I suspect feedback from the in-tank sender may vary with temperature and/or battery voltage. Test to find out, but if this is the case then you might get better results with a binary on/off type of fuel level sensor

I was thinking about that. At startup it will see battery voltage. (12-12.6VDC) When running 14-14.5VDC. I need to put a time delay relay in the circuit for the oil pressure s/w which is normally open. That could take care of the voltage difference. Not sure how much the temperature will affect the final output voltage of the sensor at 230 ohms. I guess I need to see the voltage difference between the 250 ohms empty and the 230 ohm shutdown parameter.
The fuel can be very cold and has frozen at times, when you don't have the proper stuff in it. Once running the return line from the injectors sends back hot fuel and it does get a significant temperature change.

 

The fuel tank level sensor same resistance as GM? Stewart-Warner uses, I would first connect a gauge (get one from autowrecker), you do need a fuel gauge instead of looking in the tank all the time.

Because it is in contact with flammable liquid, you do need extra care with the wiring and circuit current-limiting under fault conditions. There is a safety aspect. Typically this is keeping the tank sender away from wiring that has power on it, and using redundant (two) resisrtors to read the sender voltage.

One problem is fuel slosh- waves or filling it while the engine is running, can cause a brief false low reading.
It is customary to average the tank level reading over a long time period like a minute, like a fuel gauge does.

Think of the circuit as a low fuel light. You can read the sensor ratiometric to system voltage so it's accurate with 12V or 14V regardless.
I find lamp on spec for a Surburban 2500 is 11-13% and off at 14-16% of tank capacity, 33-240 ohm sender, and with gauge sender voltage is 1.0V or 1.1V for that.

 

You could also use a 12V regulator to feed the fuel gauge sender, so that it gets the same 12V input no matter what the battery voltage is at. That leaves only temperature to deal with, which may or may not be significant. Just for a data point, one of my old trucks would read probably 15% different depending on the battery voltage, the difference was significant. I would like to think modern cars account for this, but back in the day with purely analog gauges this was common.

 

You could also use a 12V regulator to feed the fuel gauge sender, so that it gets the same 12V input no matter what the battery voltage is at.

that's not needed if he uses a ratiometric (bridge) circuit to detect the low fuel level.

 

If a changing supply voltage may be an issue, then powering both the sensor side and the reference side will compensate. Not an original idea, it has been used for many years. Cheap and easy as well.

 

Not an original idea, it has been used for many years.

Of course.
It has been used since popularized by Sir Charles Wheatstone circa 1883.

 

that's not needed if he uses a ratiometric (bridge) circuit to detect the low fuel level.

There is almost always more than 1 solution. A 12V regulator can be had on Amazon for around $10, which makes it an easy 5 minute solution.

 

There is almost always more than 1 solution. A 12V regulator can be had on Amazon for around $10, which makes it an easy 5 minute solution.

If the TS uses a comparator solution a "bridge circuit" is almost intrinsic to make it work correctly.

 

There is almost always more than 1 solution. A 12V regulator can be had on Amazon for around $10, which makes it an easy 5 minute solution.

Of course there are always other solutions to most problems, but my maxim is to always make it as simple as possible (but no simpler).
Adding the regulator adds unneeded cost, complexity, and another active device that can fail.

And you still need a comparator circuit, whether you have the regulator or not.

 

An LM78L12 regulator costs a lot less, and to stabilize the bridge for a comparator circuit will be entirely adequate. And while the gain of a comparator amplifier may change a bit with supply voltage, it is not likely to matter in this sort of application.