• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 1314 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: Drift on Environmental Reference Electrode

Sam Matus
Prodigy 30 points
Part Number: FDC1004
Other Parts Discussed in Thread: TIDA-00317,

I'm using the FDC1004EVM module with the TIDA-00317 design (modified for a .034" thick rigid PCB) attached to a plastic square bottle using 3M adhesive. 

I set up a system that will automatically pump back and forth between two vessels while simultaneously recording data from capacitive level sensor and a lab scale that the bottle is placed on top of. The FDC1004 is configured as follows:

If I do a basic measurement (C1/C2), the sensor is pretty steady and tracks the liquid level pretty well. However I cannot use the environmental reference electrode because it causes the liquid level measurement to be inaccurate and drift, see the following data:

I believe there are two problems with the reference electrode:

  1. The capacitance of the reference electrode changes with the liquid level. I believe this is because the trace that goes to the reference electrode positioned in a way on the TIDA-00317 such that the capacitance of the trace is affected by the liquid level. I can resolve this with a simple design change.
  2. The capacitance of the reference electrode is drifting downward over time. The drift of the reference electrode causes a drift in the level measurement approximately equivalent to 1.2 cm per hour. I have very little idea what could be causing this. Our best guess is that it relates to how the shields are set up on the TIDA-00317. Referencing Figure 12 and section 8.3.1 in the FDC1004 datasheet, CIN1 and CIN2 are referenced to SHLD1, while CIN3 is referenced to SHLD 2 when doing a differential measurement. However, the TIDA-00317 is set up such that CIN2 (reference liquid) and CIN3 (reference environment) are both backed by SHLD1. Is this intended? I may be misinterpreting the datasheet. If this isn't a problem, what else could be causing this drift in the reference electrode output?

Thanks,

Sam

  • 0
    TI__Genius 17440 points
    Hi Sam,

    Could you provide a picture of your setup so I can see exactly where your sensors are attached?

    In Figure 12 in the datasheet, the capacitance of CINx is not measured in reference to the shields. The SHLD pins are driven to the same potential as the CINx channels. In the figure, SHLDx is connected to the connected to the cables' shields, while CINx is connected to the interior wire. With SHLDx surrounding CINx, there should be no capacitance between CINx and the environment in the vicinity of the cable; all environmental shifts in capacitance should be seen by the SHLDx pins. The setup of the shield is acceptable in both the datasheet and the TIDA-00317 reference design.

    I think it's more likely that this is an environmental issue or a calculation issue. How are you calculating the liquid level? Are you using equation 2 in the TIDA-00317 reference design?

    Regards,
  • 0 in reply to Kristin Jones93
    Prodigy 30 points

    Kristin,

    Thanks for the quick reply. Your explanation about the shields makes sense. Here's a photo of my setup:

    To calculate the level, I'm taking a measurement of C1, C2, and C3 when there's no liquid in the container (calling these C10, C20, C30), then filling with liquid above the ~2cm mark on the sensor. The two equations I'm using are:

    • Without environment reference: level (cm) = (C1-C10) / (C2-C20) This seems to work pretty well.
    • With environment reference: level (cm) = (C1-C10) / [ (C2-C20) - (C3-C30) ]. This one results in a lot of drift due to the drift seen in C3.

    Thanks,
    Sam

  • 0 in reply to Sam Matus
    TI__Genius 17440 points

    Hi Sam,

    I think the calculations are the problem here. This is the equation we recommend to find the liquid level:

    Your first equation is very close to this, which is why I believe it is working. In fact, we recommend that if someone is not using an environmental reference sensor, they should use CRL(0) in place of CRE. This is exactly what you are doing in the first equation.

    In the second equation, you are including CRE(0) and CRL(0), and I believe these extra terms are causing problems. Both the environmental reference and the liquid reference will see environmental drift, which will be factored out in the denominator (CRL - CRE), so you don't need extra terms. 

    If you use the recommended equation and do not solve the drift issue, it is likely an issue with your environmental reference.

  • 0 in reply to Kristin Jones93
    Prodigy 30 points

    When I calculate according to that equation, I get liquid levels ranging from between -0.36 cm an -0.19 cm because the measured capacitance of the environmental reference electrode is so high (almost 9pF), here's the same data I posted earlier, not zeroed out in air:

    Are my measurement configurations (0x08 through 0x0B) perhaps not set up properly for use with the TIDA-00317 design?

    Thanks,
    Sam

  • 0 in reply to Sam Matus
    TI__Genius 17440 points
    Hi Sam,

    Could you verify that your reference liquid and reference environmental sensors are identical in size? In addition, is the liquid water? The dielectric constant of water is much higher than air, so if the liquid is non-conductive and the sensor are the same side, the reference liquid sensor should always have a higher capacitance than the reference environmental sensor.

    Regards,