CTMU Capacitive touch – 2900 days old

[hadv215]

@Les: saw that one too and seen it more often than I like. Made me think of the Option Base directive in VB

@pic-ignorant: Not yet, I'm still experimenting with settings and thresholds.
First I'll have to get an idea of what the initial value for Timer1H/L is and what values are obtained when putting my finger in the neighbourhood.
First tests yesterday showed that approaching the pads from the other side of the pcb (1.5 mm epoxy) did not work for all 8 pads.
3mm plexiglas proofed to be an insulator, so I have to start fiddling with the values.
Will report when I know more

Regards
Harm

[top204]

That's where the SizeOf directive comes in handy. For example:

Dim MyArray[11] As Word
Dim Index As Byte
    
For Index = 0 To SizeOf(MyArray) - 1
    
Next

Or better still:

Dim MyArray[11] As Word
Dim Index As Byte
    
Index = 0
Repeat
    
    Inc Index
Until Index >= SizeOf(MyArray)

[hadv215]

Of course, seems that sometimes I forget the features that may come in handy. On the other hand, being a Delphi programmer, I'm used to using 0-based arrays so I'm not even considering this feature.
Constants and Symbols work miracles too.


I did some experimenting, using the UART to display the values.
When I changed the threshold from the original program ("TOUCHBUTTON[LOOPIE] = TMR1TOTAL + 400 ;BEST GUESS 1/2 VALUE OF DROP")
from 400 to 100 it also worked with 3 mm plexiglass.

Keep in mind that the physical design will definitively have impact on the values obtained from the Capacitive Oscillator.

My test PCB is 160 by 65 mm. The pads are circles with a diameter of 14 mm, 5mm space between them and a groundline of .76 mm between them.
MicroChip® has a tool to communicate the values read between the uC and a PC. It uses the UART and PicKit.
That's the next I'm going to try.

[hadv215]

I thought I forgot about SizeOf(), but it's some different: it is not documented but it is a protected compiler word

Regards
Harm

[top204]

Sorry. I forgot to add it to the manual. It is documented in the Whats New file and I made a note to add it to the manual, but never got round to it. I'll make another note.

[pic-ignorant]

Hi Harm,
I would be interested to see what results you get with the 3mm acrylic. I made a keypad with key scanning, which worked fine with a thin covering, but I had varying results with 3mm. I also needed some of the buttons to latch on. In the end, I used an Atmel solution which works very well

John

[hadv215]

Hi John,

what kind of results do you want to see
measurement values?
a video?

latching is, of course, no more than a software thingy

Regards,
Harm

[pic-ignorant]

Hi Harm,

I was more interested in what material types and thicknesses that you were able to get reliable repeatable results from.

[hadv215]

It's called 'plexiglas' in Dutch, a more formal name is Polymethylmethacrylaat.
Both 2 and 3 mm transparent give good results and even 10 mm works.

[pic-ignorant]

Thats very good, better than I achieved

[hadv215]

That's as far as it will go.

Question: at the moment my uC and pads are on the same PCB (just for experimenting purposes) and I find that touching the traces between the pads and the uC results in a response.
I'm considering to have the pads on a separate PCB, so make it two PCB's, to reduce this unwanted behaviour. Then I'll have to wire the PCB's and this might cause more problems than it solves like interference between wires.
Should I shield each separate wire?

Regards,
Harm

[pic-ignorant]

I made my board in 2 parts as you suggested with pads and leds on top board.(illuminated pads) and a uc board underneath with pic16f723 & relay outputs on the lower board and connected them with 2.54mm pin header and sockets. The top board was double sided with the pads on top and traces below.