• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 1050 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: Regarding shielding and interference

Ethan Lu1
Prodigy 50 points
Part Number: FDC1004

I am working on a capacitive sensor that measures distance to a floating (rather large slightly curved aluminum) object (inside an instrument) around 5mm away accurately (~0.1mm) using your chip. The idea is to use two side by side metal plates, one attached to sensor and another to ground, so that together with the object they have the geometry of two parallel plate capacitors in series (of course ignoring edge effect)

First of all, I am wondering if it is okay to shield the ground cable, if the cable length is at most 2~3m (only because the cable we got is suitable for such configuration)? 

Secondly, I noticed that the capacitance is referenced to ground. But I still do not quite understand what ground means in this case. For instance, why are things like fingers (or even floating piece of metal) grounded, and not just the ground cable. In fact even powering the chip using a battery pack (so that it's definitely not connected to the electrical ground) doesn't mitigate such effect. I am afraid that this uncontrollable grounding situation can potentially produce strong interference on the (already weak) signal (e.g. when someone touches the instrument, albeit unlikely to actually happen), and therefore believe that I would need to put a lot of thoughts into canceling them with reference sensors. Any suggestion on that matter is much appreciated.

  • 0
    TI__Expert 7599 points

    Hello User,

    Thanks for your question. Our FDC expert is on vacation and will get back to you by end of next week. Meanwhile please take a look at the FDC1004 product folder. 

  • 0 in reply to Arjun_Prakash
    TI__Genius 17440 points

    Hello Ethan,

    You could connect your cable's shield to GND, but I would not connect it to the SHLDx pins. The SHLDx pins are only effective shields for the CINx pins, because they used the exact same excitation signal. Connecting the cable's shield to SHLDx would just add a lot of load capacitance to the SHLDx pins (which can drive 400pF max) for no gain.

    Before I address your grounding questions, have you considered using inductive sensing? You mentioned that you have a fairly weak signal from your capacitive sensing implementation, but inductive sensing generally performs much better than capacitive sensing for metal targets. If you haven't considered it yet, I recommend looking through the FAQ in my signature below. Your aluminum target and target distance sound like a great fit for inductive sensing.

    Best Regards,

  • 0 in reply to Kristin Jones93
    Prodigy 50 points

    Hello Kristin,

    Thank you for the answer. I've been busy with other stuff recently and couldn't respond earlier.

    Regarding the cabling, though I am not totally sure what you have in mind, I suppose the point is that I really should separate the shield from the ground (say by feeding ground in a completely separate wire) because of the extra load on the shield. 

    I did look into inductive sensor, and it looked promising. However, I am probably going to stick with the capacitance plan for now since it's close to being done, so might as well give it the best shot. With that being said, I am wondering if the fact that the target (whose axial movement we are detecting) also spins around laterally (say on order of 1 meter/s) would be an issue.

    Finally, it would be great if you can explain a bit about what's going on with the ground. I am guessing it might be quite complex and an inherent limitation for the capacitive method, but it would be nice to know just for the sake of curiosity.

    Thanks again,

    Ethan

  • 0 in reply to Ethan Lu1
    TI__Genius 17440 points

    Hi Ethan,

    Happy new year! Thanks for your patience while we were out of office for the holidays.

    Yes, you got the point of my comment. If you can minimize proximity of SHLDx signals to GND in both layout and cabling, you'll minimize unnecessary load on the shield.

    I have seen other customers successfully use inductive sensing with spinning targets.

    Sometimes targets are connected to earth GND. When they are not, we still consider them to be virtually grounded. Even if they are not directly connected to earth GND, they still have a different potential than the FDC1004's sensors, and they also have a capacitance to earth GND. The difference in potentials creates a variable capacitance that loads the sensor.

    In order to mitigate interference from unintended targets, the SHLDx pins can be used to create a unidirectional sensor. If there is a layer of the SHLDx signal in between the sensor and the interfering object, the object will instead forma capacitance with the SHLDx pins and will not affect the measurement. 

    Best Regards,