Using capsense to sense plastic objects

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
jcsb1994
Level 4
Level 4
First solution authored 50 replies posted 50 sign-ins

IMG_20200929_133238879.jpgIMG_20200929_133147874.jpg

Hi,
I am currently working on a tactile sensor using capsense and I want the sensor to feel anything that touches it, not only fingers or conductive objects.
So far, I am testing various capsense settings with a CSX 3x5 touchpad covered by a silicone foam dielectric, which is covered by a grounded conductive cloth. This cloth lets me read a signal with plastic objects too, although very week so far...

1. Anyone has an idea of which CSX or capsense setting I should play with to help it sense objects more when touching the grounded cloth?

IMG_20200929_135645418.jpg

I am also considering ungrounding the cloth. It would become a floating conductive layer acting as the second plate of the capacitor, thus reducing the distance between the plates when an object touches it and presses on the foam dielectric.


I am confronted to a problem when I do this however. If I press on the dielectric with a plastic object, it doesn't trigger any signal, but when I release it, the elements I was touching start emitting a continuous signal as if their baseline had increased.

IMG_20200929_135527578.jpgIMG_20200929_135533912.jpg

     while pressing                                                                                     After press

2. Can someone explain what is going on and why the baseline is increasing this way? I really expected the ungrounded fabric to work!

0 Likes
1 Solution
BragadeeshV
Moderator
Moderator
Moderator
First question asked 1000 replies posted 750 replies posted

Hi JeSi_4326976​,

1. The finger is a grounded conductor. Therefore, it is not enough that you have a conductive cloth, the conductor should be grounded for the conductor to affect/ alter the field lines. Therefore a floating metal or conductor will not have much effect / signal.

2. Make sure that you do not compress the conductive object less than 0.5 mm. You can have a fixed 0.5 mm overlay on top of the sensor, one top of that you have a movable dielectric (foam of 1 or 2 mm), and place the grounded conductive material on top. You can have a overlay with high dielectric constant such as glass so that the signal is high. The rigid overlay ensures that you do not bring the conductive material less than 0.5 mm to the sensors.

Please do not crop the y axis  of the image since it contains the difference signal numbers.

Regards,

Bragadeesh

Regards,
Bragadeesh

View solution in original post

0 Likes
3 Replies
BragadeeshV
Moderator
Moderator
Moderator
First question asked 1000 replies posted 750 replies posted

Hi JeSi_4326976​,

1. Are you using any non conductive overlay on top of the sensors. It is mandatory that you need to place a non conductive material on top of the sensors? You can use the below stackup.

pastedImage_0.png

2. You should increase the Number of sub-conversion if you get very less signal.

3.

I am also considering ungrounding the cloth. It would become a floating conductive layer acting as the second plate of the capacitor, thus reducing the distance between the plates when an object touches it and presses on the foam dielectric.

I'm finding it difficult to visualize the stack-up. Could you draw an image like above and send us so that we can review it?

2. Can someone explain what is going on and why the baseline is increasing this way? I really expected the ungrounded fabric to work!

Instead of this view,  share the graph view of tuner and show us how the rawcounts and baseline behave when you do this activity.

Note that, you should not have the conductive grounded object move closer to the csx sensor than a minimum distance. Doing so will increase Cm between tx and rx segments instead of reducing. This will be seen as decrease in rawcounts. Therefore we recommend a minimum thickness of the overlay (strictly non conducting), so that you do not see this effect.

Regards,

Bragadeesh

Regards,
Bragadeesh

1. The stack would be basically the same, but the conductive cloth would be floating. Note that I just took your photo and removed the ground symbol on the cloth. I was expecting his to work as the cloth approaches the PCB, the electric field between the Tx and Rx plates would be altered by the conductive cloth. But I assume the cloth would need to have a dielectric constant similar to a finger in order for this to work?

pastedImage_0.png

2. What you describe really looks like what I am experiencing. The overlay foam is about 2.5mm thick, and can be compressed to a fraction of a mm. However, this problem occured when the cloth was not grounded, (it was floating, not connected to anything).

The following image shows what happening while pressing the sensor with a plastic non conductive stick while the cloth is not grounded. In this test, I pressed on the center element of the touchpad. The wierd thing is that while I press on the element, capsense identifies a touch on the opposite edges of where I am touching, as if the stretching of the cloth fabric in the corners was indicating a touch. (The same occurs when I press on a corner element, capsense senses a touch on the opposite side of the touchpad) When I release, however, it always can tell exactly where I was pressing with the plastic stick, but I remains  triggered as a touch forever after I release.

pastedImage_1.png

0 Likes
BragadeeshV
Moderator
Moderator
Moderator
First question asked 1000 replies posted 750 replies posted

Hi JeSi_4326976​,

1. The finger is a grounded conductor. Therefore, it is not enough that you have a conductive cloth, the conductor should be grounded for the conductor to affect/ alter the field lines. Therefore a floating metal or conductor will not have much effect / signal.

2. Make sure that you do not compress the conductive object less than 0.5 mm. You can have a fixed 0.5 mm overlay on top of the sensor, one top of that you have a movable dielectric (foam of 1 or 2 mm), and place the grounded conductive material on top. You can have a overlay with high dielectric constant such as glass so that the signal is high. The rigid overlay ensures that you do not bring the conductive material less than 0.5 mm to the sensors.

Please do not crop the y axis  of the image since it contains the difference signal numbers.

Regards,

Bragadeesh

Regards,
Bragadeesh
0 Likes