Capacitive Sensing & iPhone Touch Screen

[cepheid]

I'm trying to understand why certain objects work (i.e. register a touch) on an iPhone screen, but others don't. From what I've read, the screen consists of two transparent conductive layers separated by an insulating layer. Furthermore, each layer is etched to form parallel electrodes, and the two layers are etched in orthogonal directions in order to create an x-y coordinate grid of capacitors. One layer has a voltage applied to it, and if the capacitance at any x-y point changes, the voltage at that point on the other layer will change from its nominal value. This is my understanding. Please let me know if I am in error.

Based on this scenario, it seems to be that touching *any* conductor to the screen should work to change the capacitance between the two layers at that point. However, some things work and some don't. Reading online how to make your own "stylus" for the iPhone, I came across the suggestions of wrapping a conveniently sized object in aluminum foil and also trying a battery. These suggestions both seem to work, but others don't, like the end of a mini jack plug, a penny, and the metal tip of a mechanical pencil. I *think* that all of these last three should be conducting. So the experimental evidence seems to point to the idea that you actually have to apply an electric field to the screen. I assume that doing so helps, for the same reason why fingers work so well, and this could explain the battery, and the foil, which is thin and was in contact with my fingers directly. But based on what I've read about how this actually works, I cannot understand why it is *necessary* to do so and won't work at all otherwise.

Another weird thing: why does only the negative terminal of a battery work, and not the positive?

EDIT: In a test app (a doodling program), the penny works (draws stuff) when laid flat on the screen and moved around, but not when making contact with one of its edges. ?

EDIT 2: even that only works sometimes, not very reproducible.

 

[MATLABdude]

The following pages (from HowStuffWorks.com) go into how Apple's touchscreen works, and what makes it different from other capacitive touchscreens out there. Not sure if the proliferation of multitouch devices now results from adapting Apple's technology, or just better software and tracking with their existing touchscreens.
http://electronics.howstuffworks.com/iphone1.htm

As for an appropriate stylus, I found a few commercial ones while Googling for iPhone stylus. I'd presume that the Pogo one works decently if Thinkgeek is carrying it. None of these seem very high-res (so to speak), but this seems to be a limitation of the iPhone itself.
http://www.google.ca/search?q=iphone+stylus

 

[svenkesd]

Any capacitive touch screen, such as the iphone, will register a touch based on the dielectric of the material in contact and the amount of that material.

Think of it as literally inserting a material between two plates of a capacitor, a low dielectric material won't have much effect on the electric field, nor will a miniscule amount of a high dielectric (not contacting the plates). You need a certain amount of a high dielectric for the sensors to register the touch. The edge of a penny or the tip of a pencil will not register because there is not enough of the material in proximity to the sensing surface for the software to register as a "touch".

There is nothing special about the iphone touch, it utilizes existing technology, i.e. they put an off the shelf chip in there to do the job.

 

[cepheid]

Hmm, so nobody addressed my specific points (i.e. nobody indicated which of the specific things I surmised about how the iPhone touch screen works were correct and which weren't). That's okay though. I understand people are busy and don't have time to read a big post.

I can only assume that since my description of how the touch screen worked differed drastically from svenkesd, that I was mostly wrong in my thinking.

MATLABdude: thanks for replying. However, I had already read all of the relevant HowStuffWorks pages before posting. I found their descriptions and explanations to be inadequate, in the sense that they did not explain how the technology works with the level of precision I was looking for. Also, my objective is not to simply buy a commerical stylus, but to gain enough of an understanding of the physical principles behind the operation of the touch screen to build my own stylus using everyday materials.

svenkesd: I think we're getting closer with your post, but I'm still confused. Based on what's on that HowStuffWorks page, clearly you cannot actually insert anything between the "plates" (i.e. between the two conductive layers). Therefore, while your description might be a good conceptual way "to think about it" it does not explain what is really happening. I'd really like to know how this actually works. How is touching the screen (at a point that is entirely outside the capacitors) somehow equivalent to inserting a dielectric in the middle of one of them? Furthermore, I'd really like some clarification on the following points:

- is it necessary to supply an external electric field, or is it sufficient to somehow modify one that is already present?

- if what does the job is a material with a high dielectric constant, then why are people recommending constructing styluses out of conductive materials like aluminum foil? Furthermore, why does aluminum foil actually work when you touch one layer of it to the screen over a large enough surface area?

Well, this turned into another big post in spite of my best efforts...

-

 

[svenkesd]

In a sensing device like the screen the electric field is generated by the device, no external electric field is necessary to be introduced to make the screen work. There are various methods to do this that have different advantages for noise immunity, sensitivity, etc.

I'll start from the basics...

