Fuel Level Sensor Testing

Nearly ran out of gas the other day- fuel light never came on, and ODO showed not nearly enough miles to be empty, but it took 5 gallons to fill, so me thinks the sensor has gone or the gas pump we used was way out of whack. So I need to check the fuel level sensor in the tank for continuity.

I recall (or I dreamed reading about it) there was a Ohm or voltage test you could do on the sensor, and I've searched through the repair manual cover to cover but havent found the procedure or base line voltage or ohm test.

Does anyone have the answer?

Thanks

I believe it is a "heat" switch. It stays cool when submeged and switches when exposed and warms up. That said, mine doesn't work even after replacing the switch. There is a thread somewhere on testing the switch. I went through the testing and was confident the switch was bad, but if it was, there was more wrong.

Racer- thanks for your comments- there are a couple of threads Ive read but nothing that suggests how to bench test the sensor. In the shop manual there are tests for nearly every other sensor and diode in the electrical system, but for the life of me I cant find anything on the fuel sensor output. I drained the right tank today and the light never came on so I suspect the sensor when south. I was digging around under the tower doing some wiring prior to this, and also had the right tank off last week, so I hope I didnt pull something and create a larger issue. Ive aleady ordered a new sensor so I will check that when I get it against the one I plan to remove and report back. BTW I ordered my parts at KTM World who has just updated their website and gotten very competitive on their web pricing, they are 30% off and free shipping over $39- pretty good deal.

Shut off the right tank's fuel valve. Keep riding until the light comes on or you run out of gas.

That tells you of the sensor is working or not. Open the right valve of course if you run out of gas!!!

ABuck99,
the level sensor is a simple thermistor, which behaves pretty much as Racer111v described. When in normal ambient air you should be able to measure something around 1kOhm at the sensor with it disconnected. The level sensor circuit monitors the current passing through the sensor with a fixed driving voltage. With the sensor in the fuel its temperature is kept relatively low, its resistance stays low and in consequence the current through it is "high". When the sensor emerges from the fuel the current passing through it causes it to heat up, its resistance increases as a result and the current drops to a "low" value. The sensor circuit sees the current drop and turns on the warning light. It's worth noting that firstly the sensor takes some appreciable time to heat up once it's out of the fuel (at least 30s... I've measured it when I rebuilt my faulty sensor, but I forget!), and secondly that the sensor circuit seems to have some intelligence... it appears to remember the last known condition of the sensor when the ignition was turned off. This second point, explains why once you've refuelled, the light remains on for a while.

It's too late if you've already ordered a new sensor, but maybe it's worth remembering that a thermistor is some cents... for the next time around??

Cheers... Paul

Thanks Keener- I menant to say I drained the other right tank(left tank) and the light never came on. But yeah that's the way to confirm it for sure. Just a note though as you run the fuel down in the left tank below the pick up the FP cycles continuously- which is prob not good.

Paul- Thanks for explaining the dynamics of the sensor- I put an ohm meter on the sensor side plug and it was reading .064 ohms, or 6.4 depending on where you put the switch. Admittedly about the only thing I really understand on that damn multimeter is checking a battery . So to your point about the time it takes the thermo resister to cycle once it's In or out of the fuel. After draining the left tank as much as I could by siphoning (right tank was shut off) I started the motor and let it run for maybe 5 min. I am very confident I drained well below the sensor. The light never came on. I wonder then how long it would take the sensor to register low fuel condition.

If you have a suggestion on how to rebuild the sensor, I like that idea. I've already ordered a new sensor, but I might as well try to fix the bad one.

Thanks

Andrew

Andrew,
first of all, a correction! The thermistor is actually an NTC type in this application, so as the temperature increases the resistance decreases... not how I described above. Sorry, I was clearly in the throes of caffeine deprivation when I wrote that!

The thermistor is mounted inside small a metal can, which is soldered to a little PCB on the sensor housing. Take the whole sensor out of the tank and it should be obvious. You need to remove the can and remove the thermistor inside by desoldering. Replace it with something like this
http://cpc.farnell.com/epcos/b57164k102j/thermistor-ntc/dp/RE06868
polarity is not important, just cut the legs to match those of the old part and solder into place. Solder the can back onto the PCB and replace the sensor in the tank, being careful not to over-tighten the mounting screws.

