Is capacitive sensing possible with batteries?

[jbarchuk]

Then as soon as I switched to battery power it stopped working. It took me a while to figure out why, ie. until then I'd always had a USB cable connected to the device and my PC was forming part of the system.

The difference in capacitance was easy to measure when it's on USB - it can actually detect a hand from a few inches away. Without the USB cable the difference drops to almost nothing. I had to go from 1MHz to 8Mhz clock to be able to measure it.

Is is possible that you're looking at too large a change in F. The 'measure capacitance' references I saw were talking about measuring nF and pF variations. Is it possible that the F is changing, but your circuit isn't set to a high enough sensitivity?

Check out 'butteryfly in a jar' at youtube. (EDIT: Sorry I forgot you already referenced this one.) I'm 99% convinced it's a piezo. One guy writes 'spoiler alert - see how it works.' He plays with the butterfly, opens the lid, and shows where the batteies and switch are. And that's it. He shows how it OPERATES but not how it WORKS. So that's yet another 2:43 of my life waste that I'll never get back.  

Another forum there's a buncha discussion the butterfly, capacitance, piezo, and the thread ends where someone writes 'I'll take one apart and let you know.' That was in 2011. [checking sundial [sigh]]

Given $ vs ease of design, piezo is cheap, easy and reliable.

[dlloyd]

A piece of foil in the base of the jar works.

What if the foil was replaced with transparent conductive spray?

[fungus]

Check out 'butteryfly in a jar' at youtube. I'm 99% convinced it's a piezo. One guy writes 'spoiler alert - see how it works.' He plays with the butterfly, opens the lid, and shows where the batteies and switch are. And that's it. He shows how it OPERATES but not how it WORKS. So that's yet another 2:43 of my life waste that I'll never get back. 

Another forum there's a buncha discussion the butterfly, capacitance, piezo, and the thread ends where someone writes 'I'll take one apart and let you know.' That was in 2011. [checking sundial [sigh]]

Given $ vs ease of design, piezo is cheap, easy and reliable.

Trouble is... my jar spends a lot of time with the CPU in sleep mode. You have to put your hand on it for a second or so to wake it up via capsense. It would miss a quick tap on a piezo.

[fungus]

A piece of foil in the base of the jar works.

What if the foil was replaced with transparent conductive spray?

Would probably work...

[jbarchuk]

A youtube vid titled 'Arduino capsense with battery power,' and it apparently works. http://youtu.be/mkQbzLMWFgQ

I left a comment asking what R value he used, if he added another C, and if he made any other timing adjustments.

You could also write to him directly to ask in case he never sees the comment.

[jbarchuk]

Any new progress?

Another thread referenced this tiny one MPR121 http://www.freescale.com/files/sensors/doc/data_sheet/MPR121.pdf

Fun-With-Sparks has a board https://www.sparkfun.com/products/9695

Surface mount, annoying low voltages, but might work better than scratchbuilding a cap and trying to adapt it to conventional parts.

[fungus]

Any new progress?

It's pretty much working.

This is for a workshop where people bring their own jars to decorate so I don't have any fine control over it. I think it will work with a piece of foil and 8MHz CPU clock. I can always increase the size of the resistor if it starts failing on the day.

[fungus]

...by that I mean there's definitely a measurable effect with the piece of foil on a wire for "ballast". The only challenge is to measure it.