# Dual fuel sensors, one gauge

#### Grinchy

Joined Mar 23, 2021
17
<will use a resistor network. For the discussion, I had the ohm values swapped, it is 410 ohm empty and 15 ohm full>

i have added a second fuel tank to my rig. both tanks use the same 15 ohm empty to 410 ohm full fuel senders (resistive pots?). The gauge is a voltage based gauge and reads the voltage drop across the sender resistance relative to 12v.

I want to build a simple circuit that switches the resistance the gauge sees based on the resistance of tank 2.
example a. Just filled up. Both senders read 410 ohm. gauge shows full and should be reading the resistance of tank 2.
example b. Tank 2 is empty, tank 1 is full. Tank 2 reads 15 ohm, tank 1 reads 410 ohm. Gauge should show full, and display tank 1 Value.
example c. Tank 2 and tank 1 are empty, both at 15 ohm. Gauge shows empty, and is displaying tank 1

I have a couple basic parts kits with resistors, diodes, and npn and pnp transistors available. I’ve been reading around and searching and found some thermistor and photo resistor examples that are in the ballpark, the part I’m having trouble with is where to ‘insert’ the fuel sender pots.

sadly the tanks are not the same volume, so a simple divider won’t work, but I’m totally open to a ratio divider network.the tank volumes are tank 2 12 gal and tank 1 25 gal.

Last edited:

#### crutschow

Joined Mar 14, 2008
26,992
That all seems rather confusing.
Don't you want to know how full either tank is?
The way you describe it, you won't know how full tank 2 is.
And what happens if tank 1 is empty and tank 2 is full?

Seems like you just need a switch to go between tanks, or add a gauge for the second tank.

#### Grinchy

Joined Mar 23, 2021
17
Thanks for your response. Tank 2 must empty before tank 1 begins to empty.
the intent is to use the gauge twice per fill, on once for tank 2 then again for tank 1. The gauge can read only 410 to 15 ohm (scales against 12v), so I’m looking to Automatically switch the tank that is read based on the resistance of tank 2 if tank 2 is 15 read tank 1, if tank 2 is anything but 15, read tank 2.

they are in series. So we can ignore conditions where tank 2 has fuel and tank 1 is anything but full.
attached are the combined and desired resistance values if a divider network was to be used., it is suitably non linear so I’m looking for another solution, such as using the gauge sweep twice, once each for each tank.
it is not convenient or cost effective to add a second gauge.
I could add a physical switch and manually switch the gauge, but where is the fun in that?

#### Attachments

• 205.1 KB Views: 2
Last edited:

#### Grinchy

Joined Mar 23, 2021
17
Of course, the logic could also be, if tank 1 is 410 read tank 2, if tank 1 is less than 410 read tank 1. There might be a discontinuity at exactly 410 ohm here, where the gauge reads zero and there is fuel in tank 1, but I plan to add an led to the tank 2 side of the circuit in any cas, and it would cover this issue.

#### crutschow

Joined Mar 14, 2008
26,992
The problem I see is that the tank sensor voltage will change significantly when the gauge is connected and removed from sensor 2.
Need to know what the sensor voltage is when the gauge is attached when both empty and full in order to suggest a design.
Saying it reads the voltage drop across the sensor resistance gives me no information on the voltage value.

#### Grinchy

Joined Mar 23, 2021
17
Unfortunately I don’t have access to the voltage the gauge is reading as it is in the dash, but let’s assume it reads 12v at 15 ohm and 0 volts at 410 ohm. On the gauge scale 0 volt is Full and 12v is empty.

#### Yaakov

Joined Jan 27, 2019
2,902
The obvious (though possibly incorrect based on not enough information) solution is a comparator that switches the meter to whichever tank is producing the highest voltage and a second that disconnects thank 2 if it reaches whatever corresponds to “empty”.

#### boostbuck

Joined Oct 5, 2017
40
If you are looking at your gauge and it reads "1/3 of Full", how do you know if you've got 1/3 of Tank 2 PLUS Tank 1, or just 1/3 of Tank 1 and an empty Tank 2? The gauge reads the same for each case.

#### Yaakov

Joined Jan 27, 2019
2,902
If you are looking at your gauge and it reads "1/3 of Full", how do you know if you've got 1/3 of Tank 2 PLUS Tank 1, or just 1/3 of Tank 1 and an empty Tank 2? The gauge reads the same for each case.
I believe the TS is trying to work around the fact that the gauge is only able to deal with the resistance range of one sending unit at a time and still produce correct indications. I think I would add an indicator to show it had switched to tank one but he doesn’t seem to have access to the panel, so…

#### boostbuck

Joined Oct 5, 2017
40
It seems the TS is regarding both tanks as effectively a single reservoir of fuel, and the gauge could reflect this. Assuming for example that tank 1 is one third the volume of tank 2, then by paralleling the Tank 2 sender with 750 ohms and the tank 1 sender with 200 ohms and connecting them in series, the gauge would read 'full' when both tanks full, down to '1/3' when tank 2 is empty, and on down to 'Empty' when both are empty. Correct divider values would be calculated on the actual ratio of tank volumes. The linearity of the gauge would of course would be poor with a simple divider, but it would provide a simple indicator of the combined volumes and remaining fuel.

#### Alec_t

Joined Sep 17, 2013
12,000
Tank 2 must empty before tank 1 begins to empty.
they are in series.
Those statements seem at odds. If fuel flow is stopped at one tank then it will be stopped at the other, too. Normally, dual tanks would be in parallel, even if used sequentially.
Are you sure the sensor voltage is referenced to 12V? Modern gauges typically have a 5V stabilised supply derived from the 12V one.

