Finally we come to the point. Because it is cool to touch a metallic surface (sensor) and turn on your PC for example, but it is absolutely awesome to touch the wooden surface of your modded PC case, or the dark plexiglas with the icon and blue backlit... Isn't it?
Until now, we've talked about touch sensors with metallic touch pads. The first method we explained was the resistance. We cannot do much with this method. Then we talked about the AC Hum sensors, which is not good for switching applications. Finally, we talked extensively about the capacitance touch sensors. Now let's take a closer look to this method. No matter which capacitance sense method is used, they are all based on the fact that the human body and skin can hold charge and act as capacitors. When the sensor is touched, the human body adds a parallel capacitance to an existing capacitor somewhere on the circuit.
This happens if the sensor is metallic. But what if the sensor is covered with a plastic layer, or a piece of wood? How can we transfer the body capacitance to the sensor through this layer? As a matter of fact, we do not need to do absolutely nothing. And that is the awesome part with the capacitance sensors. The key is the dielectric constant. All insulating materials have this property. The dielectric constant, is the ability of a material to store electrical energy. Some materials have bigger dielectric constant than others and can perform larger capacitors. You can find a table with some materials and its DC in this link: wikipedia.
Check out now this drawing:
This is a very simple touch sensor plate. Actually, it is a simple PCB with a circle in the middle. The diameter can begin from a 4-5 millimeters and can grow bigger. There are 3 regions: the grounding layer, which is a copper layer connected to the ground of the power supply, the sensor which is the copper circle in the middle, and an insulating layer, which is an electrically insulated region between the grounding layer and the sensor, with a material with small dielectric constant (simple air for example). Look what happens, when the grounding layer is connected to the ground and the sensor to the input of a capacitance sensor circuit:
Here you see a cross-section of the above sensor. The grounding layer and the sensor layer can be see, as well as the insulating layer in between. The blue lines shows the electric field energy which packs in this sensor. And because the only dielectric material that it has is thin air, it has an extremely low capacitance (the constant Îµ for air is 1).
Now look what happens, if the sensor is covered with a material that has high dielectric constant, like for example glass, that has Îµ from 3.7 to 10:
The energy lines are not buried under the thick layer of glass. Instead, they flow within it much easier than air! This is something that someone does not expect to experience, unless he knows it. This property is what makes the covered capacitance switches functional.
When the finger touches the glass, the energy lines goes through the finger, which adds an extra capacitance and finally changes the overall capacitance of the measuring capacitor in the circuit.
So? What should i change in a capacitance switch, so that i can cover it and still work?
In a word: only the circuit sensitivity.
And what if i increase the sensitivity TOO MUCH?
Then you have just make yourself a capacitance proximity sensor! It will be able to sense the human presence from a distance, without even touching it. Do not expect of course something like 100 meters range. It works for very short ranges. To increase the range, you can use larger sensors with bigger area.
What materials can i use as a cover?
Almost any insulator will do, but prefer those with larger dielectric constant if you want to have thick layers. If you ask me the maximum thickness, i cannot answer it. There are several parameters to determine this. I can tell you though, that i have try with success 8mm Plexi Glass, 16 mm MDF wood and 16 mm chipboard. The sensor that i used was a rectangular 2-sided PCB with 1cm sides. The front side was the sensor and the back side was the grounding layer. This is also another technique that you can use if you want to make a single button on a single PCB. If you feel like putting more sensors on the same PCB, then make sure that there are insulating layers between the sensors. Here is an example:
The sensors are marked with the letter "S". The white regions are insulating layers, and all the rest is a grounding layer. This specific design has all the wiring at the bottom side of the PCB with vias. Instead, you can reduce or completely remove the grounding layer, and place the traces on the same layer.
