# Level switches in the fuel system for alarms - Yes, No, Why?

Discussion in 'Firewall Forward / Props / Fuel system' started by wsimpso1, Dec 14, 2018.

### Help Support HomeBuiltAirplanes Forum by donating:

1. Dec 14, 2018

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,627
2,866
Location:
Saline Michigan
The saga continues. First my setup. My fuel system has a couple wing tanks, a header tank, pumps to get fuel from the wings to the header, draw from the header to run the engine.

I already have capacitance fuel gages for the wings, but none selected yet for the header tank.

The idea is that things break, we do not want the engine to stop anytime soon just because something quit working... That includes fuel gauges.

Now we get to the point where we use your brains as well as mine. Yeah, your thinking is a welcome addition to my thinking...

Since running any pump dry is bad for the pump, I am thinking a low level warning on each wing would be a good idea. Yes, No, What made you say that?

Next how would you run fuel level sensing in the header tank? As long as the wing tank selected has fuel in it and the pump selected is running, the header tank should be full and circulating some fuel back to the selected wing tank. Do we really want a gauge showing quantity when it should always be full? Yes, no, Why did you select your answer?

I am thinking the header tank could be just fine with two simple level switches because if I can not add any more fuel to the header and the level has just started down, I have about 50 minutes at cruise speed or 100 minutes at max endurance speed. When the other triggers, I better set it down on the best chunk of dirt I can see;
One switch to tell you when the transfer pump is not keeping the header full;
The other to tell you that the engine is about to stop from fuel starvation. If you have a compelling reason to know how much is in the header tank, I want to hear that too.

Now, assuming we want alarm switches, which should we have?

There are float switches in the catalogs that run around $30. I do not know about you guys, but a limit switch with a float on a pivot stuck inside a fuel tank that might have some water in it and someplace difficult to access for replacement just sounds like an invite for trouble at several levels. Industrial solid state switches with zero moving parts (and they are small too) exist that sense if we have liquid at them or not. They even come in materials that will stand prolonged fuel immersion.$129 for the fuel safe ones.

So which would you use and why?

There is the question of the week. Help is appreciated.

Billski

2. Dec 14, 2018

### proppastie

#### Well-Known Member

Joined:
Feb 19, 2012
Messages:
3,511
1,006
Location:
NJ
seems overly complicated...Piper and Mooney have 1 engine driven fuel pump and 1 electric backup, either pump from the LH or RH tank to the engine. If you run out of fuel in one tank, you will know it before you damage the pump. (engine will stop). I believe RV is similar. Lots of solid state level sensors for the RV, but certified use float sensors. My preference is for the proven tried and true, (floats), I believe there are some bad reports about some of the solid state sensors for the RV. Those "pump back" fuel systems, have caused lots of crashes in early Beach V tail aircraft, they only pumped back to one side and if you screwed up you ended up pumping fuel overboard. My injected Mooney has a relief valve in the pump, and no pump back. I guess the float carbs in the Piper must work the same way.

I also have a fuel flow, totalizator with a programed alarm at 5 gal. . some I have heard have the rotating sensors in more than one tank, but again with a pump back fuel system you have to then have sensors to keep track of the fuel pumped back.

Mine is very accurate to about 1 gal per 50 but that has to be programed by measuring the fuel used against the reading and adjusting the K factor to get it better.

3. Dec 14, 2018

### Hot Wings

#### Well-Known MemberHBA Supporter

Joined:
Nov 14, 2009
Messages:
5,995
1,999
Location:
Rocky Mountains
To me this sounds like adding a bell or whistle that really isn't needed. I tend to do this because I kind of like gadgets.

If the fuel gauge is working and the flight plan is sound then we should know about how much fuel is where. Even if the gauge malfunctions and says there is more fuel than there actually is then, again, if the flight plan is sound we should be noticing the discrepancy?

4. Dec 14, 2018

### Toobuilder

#### Well-Known Member

Joined:
Jan 20, 2010
Messages:
4,287
3,034
Location:
Mojave, Ca
Been there, thought about that... Have a box full of float switches.