With this thermistor it should be something like 45s before the low-fuel warning light comes on. I believe the OE thermistor will be a little slower, needing at least 60s.

Let us know how you get on.

Cheers... Paul

Paul- thanks for the added info. Soldering electrical stuff is easy enough for me, but the idea of soldering electrical stuff that lives in highly flammable liquid-that I then sit on top of gives me the some second thoughts . But I'm going to give it a go cause the project sounds like fun. I will install the new sensor since its been ordered, and I dont have time to fix the old one before my trip, but I will get to that when I return.

Cool stuff- thanks!

Received the new Fuel Sensor yesterday and put a multimeter on the new sensor & compared to the "bad" one.


Bad Fuel Sensor Test
Setting the multimeter to 200 ohms revealed a value of 1
Setting the multimeter to 2k ohms revealed a value of 1
Setting the multimeter to 20k ohms revealed a value of 6.96ohms (me thinks its toast)

New Fuel Sensor test
Setting the multimeter to 200 ohms revealed a value of 1
Setting the multimeter to 2k ohms revealed a value of 0.864ohms
Setting the multimeter to 20k ohms revealed a value of 0.87ohms

Guess its time to pull the tank and get busy swapping sensors

FYI- 2006 950S Adventure @ 33k miles

My sensor is wonky. It'll flash the light a couple times but eventually the light turns off until I putter out of gas.
'05 with 42k on it.

You can pull the connector apart and test the sensor now that you know what the correct base value is. The connector will be haniging above the stator cover. Granted my new sensor was bench tested, the bad sensor is still in the tank but the fuel temp and ambient air temp are relatively the same. And based on Paul's post above, your can rebuild yours if you're inclined. New fuel sensor is about $45.

I'm not sure it's that simple. You tested with the new sensor on the bench and the old one in the tank. I know you said the fuel and ambient temp were the same. I tried that same test and when I removed the old sensor and had them on the bench side by side they both gave the same readings. The old sensor definitely didn't work correctly (didn't trigger the low fuel light) and when I installed the new unit everything worked.

Thanks for weighing in Boatman- I have a lot of respect for your experience(s) here. I would agree with your assessment, The readings in and out of the fuel may be different good or bad. I plan to test the bad sensor out of the tank just for comparison. I'll post that data up over the weekend. Thanks for your input.

I had the tank off (oil change, etc...) but I couldn't get it to work with either the new or old sensor. I'm thinking it must be an issue in the dash. I'm so used to not having it at this point I don't really care.

Looking forward to seeing what you come up with on the bench. I have a feeling the testing can only be done with the sensor at certain temps.

Ok recapping the fuel sensor install & bench test results.

Boatman- as you predicted the Ohm values for both the good sensor and the bad sensor were a little different in tank and out of the tank.

The Bad sensor
Originally in the tank the bad sensor produced a value of 6.96Ohms, and outside the tank on the bench read 5.95 Ohms so a difference of 1 Ohm. (But its still toast)

The New Sensor
On the bench the new sensor produced a value of of 0.87ohms, and in the tank produced a value of .75Ohms.

Next is road testing, but bike is half torn apart prepping for IDBDR- so wont know for another week if the new sensor did the trick. Fingers crossed.

Thanks to everyone for all of the input and commentary.

Me and my friend Alysson idle fixed the fuel sensor its a Thermistor ntc 1k, is easy to fix the cost is 1usd.

Comparing with original.


The difficult is disassembly you need take care.

Nice work.
Thanks for posting up in the other thread. I'll be looking for a thermistor to fix my spare sensor.
Was the can crimped over the edge of the board? I'm trying to understand the need for the Dremel grinder.

We use a emery tool, its in Portuguese see Google translate

https://translate.googleusercontent...ml?m=1&usg=ALkJrhivOUOEmNaHFCzIt3zkvGy80zSSMw