• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 978 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

FDC1004 and TIDA-00317 for gasoline level measurement

Glen Meloy20
Prodigy 20 points
Other Parts Discussed in Thread: TIDA-00317, FDC1004

I've been asked to make a fuel (gasoline) level sensor and I've been experimenting with the FDC1004 EVM and a homemade TIDA-00317 sensor. 

For my first test, I taped the sensor to a thin plastic container and filled it with water. It worked very well for that, ranging from about 3pF when empty to nearly 12pF when full.

For the next test, I taped the sensor to a standard red gas can, about half full of gas. I figured I could tip the gas can back and forth to simulate changing fuel levels - though obviously the thickness of fuel in front of the sensor would be uneven as a result of tipping it compared with actually filling and emptying it.

I can see a difference between full and empty, but not much - it looks like about 0.85pF change from full to empty with the channel gains set to 1, or about 2pf with the gains set to 3. The sensitivity is far far lower for this than it was for the water test and it doesn't look like there is enough range to be useful.

I see that the dielectric constant of water is about 80 and for gasoline it's only 2. Is this the primary reason for the difference in performance? I know the plastic of this container is thicker than what I used for water and that my tipping-the-container method is not ideal, but it seems like both of those effects should be relatively minor.

Do you have any advice for increasing the sensitivity to make an acceptable sensor for gas levels? We can make a new sensor if needed. It looks like making it wider will increase the sensitivity, but I don't know how much wider it needs to be to perform well.

Is this OOP technique and this sensor the right solution for fuel level sensing - or is there something better? An in-tank design would be acceptable, if capacitive sensing from outside the tank just isn't the right fit for fuel.

Thanks,

Glen

  • 0
    TI__Expert 6480 points

    Hi Glen,

    Your observations are consistent with what we would expect - the dielectric of gasoline is 40x less than that of water which is the main reason for the lower sensitivity. The thicker plastic is definitely also a factor so it would help if the container walls were thinner. We would still suggest using the OOP technique as it uses a ratiometric approach to mitigate temperature effects and hand interference. You could try using the immersive straw approach if you wanted an in-tank design.

    On another note, a dynamic range of 2pF could be acceptable depending on what you are looking for. What is the system level noise you are seeing without any liquid in the container? How many levels are you looking for?

    Thanks,

    Rachel

  • 0 in reply to Rachel Liao
    Prodigy 20 points

    Thanks for the response. I'll look into the noise and # of levels requirements questions.

    Is there a tool or a list of rules for customizing sensors?

    There is some general information in the Sensor Design Considerations section of TIDU736A about how increasing the electrode width increases sensitivity and a few other tidbits, but I'd love to have more information to help me determine how wide I should make it for my application. I feel like I need more sensitivity, but I don't know how much more I need or how much more I get by making the sensor X percent wider. 

    Also, the tank that we're trying to measure is very short - maybe 2" tall total. Is it ok to shrink the pattern vertically, which would result in pretty squashed reference liquid and reference environment pads?