Low Level Fuel Warning System

Thread Starter

strokebow

Joined Jan 26, 2008
14
G'day,

I want this to be really simple. (Then maybe advance to more complex scenarios depending on how this goes).

Rather than measuring quality I am focusing purely on level.

I have a couple of ideas in mind...

1) A potentiometer type (displacement transducer)...
When a float moves in a tank (with the level) then a resistance is changed which when calibrated relates to a change in level.

2) Some sort of tank or tubing with relays (switching devices which are closed by a magnetic field) and having a magnetic on a float which moves with the fluid level.

NOTE: I am keeping things really basic at the moment. eg) A stationary fuel tank/container.

I have these basic ideas but do not know where to start in terms of implementing either of these ideas.
If anyone has any practical suggestions or ideas for circuitry etc. it would be VERY MUCH APPRECIATED.

Thanks in advance :)
 

SgtWookie

Joined Jul 17, 2007
22,230
Autos/trucks typically have a potentiometer with a float attached to a swinging arm that changes the position of the wiper arm. One end of the potentiometer is grounded, and the other end is supplied with a very low voltage at low current to prevent any possibility of arcing/sparking inside the fuel tank, which would be disasterous. The fuel gauge is actually a milliamp meter that measures current flow in the circuit, but is marked in the traditonal "E \ | | | / F" style, and usually grossly inaccurate. :rolleyes:

I strongly suggest that for safety reasons, you do not attempt this project. It would involve drilling holes into a steel enclosure containing fuel/air mix, sealing said holes, and designing a low-voltage low-current circuit that could not possibly create a spark even if it failed. Even just "tapping in" to an existing fuel level circuit could create a very dangerous situation if not done properly.
 

hgmjr

Joined Jan 28, 2005
9,027
I agree with sgtwookie that there are serious dangers involved when you mix electricity and flammable fluids and gases.

Now, if you want to simply explore what is involved in monitoring the level of a non-flammable fluid such as water then I say go for it. There are many things you can learn about sensors and sensor interface circuitry that would be well worth the effort.

hgmjr
 

beenthere

Joined Apr 20, 2004
15,819
For a minimally intrusive possibility, you could try a thermistor inserted in the tank at a level, like 1/8 up from the bottom. With enough current (in the milliamp range) thermistors become self-heating. When immersed in the gas, the liquid will carry off heat, so the resistance will stay fairly high. When in air, the resistance will drop.

We're not speaking of creating an explosion hazard here - especially if the absolute current is limited resistively. A voltage comparitor would trip when the voltage fell and indicate low fuel. The arrangement would be a divider, with the fixed resistance limiting total current. All you would need would be a 3mm hole to insert the thermistor (and epoxy to seal it).

Looking through a catalog, I find a thermistor that is 250 ohms at 25 C. It is rated at .2mw dissipation for measurement, which means that any current above 16 microamps will cause self heating. The device is further rated up to .5 watt, so more heat than that will cause it to self-destruct.

Arranging a voltage divider fed by 12 volts, we use 6800 ohms in series with the thermistor. The current with the thermistor at 25 C will be about 1.7 ma, plenty to heat the thermistor, but below the .5 watt limit. Using 100 oms as the trip point (meaning that the thermistor is in warm gas sometime, and so will be less than 250 ohms), we see that the voltage indicating low gas will be 170 mv. That means the thermistor will be in air and heat to a resistance less than 100 ohms. This is easy to experiment with using water in a container.

I would suggest a thermistor, but my catalogs are for a more northerly area (G'day suggesting you are an Aussie). See if you can't find one with a similar rating in the local literature.
 

SgtWookie

Joined Jul 17, 2007
22,230
Beenthere's idea could be implemented external to a tank filled with fluid, by epoxying the device to the exterior of the enclosure, which would be much safer and easier to implement than penetrating the tank. Of course, the metal of the tank would sink away a good portion of heat all by itself, but fluids would sink far more, and more rapidly.

A 2nd thermistor would be needed to monitor the temperature nearer the bottom of the container. When there is a substantial difference between the two readings, the fluid is between the two levels.
 

Thread Starter

strokebow

Joined Jan 26, 2008
14
I agree with sgtwookie that there are serious dangers involved when you mix electricity and flammable fluids and gases.

Now, if you want to simply explore what is involved in monitoring the level of a non-flammable fluid such as water then I say go for it. There are many things you can learn about sensors and sensor interface circuitry that would be well worth the effort.

hgmjr
Hi,

Thanks everybody for feedback and suggestions.
I am merely exploring this avenue of monitoring fluid levels. A substance such as water is what I will be working with to do this project. I am just creating a model (not necessarily using a fuel tank).

