low fuel alert

[dacflyer]

i have a universal fuel sender. 33 ohms ( full ) and 240ohms ( empty ) and the gauge that works with it.
i am wanting to add in a low fuel alert ( led and a alarm ) possibly a "sonalert"
i need the device to have a set point to where i want the alert to activate.
i was thinking to use a OP Amp or a comparator. which would be better to use?
need something to get my attention of low fuel in case i do not check the gauge often enough.

[MrAl]

Hi,

You should probably find out how the gauge energizes the sensor and then we can figure out what to measure and what to expect from those measurements.
For example, does the gauge use a constant current or a constant voltage ?
I am assuming you still want the gauge to work after the retrofit.

[dacflyer]

not sure how the gauge works..it has 3 connections. it is a analog gauge.
the connections are
pos. neg. and sender. the sender has one side to ground. the other goes to the gauge
i have no schematic of the insides of the gauge..
the face plate has MW on it, and the back has a sticker that reads 2800, oh lets not forget
MADE IN TIWAN
that's all i can tell ya...

[dacflyer]

any updates yet ?

[MrAl]

Hi again,

Well, if it is a resistance then you should be able to measure that resistance with a circuit and then compare the measurement to the known low point and determine if the measurement falls below a threshold and that will tell you it is low.

The only problem is when the gauge itself works, it already has it's own signal that is sending some kind of test signal through the resistance to measure it. That means we have two choices: we can either learn to live with the gauge and find out what test signal it is putting out and use that to make our own measurements, or we can disconnect the gauge for s short time so we can use our own test signal to make the measurement.

To disconnect the gauge though we have to know what it will accept as a disconnecting switch. Bipolar, MOSFET, or do we have to use a reed relay? It might take some experimenting, but once this is found out then we can incorporate probably a simple circuit to make the measurement. With fuel gauges you dont have to update too often anyway, so an interruption only has to occur maybe once every 10 seconds or something like that.

The problem with using the gauge as is without disconnecting is we have to study the system to figure out what the gauge is doing to make the measurement. It could be something very simple like a current or voltage or it could be a variable signal so it might get too complicated anyway. This make disconnecting the gauge for a short time period look like a good idea.

What do you think?

[Einar M]

What I would do is measure the voltage on the 'sender' wire as the tank goes from full to empty (or vice-versa). Let us know what you measure and then we can suggest more. Most likely a comparator swicthing an oscillator or buzzer will work but we need more data.

[dacflyer]

the sender is basically a rheostat. without the gauge connected, the voltage would stay the same.. i am gonna try to research more on what makes the gauge tic. or i might have to measure voltage at the "S" ( sensor ) on the gauge as the tank level changes.

[Einar M]

It seems most likely that the sensor voltage will change from a low amount (1V?) at full to a higher reading at empty, but you will need to measure it to be sure. I can easily make a circuit that latches a led or buzzer on above a set voltage or have them flash or beep once, or flash or beep repeatedly with the pulses getting closer together as the input voltage increases above the trip point. If you have something else in mind let me know.

[dacflyer]

well the tank is 33ohms when full...and 240ohms when empty.
i been thinking to use a op amp. and just let it flash a bright red led so it is easily seen in the sun.
i think a buzzer is useless, since its a open cockpit and the engine noise will drown out the buzzer.
the red led will be easily seen.

[Bob Scott]

Comparator or op-amp? If you are operating the device with the + and - inputs at different voltages, use a comparator. An op-amp circuit always has both inputs at the same voltage. Some really old op-amps will not work, may invert the signal, or will lock-up with inputs at different voltages.

Most modern op-amps are useable as comparators and do not mind the inputs having different voltages.

On the fuel level warning circuit, I recommend using some hysteresis so that the warning device does not chatter. It should work like your furnace thermostat. It has a little positive feedback. And being in a vehicle in flight, dependability or redundancy so that you don't die.

Remember magazines like Popular Science predicting that we'd all be flying around in our own personal helicopters by now? Can you imagine the daily carnage just from commuters running out of gas? Or "fender-benders" in three dimensional space? You can't just pull over to the shoulder!

[dacflyer]

lol. ya i just do not see the flying age yet, people here cannot drive as it is, forget the air..

ok, some new news, something told me to bench test the fuel gauge before i installed it.
its bad, so i dismantled it and seen what makes it tick.. Interesting to know this unit does not have the workings of a normal meter, (coil around a magnet) nor is it a thermal unit
(bi-metal flexing with heat) this unit has 2 coils in a X-Y set up.. one coil is N-S the other is W-E
apon closer inspection i seen the midpoint of the coils connection had broken from vibration fatigue i guess ? and i am suspecting the needle armature is magnetic. so i am guessing one coil was a reference coil and the other was manipulated by the current from the sensor to deflect the needle on the gauge.. simple design, no electronics , regulators or heaters.

so with that info maybe one of you can better help me decide what sort if unit to use.
i am not sure how to incorporate a hysteresis into the unit. as far as connections go i should be able to just connect onto the sensor.

as far as calibrating it, i'd fill 3 gallons in the tank ( approx 1/4 full ) and set the unit.
( 10 gallon tank full )
and i still plan to use the bright led.. i bought a new gas gauge from O'Reillys auto parts, they have a slew of after market meters now..all sorts. and styles.

[MicroRem]

DacFlyer said "i think a buzzer is useless, since its a open cockpit and the engine noise will drown out the buzzer."

I remember my flight instructor saying there is nothing louder than the sound of your engine stopping in flight...

best

Tom

[dacflyer]

yes, i agree that a buzzer is useless , especially when the engine is just behind you. and no firewall
but i just wanted to share with all what type of fuel gauge i had..
never seen such a style before..
i plan to start bread boarding the circuit this weekend..

[rshayes]

The indicator sounds like a style of gas gauge that was used on old cars (possibly before WW II). The position of the needle depends on the ratio of the currents in the two coils and not on the magnitude of the currents. This is a superior arrangement to the hot wire indicators which followed them. They do not need a regulated power source and the calibration is probably more consistent and stable. The main disadvantage is that is is probably more expensive to manufacture and can only cover a 90 degree arc. In Detroit, expense is the major consideration. I suspect that is the reason that this type of indicator was replaced by the hot wire type.

I would expect that the power source would be connected to the common connection of the two coils. The other end of one of the coils would be connected through a resistor to ground as a reference. The other end of the second coil would be connected to the sending unit. If the two coils have the same number of turns, the resistance of the reference resistor would be equal to the resistance of the sending unit when the indicator is at half scale.

The reference resistor could be eliminated by using smaller wire and/or resistance wire in the coil connected to ground.

The coil in series with the sending unit may also have a substantial resistance. If this is the case, the voltage on the sending unit could be compared with the voltage from a voltage divider connected between the power source and ground. This would create a Wheatstone Bridge which would be balanced at one value of sending unit resistance. A comparator can easily detect this point. If the input impedance of the comparitor is much higher than the sending unit resistance, the gauge would function normally.

This gauge may be designed for a 6 volt electrical system and might burn out if operated on higher voltages. An appropriate zener diode in series with the supply voltage would allow operation on higher voltages.

[dacflyer]

actually the gauge i described is what was used on many AMC type vehicles and a few others.
the gauges can be bought at O'Reilly's Auto Parts.
http://www.iequus.com/Product/Search part # 7363
nice video there also..
and yes, it is 12 volts operated.