modifying a circuit design for a touch switch

07 May 2008 05:31:08

hi,
i'm trying to create a touch switch, which switches a light on and off
(toggle), with only one metal plate to be touched and runs with 3.6v
batteries.
here is a link to a circuit i've been looking at

http://ourworld.compuserve.com/homepages/Bill_Bowden/page9.htm

If you scroll down a little the heading "Relay Toggle Circuit Using a
555 Timer" appears.

How can this circuit be modified so that it works as a touch switch
with one touch-contact only?

thank you for any suggestions

makay

07 May 2008 07:07:07

donald

Re: modifying a circuit design for a touch switch

markus.kayser@friendstore.com wrote:
> hi,
> i'm trying to create a touch switch, which switches a light on and off
> (toggle), with only one metal plate to be touched and runs with 3.6v
> batteries.
> here is a link to a circuit i've been looking at
>
> http://ourworld.compuserve.com/homepages/Bill_Bowden/page9.htm
>
> If you scroll down a little the heading "Relay Toggle Circuit Using a
> 555 Timer" appears.
>
> How can this circuit be modified so that it works as a touch switch
> with one touch-contact only?
>
> thank you for any suggestions
>
> makay
>
>
>
Google knows how !!

http://www.google.com/search?q=touch+circuit+555

First hit of 791,000.

10 May 2008 05:45:18

Re: modifying a circuit design for a touch switch

On May 7, 3:07=A0pm, donald <Don...@dontdoithere.com > wrote:
> markus.kay...@friendstore.com wrote:
> > hi,
> > i'm trying to create atouchswitch, which switches a light on and off
> > (toggle), with only one metal plate to be touched and runs with 3.6v
> > batteries.
> > here is a link to a circuit i've been looking at
>
> >http://ourworld.compuserve.com/homepages/Bill_Bowden/page9.htm
>
> > If you scroll down a little the heading "Relay Toggle Circuit Using a
> > 555 Timer" appears.
>
> > How can this circuit be modified so that it works as atouchswitch
> > with onetouch-contact only?
>
> > thank you for any suggestions
>
> > makay
>
> Google knows how !!
>
> http://www.google.com/search?q=3Dtouch+circuit+555
>
> First hit of 791,000.

thanks for this link.
But what i am looking for is a touch switch that stays on longer than
4 seconds. It should just switch on when touched once and off again
when touched the second time and so on...
There also is a space restriction to the switch so it should be made
from as little parts as possible.
The link:
http://ourworld.compuserve.com/homepages/Bill_Bowden/page9.htm
shows a "Relay Toggle Circuit Using a 555 Timer". This circuit uses
just a few parts to create the toggle (leaving the right side of the
diagram aside) is there a way to make this into a one-plate touch
switch?

thanks for any suggestions

makay

10 May 2008 05:59:34

Re: modifying a circuit design for a touch switch

I would forget about the 555. It is mainly there to create delays.
Also, you don't really need a switch that stays in contact until the
next time you hit it. A simple electronic button will do the job. What
you need is a memory that remembers the last time you hit the switch
was for opening or closing the light. Because you want a simple
circuit, I would use a FLIPFLOP to make the memory. If you look on
your link, there are some circuit using the FLIPFLOP. If you wanna
know more about it, check wikipedia flipflop jk. I beleive using a
flipflop is the easiest way to do it.

10 May 2008 13:22:35

Jan Panteltje

Re: modifying a circuit design for a touch switch

On a sunny day (Sat, 10 May 2008 05:45:18 -0700 (PDT)) it happened
markus.kayser@friendstore.com wrote in
<c8d74dd2-d8e3-4fb6-b660-ef7cb4ca5b46@d77g2000hsb.googlegroups.com >:

>But what i am looking for is a touch switch that stays on longer than
>4 seconds. It should just switch on when touched once and off again
>when touched the second time and so on...
>There also is a space restriction to the switch so it should be made
>from as little parts as possible.

