SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By brennen
#48163
So I've tried to figure out how to do this, but I'm coming up blank. I'm working on a hand-held device that will have a coin-cell lithium battery on board. The battery will likely be visible, and I was wondering if it's possible to turn that into a touch sensor. Given the fact that every article and app note on capacitive sensing doesn't have an active source hooked up to the sensing circuitry, I haven't had much luck with internet searches.

I thought that maybe putting a DC-blocking capacitor in series with the battery might allow me to make some sort of brief measurement at the instant that somebody touched the sensor by using some sort of capacitive sensing. Any ideas?
By saipan59
#48172
I'm assuming that the device will be running from the battery.
If so, I think making a touch-sensor out of it will be difficult, because the 'touch-signal' will be common-mode with your power supply.
Or, the battery will be a relatively low impedance, and you would really like to have the touch sensor input be high impedance.

Pete
User avatar
By leon_heller
#48175
Microchip has application notes on using standard PICs for touch sensor applications. If you need an MCU you might as well do it that way. You need a couple of metal pads covered with plastic, IIRC.

Leon
By brennen
#48178
leon_heller wrote:Microchip has application notes on using standard PICs for touch sensor applications. If you need an MCU you might as well do it that way. You need a couple of metal pads covered with plastic, IIRC.
I've read their app notes on creating capacitive sensors, but unfortunately this seems to be a special case. The way they did it, I would end up shorting two voltage sources together, which is not a good thing to do. The only way to use their app notes is to isolate the supply voltage of the battery, but still somehow read the capacitance across it (which is why I had thought of maybe putting a DC-blocking cap in series with the battery).

My goal is to actually use the case of the battery as the touch sensor, instead of creating pads on the PCB. If I can't get the former to work, though, I will move on with the latter, since it's considerably easier to figure out.
By saipan59
#48182
I just thought of an idea that has a "chance" of working:

Put an inductor in series with each side of the battery, then connect a capacitor from the exposed side of the battery to the touch circuit (whatever that may be).

In theory, the inductors allow the battery to supply DC power to the circuit normally. Yet for AC signals (such as the touch circuit), the battery appears to be connected to 'nothing'.

Let me know if this scheme actually works... ;-)

Pete
By wiml
#48399
saipan59's idea is what occurred to me also. I can't decide if I think it'd work or not. For any capacitive sensor you need two electrodes anyway: normally the "other" electrode is the system ground (both sides of the battery are at AC ground). You'll still need something to be at the other end of the e-field that the user is disturbing.
By saipan59
#48438
You'll still need something to be at the other end of the e-field that the user is disturbing.
The answer there would be to provide capacitive coupling from the touch-sensor circuit to the battery. I wouldn't think that the touch-sensing would be ground-referenced per-se. So I think my original suggestion is likely all it takes (at least on paper).

Pete
By khearn
#48458
saipan59 wrote:I just thought of an idea that has a "chance" of working:

Put an inductor in series with each side of the battery, then connect a capacitor from the exposed side of the battery to the touch circuit (whatever that may be).

In theory, the inductors allow the battery to supply DC power to the circuit normally. Yet for AC signals (such as the touch circuit), the battery appears to be connected to 'nothing'.

Let me know if this scheme actually works... ;-)

Pete
Wouldn't this have problems if your draw from the battery isn't fairly constant? If you've got digital circuitry it's going to have sudden changes in current demand from the battery, and the inductors would impede that. I'm not sure how much of a problem it might be, but it's something to keep in mind.

Keith
By ylim
#48466
saipan59 wrote:
You'll still need something to be at the other end of the e-field that the user is disturbing.
The answer there would be to provide capacitive coupling from the touch-sensor circuit to the battery. I wouldn't think that the touch-sensing would be ground-referenced per-se. So I think my original suggestion is likely all it takes (at least on paper).

Pete
There are capacitive sensing transducers in the market.
By saipan59
#48477
Wouldn't this have problems if your draw from the battery isn't fairly constant?
Bypass caps on the non-battery-side of the inductors will take care of it.

Pete
By brennen
#48610
Thanks for all your replies. I talked to someone in Quantum Technology's tech support (the company that makes the QProx chips), and this is what they said:
Quantum Technologies Tech Support Guy wrote:Hi Brennen,
Thank you for taking an interest in our products.

The electrode can be made of any conductive material but it must not have a potential on it. It needs to be a conductor just connected to the sensor. So I’m afraid your idea will not work.

Best Regards,

Wei Jiun
I was actually wanting to use the QProx chips because the uC I'm working with is really small and doesn't have very many I/O pins. I may still try the circuit that Microchip uses in their docs but have the supporting hardware external to the microcontroller, just to see if i can get anything going.
#141802
May i advise you to use microcontrollers with integrated capacitance touch sensing module. I used Microchip mTouch and it is working flawless.
i was able to detect finger press through 10mm glass easy..
:)
if you want to work with it i made few tutorials as well as one complete project with source code provided.
you can see more info at these links.
here is generally about capacitance measurement with microcontroller
http://www.electronics-base.com/index.p ... controller

but i recommend you to uce CSM module because it has two current sources and there for gives better readout results.
here is info on mTouch
http://www.electronics-base.com/index.p ... controller

and complete project explained with C code so you can start your hardware in seconds

http://www.electronics-base.com/index.p ... 07-example

good luck