve better. I lost about 10L of AVGAS.

Anyways, armed with a new battery for my multimeter I checked and found the resistance between the sender plate and the firewall to be right around 0.2 ohms. I powered up my display to look at the calibration routine and then decided to check voltage on my senders, between the plate and the post. I found that matched what Skyview was seeing exactly. My conclusion from that is that adding a ground won't change anything. On a whim I removed some fuel from the right wing and checked the voltage again and it moved. I repeated the action and got the same reaction. Looks like the right side is fine. I completed the calibration on the right tank, it worked quite well. As expected, the readings topped out around 57L. I tried the left wing again and got zero movement through about 30L (right wing moved quite a bit in that range). I tried the tapping and rocking tricks to no avail. Either this sender is bad or it is jammed. I think my only recourse now is pulling it out, I'll try that in the morning.

Thanks for all of the great discussion and suggestions, it's certainly helped.
 
Well I tried a few more things tonight and I think I'm only left with one possibility - a bad sender. I drained the tank and used one of my seat bent hinge pins to feel around through the drain hole. Managed to find the float arm and confirm that it's moving freely. Couldn't manage to check the voltage with it moved, I was short an arm. But I know that the float isn't stuck. I put some fuel back in (10L - enough to register a notable change on the right side) and the reading from the sender never moved out of the 2.7588-2.7600 range.

I tried to remove the sender, but it's stuck there pretty solidly. With the tank/wing mounted there aren't many places where I can pry on that plate. Any ideas on how I can get that out without removing the tank?
 
What about making a bracket to hold a piece of thin safety wire taught, then working it behind the plate? I Prosealed mine on with a 1/16" gap and that was how I would plan to remove if and whenever I needed to.
 
Fuel Sensors Suck

It think we accept a level of performance that is simply out of touch with modern technology when it comes to fuel sensors. I hate mine! I mean, really! Van's and the rest of the fuel sensor world can't come up with a fuel sensor system that will accurately indicate my fuel level from full to empty? It makes me angry.

I found this thread while researching a problem with my right fuel tank float sensor. It has decided, after 105 hours of operation, to set the fuel level at 1.8 gallons and never move again. I'll figure that out, but as I troubleshoot, I sure would like to improve the situation such that I can know the fuel level accurately at all times. Is that really too much to ask?
 
Ground your fuel-sender plate!

Mine was ok, when weather was wet or high humidity. When cold and dry, it showed different levels down to zero.

Then I grounded the fuel-sender-plate and voila: all is fine!
trmyKziMB9gHSpcRLD6y0v5UkN8ZiWYTCXY5LQVosUY1I4g5-aad9oToGU41M5yTpKid1EW0vGiO17LTyjigdY2DPC6P32aBB76wxrhVHkWsr-L4ijaabH0iTW520AZ1x-gALVeaqSU48IUXxfh0q4YdEmWyRBdChZLvbkmtGasuOyrzBYbpd0YzDizUdUdjTn11wglwQyvl6eqQSRNbNwDRZciDWZc6tqRGw50JJWUMyqT3B1po30ux2P3vyMsA0dB36TWDWBRmqZADWaipzqbN2QSQ2QJLQy6BIaBRY1mEl512K3afozEuk2zwippok2rQVRTtWsSKMTHKIv_v4jM66hP1mLgHLuJrTSCFqvvZNbrL_4wvd0IyHI4qDLdB4Lr74CXav5Wn03WuqHgbQLS9wl1-0ZpLNJ3q0EysW-SfCX1WFQ0tVdN1dOM8CLC0pA4L-1jvb-aDpRK6xpJgtMgrfU4dfo_ShFb02GGJKD3y75S9lX9QcXzbtVKH6T_GdksRy2oUAzG5OcUlhDI7qmHKNPucoWCUU7esVKYZsvNa2HHCXV942Jy202t_Y7M4aCmlVmrCWUl8XzIDhO4azA50d_ISU85SlHgz1v0=w1340-h1005-no

It is old technologie, the floating fuel sender. But if you calibrate them with modern EFIS/EM's, then the become really accurate. Anyway, you should not count to the last liter!
 
Last edited:
Yes, but

Yes, I may be able to get my sensor working but, even at it's best, it won't indicate the level accurately over about 16 gallons in a 21 gallon tank. That, I think, is unacceptable. Perhaps we've identified a market for a new product - an accurate fuel level sensor throughout the fuel range!
 
As long as the wings (and the fuel tanks) have dihedral, I don't think there's a float based mechanical sensor that's ever going to get you to both the top and the bottom of the tank. The first ~4 gallons don't reach to outboard end and the last few gallons sit away from the inboard end. When on the inboard end, the float is at it's highest possible position well before the tank is full.

