how does the thermistor style fuel level sensor work?

[brad black]

so i'm about 21 hours into fitting an acewell dash to a 2005 or so m400ie.  it's really been kicking my arse: wiring diagram largely wrong, trying to work with the original 26 pin connector, working out if sensors will talk to the dash or not (mostly not).

last issue is the low fuel light.  i did have another m400ie in, but it left this morning before i realised i didn't know enough about how these late thermistor style ones work.  they started in the 749/999 and the plastic tank monsters have them, just a little steel body thing with a wire in one end and one soldered to the outer body.

i've read how they change resistance based on temp, and the fuel cools them down so they stay non conductive, but conduct when the fuel drops below them and they get hot.  but i need to know whether they are supplied 12v or not.  wire 3 into the fuel pump connector is the wire that goes to the dash, and i'm using it as the earthing side of my low fuel light circuit, as that's what it always was on the older bikes.  and it seems to me that this thermistor needs to have a voltage supply to work.  but i'm getting a constant light on, full to no fuel and even out of the tank the resistance is in the 1.2 - 1.5k range.  unhook the connector, light goes off as expected as there's no earth path.

so, can anyone tell me if wire 3 in the 4 pin tank connector (loom side) has battery voltage with key on?  and what resistance is there between wires 3 and 4 on the tank side?  i'm wondering if the wire from the dash is just an earth path not meant to carry voltage, and by putting voltage into it i've shit the sensor.  but then the sensor has no voltage to heat it.  but then again, if it doesn't conduct until it gets hot, how does it get hot in the first place if it's only two wires and there's no earth path?  if it had a supply, earth and signal out it'd make sense.

i'm at a loss.  and losing shitloads of money on the job.  seemed like a good idea at the time.

[battlecry]

Brad:  03 M800S.  The connector terminals are not listed in 1-2-3-4 order in the wiring diagram and the loom is covered, so I'll try to find terminal 3 from the tank side.  Wires in tank side of the connector have two pairs, RED/BLACK and WHITE /BLACK.  the RED/BLACK wires are bigger gauge, so I'm assuming they go to the pump.  The connector mating to the WHITE wire in the tank side I'll assume it is wire 3 you want has +12V when the key is turned on and persists after the duration of the pump pressurization squirt.  The resistance between the white and black wires at the tank connector with about 1L left in the tank and the tank vertical is 4.3K Ohms.  I hope this helps.

Could the sensor use the ground from the pump?

[brad black]

thanks for taking the time and effort, but i would expect an 03 should have a steel tank which has the old style sensor, which is a float in tube type.  afaik anyway.

your wiring guesses are correct.

was the light on with ignition on?

[battlecry]

Yes it is a steel tank.  Sorry, did not know about the 400.  No it wasn't on, Brad.  I lifted the tank to vertical to see if I could starve it and it was still off, so I have more gas than I think and the sensor is towards the back of the tank. 

[brad black]

the underlying issue here is that before the dash shit itself the low fuel light (as i'm told) worked correctly.  the tank has been nearly full while i've been working on it, so there's no reason for the sensor to have been upset by tank up and down and ignition on for periods of time.

it just doesn't do as it should (or as i'd hoped, more to the point) and i don't know if it's me or it.  and a new sensor, while not expensive, is ex italy.

[battlecry]

I wonder if the light circuit works by detecting the current drawn by the sensor.  No change in resistance but it would still need +12V feed.

I wish you could ping Suzy.

[ducpainter]

I wonder if the light circuit works by detecting the current drawn by the sensor.  No change in resistance but it would still need +12V feed.

I wish you could ping Suzy.

Where did she get to anyway?

[battlecry]

I think she is working here:

http://en.wikipedia.org/wiki/Australian_Square_Kilometre_Array_Pathfinder

[Speeddog]

Test bike '06 749s, test fuel pump assy '04 749s

                   Bike side harness
                            V
                            V
BLACK/blue    TAN/orange      BLACK     BROWN/white

    #4                  #3             #2             #1

Ground            0 V               Ground          0 V      >>> KEY OFF

Ground            12 V              Ground        12 V ~ 1 sec   >>> KEY ON
                                                            2 V thereafter

Small BLACK    WHITE           BIG BLACK     RED
                             ^
                             ^
                  Fuel Pump Pigtail


Thermistor is connected to white and small black on the pigtail, and thus to #3 and #4 on the bike harness.

If thermistor is not immersed, it has ~1 minute delay before the fuel light on the dash comes on.

[brad black]

ta.

i'll replace the thermistor when it arrives and see what happens.

maybe i should have powered it up when i had it out of the tank too see if it got hot or not.

[brad black]

I wonder if the light circuit works by detecting the current drawn by the sensor.  No change in resistance but it would still need +12V feed.

I wish you could ping Suzy.

i did some reading about how they work and some were saying it is a current thing, but it's over my head unfortunately.

[tuxicle]

Sorry to revive an old thread, but I was looking for the same info and came across this:



It's for the Panasonic ERTLG fuel level gauge. I can't say for sure if this is what Ducati uses, but they look similar. From the data sheet:

Operating Principle
While thermistor is immersed in a liquid, it has difficulties to be self-exothermic due to heat discharge to the liquid. Accordingly, the lamp is not turned on because of high resistance value and small current flow. When the thermistor is exposed to air due to decrease of the liquid, it is abruptly heated, thus decreasing the resistance value. At this time, large current flows turning the lamp on

Interpreting the Engrish, I think it has a small heater that tries to heat up the enclosure. When immersed in a fluid, the heat gets dissipated quickly by convection. When exposed to air, the rate of heat dissipation changes, which causes the temperature to go up. The sensor detects this, and increases the current flow enough to turn on an external lamp load. Looking further in the sheet, it seems to be able to directly drive lamp loads (up to 135 mA), and has a built-in time constant of about 300 seconds.

Interestingly, there are separate models for gasoline+alcohol.

[brad black]

bringing this back, i've had a similar thing to check today.

848, engine has to be running to make the light come on.  light off, nothing doing.  there is a delay, but the time i waited with the engine off was far longer than the time taken for it to come on with the engine running.  nothing else changed.