An electric field is generated between two metal plates (capacitor) when a potential is introduced between them. Locating a dielectric material within the electric field of a capacitor will change the electric field (i.e. change the capacitance of the capacitor). If the two metal plates are exactly parallel, almost all the electric field will be between the two plates, you can think of it as lines going from the high potential to the low potential. Even in a parallel plate situation, not all the electric field lines are directly between the two plates, around the edges of the plates the field lines will "bubble" out instead of traveleing straight from one plate to the next.

See example here http://son.nasa.gov/tass/content/electricity.htm

A capacitor does not have to be "parallel plates", any two conductors with a potential difference between them will have an electric field no matter what shape they are.

So...

A sensing device like that in the iphone will generate an electric field that is intentionally "bubbled out" so that something contacting the surface of the screen will be within the bubble of the electric field lines and therefore change the capacitance of the two conductors.

If something with sufficient dielectric to effect the field lines comes in contact with the screen, the electronics will "sense" a touch by literally measuring a change in the capacitance at that point on the screen.

A conductor essentially has a dielectric of infinity, and if you place the conductor between the two capacitive plates the electric field will be reduced to zero (and electricity will flow easily). The capacitive elements of the iphone screen and other touch screens however are covered by an insulator so it is impossible to contact the plates with a conductor, however a conductor will still alter the capacitance when placed within the electric field, same as another high dielectric by allowing the electric field lines to propagate more easily through it. Metal works well because in combination with an insulated coating it acts as the ultimate high dielectric while still not shorting out the plates.

It is hard to explain by typing a message but hopefully this helps.

 

[cepheid]

That post was helpful, thanks!

 

[diplodocus]

I've been looking for a good explanation of how capacitive screens work and I think I found it here! To add to the answers, and I might be wrong though I think svenkesd referred to it in the 2nd response, the iPhone software probably plays a big role in determining what pattern of capacitance changes (area covered in the grid, for example) will register as a touch. This might explain the results of cepheid's experiments with the penny - when flat it changes the capacitance over several contiguous grid locations, while when on its side it is largely ignored by the software because though it changes the capacitance it only does so in a very small location? Please let me know if my thinking is incorrect.

 

[RSBMD]

Your explanation is excellent. I'm surprised how tough it was to find this info.
Can you explain why certain objects - like an aluminum antenna - if placed on a capacitive touchscreen (like iPad) will activate a key and the key will not deactivate even if the hand holding the object is removed and the object alone is left on the screen. Other "stylus" material will not behave this way.
Thanks is advance.

In a sensing device like the screen the electric field is generated by the device, no external electric field is necessary to be introduced to make the screen work. There are various methods to do this that have different advantages for noise immunity, sensitivity, etc.

I'll start from the basics...

An electric field is generated between two metal plates (capacitor) when a potential is introduced between them. Locating a dielectric material within the electric field of a capacitor will change the electric field (i.e. change the capacitance of the capacitor). If the two metal plates are exactly parallel, almost all the electric field will be between the two plates, you can think of it as lines going from the high potential to the low potential. Even in a parallel plate situation, not all the electric field lines are directly between the two plates, around the edges of the plates the field lines will "bubble" out instead of traveleing straight from one plate to the next.

See example here http://son.nasa.gov/tass/content/electricity.htm

A capacitor does not have to be "parallel plates", any two conductors with a potential difference between them will have an electric field no matter what shape they are.

So...

A sensing device like that in the iphone will generate an electric field that is intentionally "bubbled out" so that something contacting the surface of the screen will be within the bubble of the electric field lines and therefore change the capacitance of the two conductors.

If something with sufficient dielectric to effect the field lines comes in contact with the screen, the electronics will "sense" a touch by literally measuring a change in the capacitance at that point on the screen.

A conductor essentially has a dielectric of infinity, and if you place the conductor between the two capacitive plates the electric field will be reduced to zero (and electricity will flow easily). The capacitive elements of the iphone screen and other touch screens however are covered by an insulator so it is impossible to contact the plates with a conductor, however a conductor will still alter the capacitance when placed within the electric field, same as another high dielectric by allowing the electric field lines to propagate more easily through it. Metal works well because in combination with an insulated coating it acts as the ultimate high dielectric while still not shorting out the plates.

It is hard to explain by typing a message but hopefully this helps.

 

[svenkesd]

It is hard to say why that would happen, so many factors come into play that it is hard to guess without experimentation.

One possibility:

There are different thresholds that capacitive sensing software will watch for when detecting objects, first a threshold to tell the device that an object is present, then a threshold in the opposite direction to tell the device the object is no longer present. Maybe the aluminum in a person's hand is enough to trigger the sensor when the aluminum alone is not, but the aluminum alone presents enough capacitance to keep the sensor triggered even with nobody touching it.