Last edited:

#### crutschow

Joined Mar 14, 2008
26,992
The obvious (though possibly incorrect based on not enough information) solution is a comparator that switches the meter to whichever tank is producing the highest voltage and a second that disconnects thank 2 if it reaches whatever corresponds to “empty”.
The problem I see is that the meter is likely part of the circuit and, when it is disconnected, the sensor voltage will go to either 12V/5V or 0V.
Without knowing how the meter/sensor works it seems problematic to come up with a working solution.

And, as noted, I don't seen how the tanks can be in series.
If the tanks are connected together then the fuel level in both would be the same.
It seems that to empty one without the other, would require valves on each tank.

#### crutschow

Joined Mar 14, 2008
26,992
let’s assume it reads 12v at 15 ohm and 0 volts at 410 ohm. On the gauge scale 0 volt is Full and 12v is empty.
Sorry, but a working circuit can't be designed on an assumption.

#### Grinchy

Joined Mar 23, 2021
17
Thanks all for jumping in to help; sorry my description wasn't complete.

Here's how the fuel system works. The two tanks are in series. Fuel is drawn only from tank 1 to the engine. However, tank 1 runs a pump that continually refills tank 1 from tank 2 until tank 2 is depleted/empty. At that point fuel continues to be drawn from tank 1 and tank 1 will be depleted.

There are electrically identical fuel senders in both tanks, which have a float. The float is on a potentiometer and changes resistance as the tank is depleted. The full resistance is 410 ohm, the empty resistance is 15 ohm.

I have access to the output of both fuel senders, and can serial, parallel, network, and otherwise use these. I have access to the resistance input of the fuel gauge.

Factory: Tank 1 fuel sender is input into Fuel gauge

Modified: Tank 2 fuel sender is input into Fuel gauge // switch condition // tank 1 fuel sender is input into fuel gauge

I see in my head a way to evaluate one of these conditions and switch the input source to the fuel gauge. They don't all have to be true, just one . . .
Condition A - Tank 2 resistance is 15 ohm (ie, tank 2 is empty)
Condition B - Tank 1 resistance is less than 410 ohm (ie, tank 2 must be empty as tank 1 is no longer full, and tank 2 will be depleted to keep tank 1 full)

The car runs on 12v (14.4 v alternator), I could get a small converter if a second reference voltage is required, but the simpler this circuit is, the better.

Yes, a voltage divider could be a solution, but the divider amount is not linear, so the approximation is quite poor. I'm ok with having the gauge sweep twice.

An LED could be added to the tank 2 side of the circuit to indicate which sweep the gauge is on.

#### Grinchy

Joined Mar 23, 2021
17
The obvious (though possibly incorrect based on not enough information) solution is a comparator that switches the meter to whichever tank is producing the highest voltage and a second that disconnects thank 2 if it reaches whatever corresponds to “empty”.
This sounds right.

#### Grinchy

Joined Mar 23, 2021
17
If you are looking at your gauge and it reads "1/3 of Full", how do you know if you've got 1/3 of Tank 2 PLUS Tank 1, or just 1/3 of Tank 1 and an empty Tank 2? The gauge reads the same for each case.
Great point on having an indicator so that it could be identified which tank is being read. An led could be added to the Tank 2 side of the circuit.

#### Grinchy

Joined Mar 23, 2021
17
I believe the TS is trying to work around the fact that the gauge is only able to deal with the resistance range of one sending unit at a time and still produce correct indications. I think I would add an indicator to show it had switched to tank one but he doesn’t seem to have access to the panel, so…
I can add an indicator under the dash, for example, to reference if I don't know which tank is being read.

#### Yaakov

Joined Jan 27, 2019
2,902
I can add an indicator under the dash, for example, to reference if I don't know which tank is being read.
You could use a bicolor LED and make it green for one and red for the other. That way you have a positive indication that it is working.

#### Grinchy

Joined Mar 23, 2021
17
It seems the TS is regarding both tanks as effectively a single reservoir of fuel, and the gauge could reflect this. Assuming for example that tank 1 is one third the volume of tank 2, then by paralleling the Tank 2 sender with 750 ohms and the tank 1 sender with 200 ohms and connecting them in series, the gauge would read 'full' when both tanks full, down to '1/3' when tank 2 is empty, and on down to 'Empty' when both are empty. Correct divider values would be calculated on the actual ratio of tank volumes. The linearity of the gauge would of course would be poor with a simple divider, but it would provide a simple indicator of the combined volumes and remaining fuel.
I am absolutely willing to explore some parallel resistances and not switching the inputs and going single sweep on the gauge!
Tank 2 (depleted first) has roughly 1/3 of the total volume (full 410 ohm empty 15 ohm)
Tank 1 (depleted second) has roughly 2/3 of the total volume (full 410 ohm empty 15 ohm)

I attached a screenshot earlier of the series values for some other gauge points . . .

#### Grinchy

Joined Mar 23, 2021
17
The problem I see is that the meter is likely part of the circuit and, when it is disconnected, the sensor voltage will go to either 12V/5V or 0V.
Without knowing how the meter/sensor works it seems problematic to come up with a working solution.

And, as noted, I don't seen how the tanks can be in series.
If the tanks are connected together then the fuel level in both would be the same.
It seems that to empty one without the other, would require valves on each tank.
Yes, you are right, there are some failure states I'm not trying to capture at the switching level, or exactly when tank 2 is empty and tank 1 is full, which tank to read.

As to the fuel flow, the fuel pump from tank 1 is used to 'siphon' from tank 2 using a turbulence vacuum when it pulls from tank 1 to the engine. It is not a gravity or common level system. I think they did this due to some unique regulations for pulling fuel only from the top of a fuel tank.

Thanks for working on my problem with me!