Part Number: FDC2214EVM
We have FDC2214EVM module connected to four plates via ~1m long triax cables.
The plates (5cmx5cm) are arranged around the test object as follows: one plate (labeled as "S") is sitting in front of it, one is above ("A"-plate), and two lateral plates on the left and right sides of the object ("L" and "R", respectively).
And we see difference in the device performance i.e. different capacitive signal patterns, depending from day to day when we connect plates to same channels or change plate connection pattern.
For instance, attached below graphs show capacitive signal patterns when we connect plates to same channels (A plate to ch0, S plate to Ch1, R to ch 2 and L to Ch3) on differnt times.
In the given experiment plates are satying stationary and the object is shifted by 1 mm steps away from S plate. As you see, the capacitive change is largest on the channel, connected to S plate, but sometimes we see and sometimes we dontsee capacitive changes on other channels. Similar phenomenon is observed when we change connection pattern.
All triaxes are connected to the FDC2214EVM in the same way: the center conductor of the triax is connected to channel input A, the inner braid is floating and the outer braid is grounded to the device ground.
The channel input B is not used. FDC2214EVM is set to work in differential mode. Single-ended configuration provides the same result.
The voltages on each channel were adjusted and tested with the scope to be ~1.77 V.
Did you have similar experience with the performance variation with cable connection?
Would you please recommend how to rectify it.
Could you share oscilloscope captures of the sensor oscillation? Although it is reaching the proper amplitude, I'd like to make sure that the waveform doesn't look distorted. It sounds like there's a lot of capacitance on the sensor with the triax cables.
In addition, you mentioned that INxB is not used, but that the FDC2214EVM is set to work in differential mode. Could you clarify what you mean by this? There's no register to configure that changes between single-ended or differential mode. Differential mode refers to the electrode configuration, which requires an electrode on either side of the LC tank (both INxA and INxB). If you have only connected an electrode to INxA, then this is single-ended mode.
Overall I think the ground on the triax cable shielding could be causing some of your problems. It looks like there is noise of similar magnitudes present on all channels, but the scaling of the graph minimizes the visibility of the noise when the capacitive response is strong. Are the sensors different sizes? Or the targets? In general, capacitive sensing is less sensitive when near to GND. Though the passive shielding may help with EMI concerns, it may be interfering with the sensing itself.
Current and Position Sensing
For more information, check out the Inductive Sensing FAQ
We are glad that we were able to resolve this issue, and will now proceed to close this thread.
If you have further questions related to this thread, you may click "Ask a related question" below. The newly created question will be automatically linked to this question.
In reply to Kristin Jones93:
Please see attached below scilloscope captures for "S" and "L" plates for two cases, when outer braid of triax is grounded and not grounded.
"S" plate is 11cm x 11cm and "L" plate is 5cm x 5cm in size. The drive curent registers for all channels are set to 21 and 19, when triax cables are grounded and not grounded, respecticely. Such registed values provide oscillation voltages to be <1.8 V on used In A of channels, while voltages on InB (not used) are typically higher-pleases see scope captures.
The differential/single ended mode register setting are related to the settings of the clock driver , described in the manual. However, your description of the these modes makes more sense, since I get same results for both, when using only INA.
As for the noise on different channels-yes, indeed it is of similar magnitudes present on all channels and comparable to open channel values, regardless of the plate size.
Could you kindly reccomend cable connection for cases when passive shielding is needed and without it?
L plate_In A_grounded triax.CSVL plate_In A_not grounded triax.CSVL plate_In B_grounded triax.CSVL plate_In B_not grounded triax.CSVS plate_In A_grounded triax.CSVS plate_In A_not grounded triax.CSVS plate_In B_grounded triax.CSVS plate_In B_not grounded triax.CSV
In reply to Alex Semeniuk:
Please give me some time to review the additional information. I'll have a response for you by the end of the week. In general, we do not have experience with testing remote electrodes using coaxial cables. I can provide general help, but am not an expert.
Any update on capacitive sensing dependence on the cable connection?
What cables would you recommend to use to get consistently same performance?
what are your cable recommendations to reduce EMI concerns?
My sincere apologies for the delayed reply. I don't have a specific cable recommendation for you, but I have some other feedback that may help. I strongly recommend using a differential sensor configuration (with electrodes connected to both INxA and INxB, not just the clocking register configuration). As you mentioned, I saw that your oscillation amplitudes are significantly different for INxA and INxB. This can cause a reduction in SNR and long term measurement robustness. On INxB, your oscillation amplitude is close to 2.5V in some cases, which is well above the recommended range of 1.2-1.8V. It would be best to use the differential configuration, but another option is to reduce the drive current so that both INxA and INxB are within the recommended range.
If only your electrode is located remotely, the LC tank should still be an effective narrow band pass filter for the signal, so grounding the cable may be unnecessary for EMI concerns. I definitely recommend keeping the LC tank as close as possible to the INxA and INxB pins. Grounding the cable could be helpful if you are concerned about physical disturbances in the vicinity of the cable.
Finally, are you using the internal oscillator or an external oscillator? While the internal oscillator is very stable across temperature, it has more jitter than the external oscillators we recommend. This can significantly increase your noise floor. If you need an oscillator recommendation, the one used in the FDC2214EVM is a good choice.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.