Cap sense advice please!

[GreyArea]

You don't.Yes read reply #42 again and again until you understand.

Well, multiple reads not helping. I see Tony's diagram and big fat finger is laid across all three plates, including those connected to "driving noise output".

[Grumpy_Mike]

No it is not a fat finger, it is a capacitive coupling path. Your body is the Antenna picking up the disturbance signal.

[GreyArea]

Regardless of what it IS, it's shown touching all the contacts. Is Tony's diagram wrong?

I think I get your point though...the body acts as antenna picking up the fluctuations in signals...eg, 50Hz mains power. Tony's method creates a signal that means it could work even in the absence of mains power. This is why you suggested driving a pulse wave on that channel...a signal with higher amplitude creates more disturbance.

Am I close?

[Grumpy_Mike]

Am I close?

Yes spot on I would say.
The only thing different is that I suggested upping the PWM frequency to about 40KHz to make it radiate better.

[GreyArea]

Right. Back to this again

Happy to take advice from anyone, but particularly hope TonyWilk will contribute what he learned from his trials.

But first, start with what I've got.
1. Acrylic pole approx 30mm diameter, 1m long.
2. Layer of vinyl with design cut into it wrapped around 1. Above. Design is about 70 cm long.
3. Copper mesh 1m Wide, single wrap about 12-15 cm, cast onto pole in clear epoxy resin over the top of the vinyl design

For those who have read backi switched to the mesh becaus3 the copper foil is sold and blocks light. The meshallows light to passwhich mean when puttin LEDs in the tube, the design is still visible.

I know I need to make two triangul plates from the meshplan. I intend to do this by cutting them in-situ through the epoxy in which they are bonded. This should avoid any movement and allow good accuracy and hopefully the cuts will be filled back in with a last wrap of 80gsm fibreglass weave in a polyester resin.

My question surrounds your "driver" line.

My copper mesh is a single wraparound on the pole. I intend to use tape to mark a single turn "helter skelter" down the pole, which when cut should provide the two triangles. There will be two gaps between the triangles...a straight one where the opposite ends of the wrap of copper meet and the cut "helter skelter" cut

But...where to put the driver line? As I see it I have three options;

1. Inside the pole
2. Outside the pole as a single wire set into the resin in the original straight join
3. Outside the pole as a single wire set into the resin in the helter skelter join.

Which do you think would be best?

I've got some beefy 10M Ohm resistors to try and cut through the resin and gf that will be between the finger/hand and the contact plate.

I'm going to post a video of the lights as they are now. I've a finishing process that renders the design invisible when the LEDs are off...it just looks like a plain black pole. The effect I'm looking for is to sweep a hand over it and have it "magically" appear. From a programming point of view all I have to do is relate the input of the capacitive arrangement to the number of LEDs active in the strip. Ideally I'd like the "resolution" of that relationship to be +/-1...ie one LED on or off at a time...but if it has to be in chunks of 4 or 5, so be it.

I've probably missed out something critical...I'm sure someone will let me know...

[GreyArea]

Video as promised;

"All that is gold does not glitter"

[GreyArea]

Also, would it be an advantage to set up the capacitive sensing routine as an interrupt? I've only just started reading about them, but it seems it might have advantages? Thanks!