With the advent of very accurate FF transducers I have contemplated dumping all direct reading QTY indicators in favor of a simple three color LED indicator. Two float switches can drive a tri color LED and depending on position, can give you (for example): Green - within 70% total capacity; Yellow 69% - 15%; Red - less than 15%. That was my plan for the mains, and the LED was going to be mounted at each position on the fuel selector for quick reference (dont switch to a RED if things go sideways).

For your header, I'll fall back to a military jet I'm familiar with. In this jet, the inboard wing tanks drain first, then the system sequences to the outboards. There is NO direct reading of wing tank fuel level. The pilot has a fuel flow totalizer (set before takeoff), and the sump tank quantity. When the inboards run dry, the sump QTY drops about 10 gallons and this opens float valves for the outboards. The salient point is the pilot knows what his main tank is doing based on the state of the sump tank level.

So, LED's and float switches can give you a heads up concerning the main tank feed state. In your system the normal header state is "full" (whatever that level means). In this case you'd be green and all is good. If a main tank quits feeding, the header will come off full and a change in state of a float switch will activate a yellow (and any other aural/visual alarm you feep appropriate). With luck, this yellow period gives you plenty of time to figure out the issue or start heading for an airport.

There are many types of float switches available. I purchased several types of magnetic proximity switches to play with.

5. Dec 14, 2018

#### Well-Known Member

Joined:
Dec 9, 2004
Messages:
388
151
Location:
Hamilton New Zealand
A fuel flow totalizer requires good discipline to use reliably. For military jets that makes sense (lots of tanks that are often hard to gauge & military trainers built to resemble the systems of fighters), for GA level that's just another layer of added risk. Unless at start up one can automate the importation of the fuel quantity into the totalizator.

Low level switches in the wing tanks does allow one to plan having to switch tanks, rather than the "Oh s**t I have to change tanks right now" at a moment of high work load. Plus one can safely put an obnoxious horn on the header tank float switch with out the risk of hearing it often.

6. Dec 14, 2018

### pwood66889

#### Well-Known Member

Joined:
Feb 10, 2007
Messages:
1,323
120
Location:
Sopchoppy, Florida, USA
Please consider the Ercoupe system: wing tanks pumped to a header, then gravity fed to gascolator. One "gage" which is usually covered by upolstry on one of two wing tanks, then a float on a wire for the header. Experience flying the rig will tip you off when fuel is coming from the header only.

7. Dec 14, 2018

### pictsidhe

#### Well-Known Member

Joined:
Jul 15, 2014
Messages:
5,890
1,500
Location:
North Carolina
It seems a bit complex to me. Like Too, I'm thinking about crude level sensing with a couple of level sensors per wing tank.

I have 3 of these inductive level sensors running an overly-complex water storage and filtration system for a few years now. Dunno what they will make of gas, but cheap enough to try one...

