projected capacitive sensors, theory and design.(Touchscreen explained)
ฝัง
- เผยแพร่เมื่อ 31 มี.ค. 2014
- This video explains the physics behind the electrodes of the projected capacitive sensors. The typical touch-signal int the projected sensors starts with a initial capacitance, dips and the shoots up as the finger moves towards the electrodes. This unique response characteristic is counter-intuitive and cannot be explained by circuit theory alone. This video shows that the signal comprises 3 components, caused by the cross-component influence, the parallel plate capacitance and the initial coupling. Once the physics is understood, the design method naturally follow.
- วิทยาศาสตร์และเทคโนโลยี
Fantastic video! You always explain tricky concepts so that they are easy to understand, thanks!
Absolutely clear explanation.... I liked it a lot to the point I kept replaying a couple of times..... Great
Once again, a fantastic explanation!
Very help full.Thank you for best explanation.
I like your videos. You are extremely clear, precise, and I understand every word you spoke. Thumbs up!
Thank you for the thumbs up. It is a bonus for what I enjoy doing.
Pebble Soup hi Pebbles,
I have a project and I want to develop a sensor for my study. Do you have a website?
Once again, a very helpful video with great explanations. I hope you continue to create more videos, possibly on other topics besides capacitive sensors. Perhaps some simple circuits utilizing a home-made capacitive sensor.
very nice explenation! thanks for this clear information :)
Thums up! very professional
Thank you for liking it
very informative
Great video and thanks for clarifying the conventional way of interpreting the cap sense. Please help me understand how shield electrode can reduce self capacitnce. Is it because of the reduction in amount of charges between sensor electrode and shield electrode, since they are equal in phase and amplitude.
thanks about your information
Thanks for your illustrative video. Can you share the calculations and your software which has graphic outputs. I want to use them to lecture.
Very good video
thank you for watching and liking it.
Wow I can't thank you enough!
Thank you for enjoying the video
nice explanation! May I ask what do I need to have a demonstration of track panel as in your video (6'31'')? Thanks!
+Bonn Zhu This item can be found in: www.azoteq.com/design/eval_kits/3-products/products/115-products-ev-kits-iqs550ev03.html
hi sir,.. how were you able to project the output of the Touch Panel to the computer?
Thank you for the great content! I would be grateful for your insights on some of my questions:
- is your virtual ground at the same potential as the earth ground?
- In a numerical simulation where there is just the sensor electrodes+a piece of dielectric and no ground, the mutual capacitance should increase right (how is the mutual capacitance defined here)? Because in your earlier video "Capacitive sensor, theory and design", you mention that the mutual capacitance increases due to the polarization from the approaching dielectric medium, which draws further charges from the +V and -V sources; but in this video, you mention that the mutual capacitance decreases. What are the differences between these two cases?
- Lastly, assuming that I have the information of the charges, how can I obtain the mutual capacitance numerically? Is it correct to find the energy enclosed in the simulation space (E = 1/2 integral of D dot E dV) and do C=2W/V^2? If so, what about doing this energy calculation with a lossy dielectric with both real and imaginary part of the complex permittivity?
Thank you so much for all your wonderful content, and I would be really grateful any comments at all! Wishing you a nice day!
The working principal of capacitive sensor does not need the earth ground.
In other words, the circuit ground and the object (finger+body) are sufficient to influence the charges of the electrode(s), so the simulation does not take the earth ground into consideration. This is a important point, because capacitive sensor works near the earth surface, or in the mid air.
If a circuit has zero or negligible ground area, then the self-capacitive sensing will not work, but the projected capacitive sensor will work like a self-capacitive sensor. Self-capacitive sensors need 2 electrodes whereas projected capacitive sensor need 3 electrodes.
In the self-capacitive sensor, the transmitter and the ground plane (total two electrodes) form the transducer to detect the object (human body + fingertip), whereas in the projected capacitive sensor, the transmitter, the receiver and the ground plane (total three electrodes) form the transducer.
In the self-capacitive sensor, when the finger-tip moves to the touch-electrode, the charges on the touch electrode increases because the electric field polarize the human body, and the charges on the fingertip pull more charges from the source to the electrode. The increase of charges on the touch electrode is regarded as the increase of self-capacitance in the industry. In precise scientific sense, the increase of the charges on the electrode is mainly due to the increase of MUTUAL capacitance between the electrode and the fingertip, but this complicate the matters and I haven’t seen anyone talk elaborate on the misuse of the term SELF-capacitive sensor.
In the projected capacitive sensor, the electric field is projected from the transmitter electrode (1), reflected by the fingertip (3) and bounce to the receiver electrode (2) to reduces the charges on the receiver electrode. The analysis cannot be done by lump capacitor, because lump capacitor confines the electric field within the capacitor and will not influences the nearby object. The reduction of charges on the receiver electrode is regarded as the reduction of the mutual capacitance C21. In precise scientific sense, the charges on the receiver electrode is due to C21, C22, C23, and as the finger moves and the multi-body capacitor changes shape, C21, c22 and C23 combine to influence the charges on the electrode. C21 can be negative, and when the fingertip is very near the receiver electrode, C23 can overwrite C21 to give some unexpected turn-around response.