Another possibility:

A build-up of charge can effect certain methods of capacitive sensing, maybe enough charge is built up in the aluminum when a person touches it to keep the sensor triggered. Try grounding the aluminum after placing on the touch screen to see if that makes a difference. (must be same ground reference as device)

 

[funtongue]

Another weird thing: why does only the negative terminal of a battery work, and not the positive?

I experimented using a AA battery, and the negative terminal only activated the touchscreen while I was holding the battery. If I let the battery stand on the screen by itself, no activation. If you remove the wrapper from a battery, you'll notice the entire body of the battery is the cathode, and only the very top is the positive terminal. In other words, the negative terminal of the battery works because it's in contact with your hand! ;-)

 

[Elvislaur]

Hi, I'm trying to find out are the capacitors in these touchscreens are in series or in parallel, or maybe in mutual capacitive touch screens they're in series and in self capacitande-in parallel. I'm very confused because they're asking me what would happen if they were in parralel or in series?

 

[Claref]

The following pages (from HowStuffWorks.com) go into how Apple's touchscreen works, and what makes it different from other capacitive touchscreens out there. Not sure if the proliferation of multitouch devices now results from adapting Apple's technology, or just better software and tracking with their existing touchscreens.
http://www.tmart.com/iPhone-Stylus-Pens/

As for an appropriate stylus, I found a few commercial ones while Googling for iPhone stylus. I'd presume that the Pogo one works decently if Thinkgeek is carrying it. None of these seem very high-res (so to speak), but this seems to be a limitation of the iPhone itself.
http://www.google.ca/search?q=iphone+stylus

With this framework, the paper designed and implemented a multimodal wireless switch system, which supported pen and speech input. this is even hope. now iphone stylus pen was worked out and its very popular now, because not only convenient and quick and protect the screen.

 

[hassanshirazi]

I am not an expert on this but ..
for any circuit, capacitance in this case, to work .. current needs too flow through it .. so for the current to flow, there has to be a potential difference ..
when you touch the screen with your finger some current flows through your finger to the ground and makes a parralell capacitive effect in that area of the screen .. this lower the capacitance of that part of the screen and hence is detected by the circuitry and software .. the current through finger flows because your body is grounded .. the negative terminal of the battery should only work if you are holding the positive .. and about the part that why only negative terminal works, it might be because the potential on the screen is positive ..
the radio antennae works as it is continuously oscillating its voltage levels .. this allows the current to flow to it and as the frequency is so high .. it fells as if the button kept pressed on ..
so for a touch to be recognized the touching object must be capable of accepting current flow (needs not only be conductive but also properly connected to ground) and it needs to have sufficient area of contact with the screen to get a acceptable change in cappacitance .. by acceptable i mean the threshold level set by manufacturer..

i hope it helped ..

 

[willie001]

If something with sufficient dielectric to effect the field lines comes in contact with the screen, the electronics will "sense" a touch by literally measuring a change in the capacitance at that point on the screen.http://www.amzcard.info/g.gif

 

[conspiritech]

I have been looking for a tablet that's compatible with a pointy stylus. Although some people apparently do well with the fat capacitive styli for drawing, I am skeptical. Something like this shape would be ideal:
http://images.nittanyweb.com/scs/images/products/81/large/mehaz_mehaz_signature_cuticle_pusher_and_nail_cleaner_332_p23811.jpg

I'll have to go to the store and actually try it (fat stylus) out one of these days, I guess, but in my estimation as someone who draws regularly and wants to be able to take notes on a tablet as well, a fat stylus is simply not acceptable.

My friend however tells me resistive screens are aweful, and I think I'd have to buy an old device to get one... so the industry has failed me... the Asus eee Note might have worked, but it had problems too, apparently and is anyway obsolete now, I guess.

 

[JJinLA]

Build your own stylus with. . . .(drumroll) a Pipecleaner!
Yes, I was also looking for material with which I could fashion some kind of stylus, as my own "Mom" thumbs were simple too big for the little iPhone touch keyboard. I needed some kind of small tip accuracy, something I could use on both thumbs, since the few commercial styluses I found were pen-like. But I prefer to use my own "thumb-drives", which are faster and more convenient -- if only I could find a discrete "attachment" tip. Thimble? no. Cake decorating tips? didn't work. But wrap an inexpensive pipecleaner around my thumb (or both) with the endpoint doubled and twisted to optimum position, it does the job.

And they're available in many colors to coordinate with any outfit! Although perhaps physicists don't care that much about a color spectrum.

 

[oronymo]

This is an excellent topic! Does anyone know if you could create the same change in capacitance using magnet or electric fields? (Eg with minimal or no physical contact but instead using an electric field to initiate a registered touch response)?

If so, how might you actually test this?