Capacitive touch trigger as an isolated system

[esahakian]

I need a way to trigger a capacitive touch screen device (like a tablet) using an isolated system without touching it. For example a kid places a small gadget on a tablet screen and the gadget will trigger a capacitive touch on the tablet without requiring anyone to be holding the gadget or the gadget being tethered to anything. So as long as the gadget is sitting on the tablet screen it is continuously emitting a touch trigger to the tablet. Basically a stylus that is an isolated (independent) system.

Do anyone know of such a circuit using an Arduino or something else? Or perhaps you can point me to the right resource for building such a circuit?

[Marco]

AFAIK small objects are usually invisible to the capacitive touch screen even when conductive, unless they have a big old capacitor to the environment touching them (ie. you). In theory a capacitive touch screen could measure the mutual capacitance between multiple rows/columns instead of just individual intersections to detect conductive objects but I don't think they do.

With an active device you could probably "amplify" the capacitance change the touch screen detects to reliably trigger it even with a small object.

The simplest experiment  would be to simply have a gadget with a bottom and top electrode and generate a high frequency high voltage oscillating voltage between them (maybe at two separate frequencies to prevent an accidental synchronization with the device's measurement frequency). You need to correct for the much smaller parasitic capacitance of the gadget relative to the human body by using a higher voltage, to generate displacement currents of the same order. Obviously you need to be careful with the insulation and the amount of power the device can deliver and how much energy is stored. This approach might confuse the automatic calibration algorithms in these devices though.

[esahakian]

Thanks, would you recommend using an inductor or a capacitor to generate the field? Someone else told me that maybe an oscillating field using an unshielded inductor might in theory trick the device and trigger the touch...

[Marco]

The simplest experiment  would be to simply have a gadget with a bottom and top electrode and generate a high frequency high voltage oscillating voltage between them

So obviously a capacitor ... because that's what two parallel electrodes are. When I say high I mean 100s of volt BTW.

A magnetic field will do bugger all to a touch screen. What the digitizer is doing is putting an excitation waveform on a column wire and seeing how much displacement current it causes on a row wire, when your finger is near it that changes. What the gadget I describe would be doing is forcing its own displacement current into the wires, at it's own frequency, in the hope of causing the device to sense an object. Which might work, or might completely screw up its automatic calibration algorithms.

Sensing the excitation itself and amplifying the effects to correct for the much lower parasitic capacitance of the gadget relative to the human body might be possible too, but would make for a significantly more sophisticated circuit.

PS. this is not an easy problem, which is why no one has solved it yet. The fiducials which just detect the patterned electrode on the bottom while you are touching might not exactly meet your requirements, but they are a lot easier to make.

[Marco]

One option might be to combine a touch screen with an IR touch frame, use the fiducial pattern method to detect what exactly is on the screen, use the IR frame to detect whether it was picked up or they just let go.

Another option would be to put one of the electrodes on the gadget on a spring, that way the electrode with a spring would remain in contact with the screen for some extra time when it's taken off. If the refresh rate of the digitizer is high enough that would allow you to detect it being picked up.