• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 216 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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: FDC1004 has large errors in output

Victor Grinberg
Prodigy 70 points
Part Number: FDC1004

Hello,

We have a design where the FDC1004 is used for water level sensing with LEVEL and REF electrodes, per respective app note.  Everything works at room temp, but has large errors when the device heats up.

When I point hot air (50degC) at the IC, the Clevel and Cref outputs jump up.  In the plot below, orange is Clevel and yellow is Cref.  As you can see, the capacitance levels increase quite significantly with heat, and as the device cools off, there is some weird quantization in both Clevel and Cref.  The quantization jumps are on the order of 0.05pF and completely mess up our measurements, of course.  Also, being on the order of 0.05pF, they are much larger than the accuracy stated in electrical specifications.

Any idea what might be going on?

I re-checked the layout and bypass capacitance, no obvious problems there.  I am running single acquisitions, so no I2C transitions occur during measurements.

  • 0
    TI__Expert 7599 points

    Hello,

    Thanks for using the TI forum and thanks for using the FDC1004. 

    Also, as per the datasheet there is an offset deviation across temperature error of 46fF and its a typical parameter. The 0.05pF seems to be inline with these measurements. If you like to measure better than 0.05pF a temperature calibration needs to be performed. 

  • 0 in reply to Arjun_Prakash
    Prodigy 70 points

    Arjun, thanks for the prompt reply!

    Please take a look at the plot in the original post.  As you can see:
    (1) the 50fF noise is occurring randomly.  It only happens at a specific temperature, not over temperature. When it happens, it happens randomly over time, at a fixed temperature.
    (2) the noise in the reference (yellow) channel has double 50fF steps, and totals at about 100fF.
    (3) the noise is in addition to the "smooth" thermally-related offset deviation, so the total error even higher still.

    For random noise the datasheet specifies RES, EON, and ERR parameters, and the combination of those is much smaller than the 100fF of random noise we are observing.

    In other words, it looks like we have this large random noise IN ADDITION to the thermal offset/gain deviation, and it's on top of the deviation and not a part of it.  Because this is time-dependent and random we can't calibrate it away.  Also, because it occurs at different temperatures for different channels, detecting it wouldn't allow calibrating it away either.

    How do we deal with this thing?  Is there some layout issue that might be causing this?

    P.S. I found another thread describing what looks like a manifestation of the same problem (not entirely the same observations, but very similar): https://e2e.ti.com/support/sensors/f/1023/t/562860?FDC1004-Temperature-Related-Offset-Change
    This thread is from 2016-2017 and does NOT include a solution, unfortunately.  It does confirm that our problem is not unique, though.

  • 0 in reply to Victor Grinberg
    TI__Expert 7599 points

    Hello, 

    Thanks for the detailed information. I understand the concern here in regards to calibration at temperature. 

    The part specifications are typical and the across temperature specs are not guaranteed so the part can vary and it needs to taken account by measuring outside of the ambient 25C.  

    To help proceed and see if there is any issue with the schematics and layout if you can send us the schematics and layout we can look into this.