Thanks
 

KMoffett

Joined Dec 19, 2007
2,918
You might also look into "Capacitive" liquid/fuel level sensors.

Here are a couple of examples:

http://www.imagineeringezine.com/PDF-FILES/capgage.pdf
http://www.trv7.com/fuel-gauges.html

I also attached a schematic and PCB layout for a capacitive level detector I made based on a QPROX IC. This one uses foil tape around the fuel line. This could be used if you had an external "site tube" on your tank. I also used the same configuration to maintain the level of alcohol in a plastic lab jar by having two bands (third band is to reduce external fields) on the outside of the container, and having the circuit drive a solenoid valve supplying alcohol.

Ken
 

Attachments

Søren

Joined Sep 2, 2006
472
Hi,

Other means would be ultra sound reflecting off the surface, weighing the tank or to build on your float principle, an opaque float could breach an IR-gate.

I've also seen it done with an optical emitter/detector with a clear acrylic (or whatever material) stick made into a right angled prism at the far end - while in air, the prism will bounce the light back to the other end, but when in a liquid, most of it will escape out the end (amount depends on the refractive indexes of course).

So many ways to skin a cat :)
 

thingmaker3

Joined May 16, 2005
5,083
I've often wondered if a variation on the sight-glass could be used for this kind of thing. One might be able to use an actual sight-glass with LEDs and photo-detectors. Optionally, one might mount the LEDs and photo-detectors in tubes inserted in the fuel tank.

Using the float idea, I've seen bilge sensors using magnetic reed switches and a magnet on the float. Multiple reed switches or Hall-effect sensors could be sealed in a tube with the float moving up and down the outside of the tube.
 

Søren

Joined Sep 2, 2006
472
Hi,

I've often wondered if a variation on the sight-glass could be used for this kind of thing. One might be able to use an actual sight-glass with LEDs and photo-detectors. Optionally, one might mount the LEDs and photo-detectors in tubes inserted in the fuel tank.
If a narrow beam LED (or perhaps a semi-LASER) was to shine through the glass at an angle, the change in refractive index (liquid/no liquid) should be detectable - never tried though, so not really sure how much it would deflect a light beam.

Should be easy to test with a LASER pointer and such a glass, but I have no access to a tank so equipped.
 

pebe

Joined Oct 11, 2004
626
Søren;53678 said:
Hi,


If a narrow beam LED (or perhaps a semi-LASER) was to shine through the glass at an angle, the change in refractive index (liquid/no liquid) should be detectable - never tried though, so not really sure how much it would deflect a light beam.

Should be easy to test with a LASER pointer and such a glass, but I have no access to a tank so equipped.
Most Canon printers use an LED to reflect light through a prism in their ink cartridges to check for ink level.
 

KMoffett

Joined Dec 19, 2007
2,918
Thingmaker1,

I used the float-magnet/reed switch to control an automatic Christmas tree watering system. We always cut our own trees, so thy are fresh and drink water at the rate of up to 1/2 gallon a day. The reed is in a 1/4" closed-bottom brass tube, with a donut shaped magnet on top of a styrofoam float. Only problem is there are 2 pump-ON conditions...with the magnet below the reed switch :) ...and with the magnet above the switch. :(

Ken
 

thingmaker3

Joined May 16, 2005
5,083
IRRC, the bilge sensors had physical stops. One kept the float from dropping below the "pump off" position, the other kept the float from rising above the "pump on" position.

Ergo I never saw the problem described from the tree stand. Said problem might be turned to advantage if a series of switches could be positioned for quadrature encoding.
 

KMoffett

Joined Dec 19, 2007
2,918
"quadrature encoding" with reed switchs...gross positioning...but I like that concept. I think I recall some company that uses something like that (probably Hall effect though) as linear encoders for things like milling machine tables.

Ken
 

Thread Starter

strokebow

Joined Jan 26, 2008
14
I've often wondered if a variation on the sight-glass could be used for this kind of thing. One might be able to use an actual sight-glass with LEDs and photo-detectors. Optionally, one might mount the LEDs and photo-detectors in tubes inserted in the fuel tank.

Using the float idea, I've seen bilge sensors using magnetic reed switches and a magnet on the float. Multiple reed switches or Hall-effect sensors could be sealed in a tube with the float moving up and down the outside of the tube.
I like the idea of using magnetic reed switches and the float. Anyone have a type of circuit that would be useful or a circuit layout? or any suggestions for a simple design?
 

cat3rn

Joined Jun 3, 2008
117
Maybe I could use the earlier suggestion of using 2 stainless steel rods but put them inside a vinyl tubes (with a tight fit) so as to not have to worry about the foam at the top causing a false reading.
 
Top