- #1
GRASBOCK
- 1
- 0
Hello, I'm posting the first time on this forum.
I want to create a robot controlled by an Arduino which is able to make a touch signal without using a mechanic arm or finger ( here is a robot with the same goal but using an arm video). I want to take the electrical approach.
A modern touchscreen works like this: there are two electrode grids perpendicular to each other and in the middle there is a non-conducting material like in a capacitor. That way, if there is voltage between the electrodes they act like a capacitor and will get charged up because of the resulting electrical field. If a finger nears one of the electrodes it will act like another electrode and the electrons in the finger will get either pushed away or attracted to the tip of the finger making the electrical field even stronger and charging the electrodes even more. Because of the AC current of the touchscreen it can measure the difference between the electric charge of the electrode with the finger close by and the ones with no finger close by. Because of the grid and the difference to the other electrodes to the one of the finger the touchscreen can calculate where the finger is.
This is just what needs to be known to understand what I am trying to do.
So we know, that we basicly need to alter the capacitance of the electrodes in the touchscreen.
We need something with enough electrons and conductivity to make a significant change to the capacity of the electrodes of the touchscreen to make it notice a difference. I tried it with some metal for example and it worked (isolated from my hands of course). But using a very tiny piece of metal doesn't have an effect on the touchscreen (the volume matters). That made me think, that maybe I could use that to make something like a switch that allows access to enough electrons to register the flow as touch making the device think that there is a finger touching it.
I want to use an NPN transistor for the switch, because it symetrically build on from collector and emitter side allowing flow of electrons in both directions. Then using the basis i can open the access to the collector (where most of the electrons lie in; Something like my finger/ a piece of metal) while the emitter has a contact to the touchscreen.
The problem I'm witnessing is that if I come in contact the basis of my transisor while the emitter sits on the touchscreen, a touch signal is being made. That at first confused me because without supplying the basis with positive voltage, there shouldn't be any current flowing. It makes sense though, since the transistor is just a modifyable resistor. So i wanted to put enough resistance behind the basis so it won't be so easily triggered.
The real problem now is that I get random results. Sometimes a touch signal is registered with a resistance about 800 - 250 kOhm and sometimes none is registered with 0 Ohm resistance.
I am running out of Ideas what it might be.
I made sure that there is contact (fluid like water) and the area of contact is the same.
I need help finding the factors that make my results unaccurate.
I want to create a robot controlled by an Arduino which is able to make a touch signal without using a mechanic arm or finger ( here is a robot with the same goal but using an arm video). I want to take the electrical approach.
A modern touchscreen works like this: there are two electrode grids perpendicular to each other and in the middle there is a non-conducting material like in a capacitor. That way, if there is voltage between the electrodes they act like a capacitor and will get charged up because of the resulting electrical field. If a finger nears one of the electrodes it will act like another electrode and the electrons in the finger will get either pushed away or attracted to the tip of the finger making the electrical field even stronger and charging the electrodes even more. Because of the AC current of the touchscreen it can measure the difference between the electric charge of the electrode with the finger close by and the ones with no finger close by. Because of the grid and the difference to the other electrodes to the one of the finger the touchscreen can calculate where the finger is.
This is just what needs to be known to understand what I am trying to do.
So we know, that we basicly need to alter the capacitance of the electrodes in the touchscreen.
We need something with enough electrons and conductivity to make a significant change to the capacity of the electrodes of the touchscreen to make it notice a difference. I tried it with some metal for example and it worked (isolated from my hands of course). But using a very tiny piece of metal doesn't have an effect on the touchscreen (the volume matters). That made me think, that maybe I could use that to make something like a switch that allows access to enough electrons to register the flow as touch making the device think that there is a finger touching it.
I want to use an NPN transistor for the switch, because it symetrically build on from collector and emitter side allowing flow of electrons in both directions. Then using the basis i can open the access to the collector (where most of the electrons lie in; Something like my finger/ a piece of metal) while the emitter has a contact to the touchscreen.
The problem I'm witnessing is that if I come in contact the basis of my transisor while the emitter sits on the touchscreen, a touch signal is being made. That at first confused me because without supplying the basis with positive voltage, there shouldn't be any current flowing. It makes sense though, since the transistor is just a modifyable resistor. So i wanted to put enough resistance behind the basis so it won't be so easily triggered.
The real problem now is that I get random results. Sometimes a touch signal is registered with a resistance about 800 - 250 kOhm and sometimes none is registered with 0 Ohm resistance.
I am running out of Ideas what it might be.
I made sure that there is contact (fluid like water) and the area of contact is the same.
I need help finding the factors that make my results unaccurate.