8. Dec 14, 2018

### lr27

Joined:
Nov 3, 2007
Messages:
3,216
463
$139 sounds like a lot. My guess is that it's fairly simple and that the actual components don't cost much. Also that appropriate, simple circuits are found on the net. You could possibly have a fixed float with a strain gauge*. Or a strain gauge on part of the tank mount, assuming the mount is a material that's got a low elastic modulus. Or even the tank itself. Or use capacitance. Also, I'm pretty sure there are pressure sensors sensitive enough to use. You might need one in the top of the tank and one at the bottom to account for any pressurization of the tank. You can probably find really inexpensive sensors in Digikey or Newark or someplace like that. Strain gauges may be a little more specialized, but if I recall correctly, they're not that big of a deal. We used to put them in fiberglass oceanographic buoys to see if the tether was still attached. If it was, wave action would give a varying voltage from the strain gauge. Much better than the two exterior electrodes we used previously which might get fouled after a few months. Probably all of these methods aren't worth it if your time is worth much, unless you like fiddling around with them. I'm guessing you've installed the tank where you can't see it? That's the simple way. Especially if your tank is made of something translucent. (Is there fuel resistant epoxy?) I'm also guessing you wouldn't like the drag from the way it was done on the Taylorcraft I rode in once. A wire with a cork** on one end that stuck up through the top of the tank (maybe through the fuel cap?). How high the wire was told you how much fuel was left. * Strain gauges aren't as sophisticated as they sound. A little piece of plastic with metal on it etched the way PC boards are. Preferably in a network of 4 hooked up like a Wheatstone bridge. A very simple amplifier gives you a usable voltage. (A Whitestone bridge is too heavy for aircraft.) **Ok, I don't know if it was really a cork. 9. Dec 14, 2018 ### Chilton ### Chilton #### Well-Known Member Joined: Jun 28, 2014 Messages: 63 Likes Received: 15 Location: Arusha Considering the header tank holds more fuel than your legally required reserve I would certainly put a gauge on that, other than that you would have to assume that your reserve is the full header tank quantity which reduces your true range. For the main tanks I would put a low level warning which is silenced when the tank is not selected or which can otherwise be silenced due to the amount of fuel in the header, but it would warn you if you ran a tank low while selected. I would go with solid state switches for the warning just because I have had to change too many float sensors and if they are hard to reach even more so! 10. Dec 14, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan 11. Dec 14, 2018 ### Dan Thomas ### Dan Thomas #### Well-Known Member Joined: Sep 18, 2008 Messages: 4,714 Likes Received: 1,946 With fuel sloshing around in flight, the alarm would be intermittently sounding or flashing or whatever. You'd want a damper circuit: a transistor to fire the alarm, and a couple of resistors and capacitors to delay the triggering of the transistor's base. Or get fancy and use a 555 timer chip. Set it up to require 15 seconds of constant switch activation to sound the alarm. wsimpso1 likes this. 12. Dec 14, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan Yeah, I KNOW that the low level alarm is redundant to the gauge. Redundancy has its advantages, particularly in terms of lowering exposures. Billski 13. Dec 14, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan Looks interesting on the Dynon. I would have to add FF sensors for the flow out to the fuel rail and another for the return line to do that. Another possibility is in using the PWM signal to each injector and the rated fuel flow of the injectors to get an estimate of the fuel flow, much like the way many cars estimate their economy. I wonder if that is possibe through the Dynon box... None of that is here nor there on low level sensors and changing tanks before they go dry. Billski 14. Dec 14, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan Oh, agreed on both points. Billski 15. Dec 14, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan On here http://www.homebuiltairplanes.com/forums/showthread.php?t=30820 we talked about my system. Redundancy in pumps is a good thing, particularly in electrically dependent airplanes. As for the float on a wire, that is kind of impractical. Vent air and overflow fuel is piped back to the tank it came from using a duplex tank selector. The float and wire method is nice but would overflow fuel onto the windshield instead of returning it to the source tank. No thanks. Billski 16. Dec 15, 2018 ### wsimpso1 ### wsimpso1 #### Well-Known Member Joined: Oct 19, 2003 Messages: 5,627 Likes Received: 2,866 Location: Saline Michigan That is the idea. The solid state things are tiny and rock solid.$129 dollars is too much but inventing new sensor systems based upon strain gages is OK? I don't have that kind of need to invent, and I do have Patents. I applied and ran my first strain gages in 1978. Even after we calibrate it and tame the output with inductors and capacitors, we would have to figure out how to get the Dynon to display the info.

Widely used on Long Ez's and their derivatives, but impractical on my ship. There is an IFR panel planned for the space between the header tank and the flight crew, which is why a coreless region on the tank or even a sightglass really is not practical.

What about the COTS low level sensors I proposed?

Billski

17. Dec 15, 2018

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,627
2,866
Location:
Saline Michigan
I can guess as to why, but would really rather have your rationale as to why...

I am all over doing it that way, and it seems your reason is you have seen too many failed float type sensors. Me too. Anybody have any information on the reliability of the solid state sensors?

Billski

18. Dec 15, 2018

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,627
2,866
Location:
Saline Michigan
Dan,

Now that right there is some useful information. Have you got any ideas on what the circuit would look like and what the cap and resister specs should be for the first try at it?

Billski

19. Dec 15, 2018

### Chris In Marshfield

#### Well-Known MemberHBA Supporter

Joined:
Jan 29, 2008
Messages:
1,200
984
Location:
Germantown, WI USA
wsimpso1 likes this.
20. Dec 15, 2018

Joined:
Mar 14, 2008
Messages:
4,555