Capacitive sensing IC has no idea what happens outside the IC, so it doesn’t know the charges and voltage on the external object (human body and fingertip). It simply amplifies the charges on the electrode to get the trip, and regard the charges proportional to self- or mutual capacitance. As mentioned above, this is not rigorous academic definition. If you want to follow this practical but somewhat incorrect definition, then from the numerical simulation, just take the charges on the electrode and divided by the excitation voltage, you get the capacitance and the tripping mechanism like actual IC.
In my simulation engine (method of moment), I was able to control or force the voltage on each body (this is a multi-body capacitor problem). To calculate the mutual capacitance, I make the transmitter voltage = 1, and force the object voltage to 0, then charges on the receiver is exactly the mutual capacitance. In reality, some non-zero voltage is induced on the object, and this voltage contributes charges to the receiver electrode as well, so the charges on the receiver electrode is not simply related to one particular mutual capacitance.
@@SiliconSoup Wow, thank you for amazingly clear response! Really appreciate it!!! All the best 😀
Ciao, ottimo video... posso chiederti se è possibile collegare 4 tasti touch in serie per poter accendere e spegnere un motore passo passo da varie postazioni. Grazie per i video che posti... complimenti!!! ciao
I found your videos very interesting. Which numerical simulator is used for the sensor simulations?
Matlab, method of moment, my own coding.
How to optimize the electrode size if we want to measure the capacitance and impedance for a moist body (e.g., wood)? The capacitance and impedance should be higher or lower? which capacitance and impedance value (high or low) will provide the best moisture information. Thanks.
I think the self-capacitive sening method is better than the projected capacitive sensing method for measuring the material property.
The capacitive sensing ICs I used as examples in my videos are for user-interface purpose. If you want to measure material property, then the "capacitance meter" must be "calibrated". Most user-interface capacitive sensor has tracking mechanism to detect delta only, and therefore no calibration.
The self-capacitor sensor measures the capacitance between the sense-pad and the ground plane. The object (conductor) between the plates short-circuit the space and shorten the gap, hence increases the capacitance. The moisture also short circuit the space and increase the capacitance. There should be correlation between the capacitance and the moisture.
Best explanation for capacitive screen I've found!
Could you explain a little bit why the decrease of negative charges on R plate leads to a drop of mutual capacitance?
Charges on Tx, Rx and the fingertip are all proportional to the excitation voltage V_Tx, so it makes sense to talk about the charges on a per-voltage basis, e.g. Q_Rx / V_Tx, which is the definition of capacitance. The charges on one conductor due to 1V excitation on another conductor (and strictly no influence from the rest of the conductors, i.e.g 3rd or 4th conductors) is known as the mutual capacitance. Influence from the rest of the conductors can be eliminated by grounding them, so this is done when calculating or measuring mutual capacitance. In the matrix-type electrodes, Q_Rx/V_Tx is measured when other rows and columns are grounded, so this is almost like the mutual capacitance. (almost but not exactly, because the human body, the biggest conductor in this multi-body capacitive system, is not grounded during the measurment).
Thanks for your reply~
I'm a little confused with the definition of mutual capacitance here.
Are u suggesting that mutual capacitance btw Tx and Rx plates is almost determined by Q_Rx/V_Tx ?
If it is, then what does Q_Tx/V_Tx stand for? is it another mutual capacitance of reverse direction?
Precisely speaking, when all electrodes and human body are grounded, Q_Rx/V_Tx is exactly the mutual capacitance and Q_Tx/V_Tx is exactly the self-capacitance. But capacitive sensors do measurement with the human body floating, so Q_Rx/V_Tx is almost, but not exactly equal to mutual capacitance, and Q_Tx/V_Tx is almost but not exactly equal to the self-capacitance.
how does it measure capacitance in the first place
Very clear explanation! I’m wondering if there is any material you could provide for reference? I’ll be appreciated for that, thanks.
As far as I know, the explanation given by others use different models, which don't take into the account the effect of ground plane, and don't explain the turn-around phenomenon, The explanation in this video is original, so there is no similar reference article. But be assured, this model is backed up by years of experiments and numerical simulation.
Thanks for sharing the technology. From my understanding of your video, basically, the mutual capacitance decreases when touching before the so call "turning point", which is controlled by the overlay thickness. My question is: is this a common way to do it? why don't reduces the overlay thickness and detect the increase of the mutual capacitance? In another word, what's the benefit of detecting the decrease of mutual capacitance?
