I’ve a potentially simple question that I’m hoping to get some help with.
I’m looking for a way to create a super simple fuel gauge display for my vintage motorscooter.
I’ve bought this digital sender unit that uses a pressure transducer on the fuel line to give a resistance value based on the amount of fuel in the tank (much like a physical fuel sender would).
Because I was wanting to keep the instrumentation as minimal as possible I was thinking it’d be great to use a basic digital voltmeter as the gauge as wanted to show the battery level anyhow (so could even switch between the two and use one gauge) or even a dual gauge like
The fuel sender can give a range of 0 to 330 ohm or 0 to 510 ohm (or vice versa).
I’m also using a Motogadget M-unit so can get a clean power output from the Aux terminal on this.
Ideally I could use the signal from the sender and produce a voltage between 0 and 10 volts so the voltmeter would show 0.0 for empty and 10.0 for full (which is equivalent to 100%) so easy to see remaining fuel at a glance.
Is there an easy way to create this voltage reading for the voltmeter?
I’m pretty new to electronics etc though happy to muddle my way through : )
In any case, converting the resistance range to a voltage signal is going to require a little bit of work.
I don’t think a simple voltage divider will cut it when you need to cover 0-10V since when a single resistor in the divider is varied I think the response is nonlinear.
My intuition tells me you might need to look at building a wheatstone bridge. My example uses the sensor as Rx and assumes 540->0 ohm range
The bridge will create 0-6V (the voltage is negative, just reverse the measurement probes
from here you can put the voltage through a gain and you’ve got 0-10V for your meter.
Be careful with referencing ground here: the Vout voltage is not referenced to ground, but measured across the two terminals of Vout.
That is, a non-inverting amplifier where the reference voltage is the negative terminal of Vout, and not ground
Thanks so much for the detailed suggestion, super helpful!
I suspect the non linear nature might be a problem - will have a better look at it all.
Was also thinking using an Arduino might be a good solution, from the little I know I can output up to 5V and then just double the voltage for the meter?
Hi Matt
Referring to linearity. How linear is your fuel tank itself.
On a thing like a motor scooter it could have all sorts of shapes. It say half the vertical level (is what you will be actually measuring, the vertical level) actually half the volume.
Just a thought when you decide what level of inaccuracy you can tolerate.
Cheers Bob
Another thing. Beware of electrical bits in the vicinity of the petrol tank. O personally would go for a calibrated piece of wood.
The tank is actually reasonably linear, though the pressure transducer sender (that is just attached to the fuel line) can be calibrated at 0% 25% 50% 75% and 100% so this will mean the resistance values will relate to the remaining fuel pretty accurately.
So apart from this there won’t be any electrical bits near the fuel tank.
Hi Matt
If the transducer and sender are designed for fuel you will (should) be OK. But if the bits are designed for a water tank and you are adapting at all I would be a bit (read very) careful.
Cheers Bob
I had a Vespa back in the very late 50s/early 60s but I can’t think what the fuel system looked like. Maybe I just looked into the top to see how much was left. Can’t remember. Lambretta was another one around at the time. The Port Kembla Steel Works had about 20 of these.
Pretty modern in comparison. Built nearly 20 years after I had my last ride on one. It was Miles then (not km) but that does non sound like good economy for that type of vehicle.
Cheers Bob
Hi Matt
Just had a quick look at the installation manual for the device you linked and I don’t think it is as simple as it first looks.
I am not too sure how fuel gauges work actually as they must be independent of battery voltage or they would go up and down with varying alternator voltage across the battery. I think a Wheatstone bridge comes into it somewhere but I also think different manufacturers have their own ideas about how it is done.
Concerning your digital volt meter. This has to be powered. The simplest hook up uses the measured voltage to power the meter. That limits the measurement range to a minimum of about 4V. You can use an external source to power the meter such as the 12V battery but the negative is usually common ground. That is measured voltage and meter supply so you would not be able to connect it to the Wheatstone bridge as with this configuration the measuring points are floating.
Bear in mind I am referring here to the simple digital meter type sold by Core and others which measure Volts and Amps and are sold for very few dollars. There are lots of other configuration options around and I have no intention of researching them all. I will leave that up to you. At this stage we don’t even know which meter you have or maybe you haven’t purchased yet.
This LSK device looks pretty handy for after market fit to anything but you really need to know which Gauge it is designed to operate with and most importantly exactly how it works. I think it will work with 3rd party gauges but they have to fit into this system’s resistance range.
In a nutshell the simplest way would be to purchase a gauge from LSK which will work with this system.
Not as much fun though is it.
A Google on “Fuel Gauge” might yield some more info but I don’t really have the time just at the moment.
Cheers Bob
You will note the indicator is heavily damped. By the slow response of a heating coil and in the case of microprocessor control averaging the readings over a considerable time.
If you just arranges the sender resistance to be a part of a voltage divider and connected a voltmeter to that the meter would be jumping all over the place. You might slow it down with a cap of a few hundred or so µF (or thousands). You will need to regulate the supply so you will not get 10V with a linear regulator. If you allow for a battery fluctuation of between 13.8 and about 11V so allowing a bit of required head room your regulated DC would be 8 - 9V. May be possible with a buck/boost switch mode regulator adjusted for 10V. Don’t know.
Have a think about it.
Cheers Bob
Should work. I have used 5V as a regulated supply to give you a bit of head room and allow for battery voltage variations. If you set your sensor for 0-510Ω or 510-0Ω you can use a 510Ω series resistor as a voltage divider. You must use a voltage divider as if you try to get 0-10V swing you will have a catastrophe on your hands as at some stage the 12V supply will be nearly shorted.
This will give you a 0-2.5V swing at the divider output. If you want 0-10V you will need to multiply that by 4, hence the Op Amp. Resistor values chosen will give that gain. The resistors are 1% metal film preferred values commonly available at Jaycar and Altronics.
Ci will give you a smoother read on the voltmeter otherwise you might not see anything meaningful.
Could be quite a few µF. Suggest if a moderate value you use Tantalum types.
The Op Amp MUST be a Rail to Rail type. The one drawn (LM6482) is commonly available and should be OK. Even though these Op Amps are quoted as being Rail to Rail they will not get down to 0V. The lowest practical voltage will be about 0.1V with a high resistance load like the voltmeter which you can probably live with. If you really want to get down to 0V you will have to find a couple of negative volts somewhere to connect to the negative supply pin of the IC.
As there ar only a few mA involved you should get away with a linear 5V regulator such as a 1117 5V or a 7805. You won’t have to worry about switching artefacts then.
Cheers Bob
1 Like
And you can get our latest projects and tips straight away by following us on:
