Hello ! I’m building a prototype and using an MPR121 (12 channels) to detect proximity/touch of metal objects held in my bare hand. To protect the electronics and diffuse an LED, I cover the electrodes with a thin sheet of TPU hydrogel (basically a phone-screen protector).
It sometimes works with a finger , but when I bring the metal tool close it doesn’t register at all even after connecting the tool to my hand to “boost” the capacitance.
If anyone knows a multichannel sensor that can better penetrate thin covers while supporting multiple electrodes, or an alternative material to Ecoflex 00-20 and TPU hydrogel, or any general advice, I would greatly appreciate it.
In this case, I have a copper-tape electrode and the material is Ecoflex 00-20. The interface between the electrode and the tool— the Ecoflex thickness in that area—is less than 0.5 mm, so it tears when I press hard. It only detects my metal tool when I press down very hard—so hard that I end up tearing it. I replaced it with hydrogel, but it still doesn’t register. I’m not sure if it’s also because my MPR121 is a generic one.
+1
Try copper strips of the same width on both sides of the touch strip, and connect those to ground.
Don't know hydrogel, but the water in it could short out things.
A capacitive touch sensor should work fine through an insulator (perspex, glass, wood etc.).
Leo..
Thank you so much for your advice! I tried to replicate something similar to the diagram I will share. I used the GND from the ESP32 (MY SIGNAL FELL) and also modified the MPR121 configuration so it would work with that as ground (didnt work as I expected).
The best results I’ve had so far are with a setup where I place a piece of tempered glass (a phone screen protector) between the electrodes, and then a PET film (0.2 mm) on top of the active electrode. With this, the sensor reliably detects my finger, but it still doesn’t respond when I use a metal tool.
Do you have any tips on how to improve detection of metal objects? Maybe you could share one of your previous projects or setups that worked well—I’d really appreciate it!
The diagrams are wrong.
The ground plane should be next to (around) the electrode, not under it.
Under it creates another (paracitic) capacitor, which lowers sensitivity.
The PET film makes this even worse.
Also note that the sensor chip calibrates the electrode on power-up.
So reboot after making a change to the layout.
Leo..
I wouldn't expect it to. This is because your metal tool is acting like an antenna and picking up RF interference. Also you touching your metal tool brings a whole lot of other things into the equation.
Thanks for your help! I’m using this configuration: the ring is connected to ground, and the electrode sits in the center. It detected both my finger and a metal tool. I’ll try to make it smaller.
I’m using a metal attachment because I’m developing a simulator where users train with real metal tweezers. Therefore, I need the sensor to detect the touch of a tool that behaves exactly like the one they’ll use, but I could create my own tool; What material would you recommend—something similar to a capacitive touchscreen stylus?
Good start.
You can make the touch area much smaller and the ground plane much larger.
I once used a piece of virgin single-sided circuit board and made small touch field islands on it with a dremel tool. I then used the board upside-down, so no glass was needed.
Leo..
I’m building a prototype and using an MPR121 (12 channels) to detect proximity/touch of metal objects held in my bare hand. To protect the electronics and diffuse an LED, I cover the electrodes with a thin sheet of TPU hydrogel (basically a phone-screen protector).
, but when I bring the metal tool close it doesn’t register at all even after connecting the tool to my hand to “boost” the capacitance.