I checked a few implementations of projected capacitive sensors, and all of them detect the decrease in the mutual capacitance, so it is very common. Every capacitive sensor competes for sensing distance. The decrease in the mutual capacitance occurs first, followed by the turnaround. Therefore sensing decrease in the mutual capacitance gives longer distance than sensing the turnaround. In the lab, we can get > 20cm distance. Also, the decrease in the mutual capacitance in the projected capacitive sensor occur at a longer distance than the increase in the self-capacitance in the self capacitive sensor. Therefore, the projected capacitive sensor can sense further than the self-capacitive sensor.
In 2D sensing, the location is calculated from the signal profile of all the neighboring receiving electrodes.The neighboring electrodes are "far away" and the signals are in the dip, so can't use the shoot up signal to compute the xy coordinates.
Interestingly, Azoteq uses the shoot-up signal to implement an additional "snap" function superimposed on the 2D sensing. A metal dome structure, when "clicked" will move so close the the receiver that the mutual capacitance shoot up above the "long term average", and therefore the click is detected. As long as the turnaround go above the average value, there is no ambiguity.
Pebble Soup What company or companies offer capacitive touch screens with that >20cm distance? If there are none outside of the lab, then what companies offer products with the longest sensing distance? Thank you.
Jared Graham In the lab, I can setup Azoteq IC to sense my palm at 20 cm distance from the electrode. Other brands i tried can reach only a few cm. In real world environment where a moving object could induce more signal than the palm signal at 20cm, then it is not the IC, but the physics that determines whether 20cm-sensing is reliable.
dear Silicon Soup;
I found your video about the capacitive touch very interesting, it helped me a lot to understand the Technology behind it. I'm going to make a Presentation about capacitive touch technology for school (around 30 People) and curently i'm at the point of serching for visuals for an simpler explanation, so i decided to ask you, if it would be posible for you to send me the one's you used in youur Video. Of course I would name you as the source and i won't use them for anything else.
Thanks in advance.
Letsplesureone
I am glad that you find my video interesting. What material do you need and how to email to you?
Thanks for the video :D !!!
I fully enjoyed :D
I have a question... How can i detect a finger near of a metal sheet without contact? I need a on/off output (3.3 volts)
You may search for CAPPO sensor, sensing the change of capacitance due to the micro-depression of metal plate.
@@SiliconSoup Thanks a lot :D
I'll search about :D
Thanks 🙏
Your information is very useful. But out in the world and into the real touch screen phones, only samsung high end phones provide such features. They are also being very selfish, because they have not even released the API to let developers make use of it. But samsung's secret is no more a secret , thanks to you.
If you would like to take one step further then you should connect the dots between the hardware and software parts and make some sort of API, so that developers can finally make some working apps using the air view feature in samsung phones.
And I'm an eager s4 user waiting for third party apps to use that feature.
Thanks once again
can you please make a video on ttp223?
Ttp223 is a self-capacitive sensor, and my other video on self capacitive sensor is more relevant. th-cam.com/video/QItuf6lNvmI/w-d-xo.html
Hi Pebble, I saw your video & I feel it's great. Pls help me, can i using this IC for my application "water Level"?
Thank you so much.
+Thuong Pham Ngoc Capacitive sensor can be used for detecting water level. Most people use self-capacitive senor (see my other video) to do it, because it is most straight forward. The projected capacitive sensor potentially has some advantages, but you must be aware that if the electrode is not designed properly, the turn-around (in this video) will work against you.
Thank you so much for your help. can you let me know, what's the IC I can use it to my application.
Can you help me with the false triggering of touch in my product? If yes, then I can share the PCB files with you.
Is your design for a portable device or appliances? As portable devices have limited ground plane area, it is important to make sure the touch pad area is much smaller than the ground plane area.
Could you please tell me what tool you are using to measure the self and mutual cap?
+Chethan_D I used IQS253 eval-kit and CT210 (www.azoteq.com) to measure the self and mutual capacitance. The tool (CT210) is connected to the target board to read the capacitance and display the data on the PC. This typical connection also connects the computer ground plane to the target board so the overall ground plane is bigger than the actual target board ground plane. To avoid the effect of PC ground plane, the tool must be connected via isolated interface (the device in 2:35 is an isolated interface) rather than direct connection.
Hi,
Thanks for the info. Sorry for not being clear earlier. I am interested in the multi-physics tool that you are using to simulate the sensor. Could you please tell me what tool u have used here?
By the way, i read your article on self-cap and mutual-cap in EDN, they were really useful.
+Chethan_D I used Matlab to run a MoM(method of moment) code written by myself. Of course I tested it with known cases to verify the accuracy of the code. The algorithm (or the equations) for two plate capacitor of equal size can be found in many academic articles. For multi-body of unequal size, the equation can be quite tricky though.
Pebble Soup Thanks a lot for the info!!!
Pebble Soup hi may I have your article for self-cap please? Many thanks
how to make "inverted" touch screen?
you stick this to a screen and simulate finger touch.
блин, без перевода тяжело(