Here is another interesting technique. Suppose that you want to cover the keypad with a thick plexiglas. Then you need to increase the sensitivity of the buttons. But this may result in false button presses, if the operator presses a few millimeters displaced from the correct button. You can effectively direct the energy lines to the point that you want, by simply removing material from the cover, above the regions that you do not want the energy lines to go. Look for example this cross section:
What you see, is a cross section of a sensor with 3 buttons on the top side of the PCB, and a grounding layer on the bottom. Above the buttons there is a thick plexiglas cover. To prevent the energy lines to go outside of the region of each buttons, the material between them is removed. This leaves an area with air, and as we said before, the air has very low Îµ. So, the flux will not be expanded sideways too much to cause problems when pressing buttons.
Microchip has done a very good research in capacitance sensors. I strongly recommend you take a look the mTouch(R) pages. Also, do not forget to take a look at this application note, and more specifically in pages 11 through 13, in which they explain how to make paired-keys capacitance touch sensors, how to implement key matrices with capacitance sensors and how to make linear slides.
You can go even deeper by viewing the webminars that Microchip has:
@Marc Hello. The 555 solution was not very reliable so i have not post the circuit. It seems though that many people keep asking me for this circuit and i may consider to analyze it one day. Maybe.
And, i'm currently writing the BJT transistor theory. When i finish with this, the next step is the PIC theory... Stay tuned
i have encapsulated condenser capsule head in steel basket,and pcb electronics in steel tube,both hardwired to ground...and there was still no reduction in hum. the amplifier is JFET source follower class A,can a class A amplifier hum
hello sir,thankyou for your capacitance switch videos. i have a diy condenser microphone with class A JFET amplifier [only one transistor] powered by 48V phantom power from a Behringer Xenyx 802 mixer.
it hums loudly [0dbV] ,but when i grip the ground the hum reduces. i removed the source resistor off the JFET ,so that the amp was dead, and the hum was gone. so i knew it was not produced by the rest of the circuit.
i feel this hum could be produced by capacitance. the circuit in your capacitance switch video,how does one remove the capacitance?i appreciate very much any reply. i have a schematic but not an email address but mine is firstname.lastname@example.org.
I WILL TRY THIS THEN REPLY U. BUT I REQUIRE HEX CODE FOR PIC MICROCONTROLLER FOR ALL 16 INPUTS. ALSO PL EXPLAIN HOW MAKE 64 INPUTS THROUGH 74HCT573. & HEX CODE PLEASE TO MY MAIL ID: email@example.com;
I was building a circuit with the capacitive sensing. I build a touch pad by connecting a wire out of a circuit to a aluminum foil. And the sensor works by directly touch the aluminum foil using your finger.
But then the TA told me direct touch is not really a capacitive sensor, because there's no distance between the two plates. But the circuit works. But according to you, number two method is the direct touch, and the finger is actually a capacitor itself.
@Pedro Sousa for the resistors, you need to make some tests yourself to find the proper values. You can start with RP=220 and RB=22K, which is a good selection.
As for the CMOS solution, you can select for example a quad AND gate (4081). Keep one input H and the other input like the transistor input. Bur i would still stick with the transistor method with SMD transistors
I'd like to do a panel with 50 Resistance Touch Switches for 5 V DC.
In your video clip you are using a 2N2222 transistor and in the text you say I can instead use a CMOS or other devices.
I'm looking for reliable and cheap switches.
1. With 2N2222, what will be the values for RB and RP.
2. If I use a CMOS:
- The circuit is exactly the same?
- Which CMOS should I purchase?
- What will be the values for RB and RP?
- Is it more expensive?
3. For so many buttons, should I use other solution?
@saravanaeswaran first, you do not need to spam a message in all boards asking the same thing, it is not helpful for other users. You can find similar circuits with capacitance sensors in the "circuits" section of the site:
All circuits have complete documentation, schematics and PIC code. If you ask for another variation of a circuit we do not make circuits on demand.
thanks for the diagram of the resistance touch switch. I guessed kind of close to the right way but this clears it up. It took me forever to find a diagram! Now if only the resistor values were there...
I like the capacitance touch circuit. Can this circuit work in bistable mode using relay for on/Off the appliances. If possible please mail me the part list & schemetic for the same. Thanking u..............