I tried adding a ground to my non-working sender and it didn't help. I'm going to have to remove it (the cheese slicer trick with safety wire sounds promising). I think I'm going to leave it alone for the time being and go ahead with the final inspection with this outstanding. I do have a few other sensor related issues though (oil pressure, CHT #4 both work fine unless the engine is running) and my alternator isn't charging - so I may be having to move the inspection out anyways.
 
Fuel Senders

I understand the mechanical limitations of a float based sender. I'm just surprised that nobody has come up with a full range solution to the problem!
 
Fuel gauging.

Mooney’s have a 2nd additional float type fuel sender on the outboard tank rib that must be in series with the inboard sender. Numerous times after fuel tank work I’ve filled them up a few gallons at a time and indicated quantity would match the truck within a 1-2 gal empty to full. Wondering how much sense it makes to add that cost, weight, complexity, addition failure modes to a small plane that likley has a totalizator on board. But hey, this is the best part of E/AB. All we need is a drill and a little common sense to do almost whatever we want. Gotta love this country and our aviation rules. Personally I’m ok with the gauges not showing the upper 1/3 of the quantity if they work on the bottom 2/3. Small plane gauging systems have never been great and it’s wise to treat what the gauge shows as only a opinion on what you already know.

Don Broussard
RV9 Rebuild in Progress
57 Pacer
 
Last edited:
I understand the mechanical limitations of a float based sender. I'm just surprised that nobody has come up with a full range solution to the problem!

It is probably because no-one wants to spend 10s of thousands of dollars designing something unnecessary and never recoup their investment. This might be an opportunity for you to come with something. :)
 
Easy peasey

Actually, I think with modern electronics it is easy to make the fuel senders read from full to empty. Just add one at tip of the tank and wire them in series. The tip rib sender would need to be tweaked so it is not referenced to structure ground, but this is easy to do during installation. It might be possible to have the sensors in parrell too, so that both are reference to chassis ground; i am not there yet in my build so dont know the details of the G3x fuel level calibration ranges.

I was talked out of adding a sender in the tank tip rib; now i wish I would have done it during tank build.oh well.
 
Capacitive senders are much more accurate than float senders - but both suffer from the same problem with respect to the dihedral on the wing. Either way, to be really accurate you need an inboard sensor and an outboard sensor to accurately measure fuel level through the whole range. With the current level of tech, I believe the idea of standard resistive float sensors (one inboard, one outboard) in series might be a good place to start, as it will calibrate similar to the single inboard sensor most systems are designed to accept. In any case, it should be easy to achieve (even with dual in-series sensors) more than .80 volts range of the standard capacitive sensors on the market - which would be a great improvement.
 
Last edited:
I understand the mechanical limitations of a float based sender. I'm just surprised that nobody has come up with a full range solution to the problem!

I'm still very happy with the Van's capacitive senders I installed, zero problems and very accurate empty to full. Not sure why they discontinued them. And it doesn't get much simplier, 2 pieces of alum and some wire.
 
I'm still very happy with the Van's capacitive senders I installed, zero problems and very accurate empty to full. Not sure why they discontinued them. And it doesn't get much simplier, 2 pieces of alum and some wire.

Oh, agreed - they work great - as long as you just run one type of fuel to keep them calibrated. My issue revolved around running a variable mixture of 91E10 and 100LL, they have different dielectric values and I couldn't use the capacitive senders as a result.
 
Mystery solved

Well I found the definitive cause of my left fuel sender not working tonight. I managed to get the sender removed (cut away the bulge of proseal around the outside and got it moving with a flat screwdriver and a hammer). When I pulled the sender out of the tank there was something obvious missing - the float and the arm. Somehow it got out of the plastic clip that holds it into the sender unit. Managed to to fish out the float and arm, thankfully the arm is magnetic. I have a new sender on its way that I don't think I'll need, but the proseal that's coming with it will definitely get used to reattach the sender. The good news is that I should have it all working by the end of next week when I expect the final paperwork to go flying.
 
Well I found the definitive cause of my left fuel sender not working tonight. I managed to get the sender removed (cut away the bulge of proseal around the outside and got it moving with a flat screwdriver and a hammer). When I pulled the sender out of the tank there was something obvious missing - the float and the arm. Somehow it got out of the plastic clip that holds it into the sender unit. Managed to to fish out the float and arm, thankfully the arm is magnetic. I have a new sender on its way that I don't think I'll need, but the proseal that's coming with it will definitely get used to reattach the sender. The good news is that I should have it all working by the end of next week when I expect the final paperwork to go flying.

That's great news. When I was fitting and bending the float arms on my tanks before installation, I was not happy with the float arm retention clip, so I put a few small dabs of proseal to ensure they would not pop out.
 
SNIP... When I pulled the sender out of the tank there was something obvious missing - the float and the arm. Somehow it got out of the plastic clip that holds it into the sender unit. ....SNIP

I found the same problem on a builder?s project. It was the classic mistake of not making the 90 degree bend on the float arm that goes though the sender clip arm hole.

Van?s instruction are still not perfectly clear on this.

Carl
 
Back
Top