2 parts:

C1 C2
touch here
------------------- copper
============================================== thin PCB
------- -------- copper
| |
| |
|osc |
-------- |
out | |comp |
<----| PIC |----- |
| | |
------- R
| |
/// ///

This is a capacitive touch switch.
The advantage is that there is no physical contact with the circuit if you touch it.
The PIC runs on the internal osc.
It can be a small 8 pin PIC, one with a comparator or ADC.
The internal osc is output at 'osc', and via the capacitance C1 and C2,
between top and bottom PCB, interrupts the PIC constantly.
The interrupt routine resets a counter that is incremented in main.
When one touches the PCB top side, then the RF clock is shorted,
and the interrupts stop, the counter reaches a preset point,
and the PIC will flip 'out'.

I have not tried this cicuit, it is just a minimal parts idea.
There are of course touch switch chips, with more then one switches in
a package.

10 May 2008 10:26:27

Re: modifying a circuit design for a touch switch

On May 10, 2:22=A0pm, Jan Panteltje <pNaonStpealm...@yahoo.com > wrote:
> On a sunny day (Sat, 10 May 2008 05:45:18 -0700 (PDT)) it happened
> markus.kay...@friendstore.com wrote in
> <c8d74dd2-d8e3-4fb6-b660-ef7cb4ca5...@d77g2000hsb.googlegroups.com>:
>
> >But what i am looking for is atouchswitchthat stays on longer than
> >4 seconds. It should justswitchon when touched once and off again
> >when touched the second time and so on...
> >There also is a space restriction to theswitchso it should be made
> >from as little parts as possible.
>
> 2 parts:
>
> =A0 =A0 =A0 =A0 C1 =A0 =A0 =A0 =A0 =A0 C2 =A0 =A0 =A0 =A0
> =A0 =A0 =A0 =A0 =A0 =A0touchhere
> =A0 =A0 =A0 =A0------------------- copper
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D thin PCB
> =A0 =A0 =A0 =A0------- =A0 =A0-------- copper
> =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0|
> =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0|
> =A0 =A0 =A0 =A0 =A0|osc =A0 =A0 =A0 =A0 |
> =A0 =A0 =A0 =A0 -------- =A0 =A0 =A0| =A0 =A0 =A0
> out =A0 =A0| =A0 =A0 =A0 |comp =A0|
> =A0 <----| PIC =A0 |----- |
> =A0 =A0 =A0 =A0| =A0 =A0 =A0 | =A0 =A0 =A0|
> =A0 =A0 =A0 =A0 ------- =A0 =A0 =A0 R
> =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0 =A0 |
> =A0 =A0 =A0 =A0 =A0/// =A0 =A0 =A0 =A0 ///
>
> This is a capacitivetouchswitch.
> The advantage is that there is no physical contact with the circuit if you=
touchit.
> The PIC runs on the internal osc.
> It can be a small 8 pin PIC, one with a comparator or ADC.
> The internal osc is output at 'osc', and via the capacitance C1 and C2,
> between top and bottom PCB, interrupts the PIC constantly.
> The interrupt routine resets a counter that is incremented in main.
> When one touches the PCB top side, then the RF clock is shorted,
> and the interrupts stop, the counter reaches a preset point,
> and the PIC will flip 'out'.
>
> I have not tried this cicuit, it is just a minimal parts idea.
> There are of coursetouchswitchchips, with more then one switches in
> a package.

hi
thanks for these suggestions!
i am sorry but i'm not very experienced in electronics and have
problems understanding the diagram and the description.
if you could clarify what you mean (for stupid) that would be great!
the object which this touch switch needs to go into has got a metal
plug-socket anyway so i would not mind using this as the touch plate
but if it would work without it ...that would be fine as well.
one more important thing is that it works with three AAA rechargeable
batteries, so 3.6v.

thanks again for all contributions
m

10 May 2008 17:46:50

Jan Panteltje

Re: modifying a circuit design for a touch switch

On a sunny day (Sat, 10 May 2008 10:26:27 -0700 (PDT)) it happened
markus.kayser@friendstore.com wrote in
<d187fb6e-9e2d-411b-8077-59c1162ab0f7@27g2000hsf.googlegroups.com >:

>> >There also is a space restriction to theswitchso it should be made
>> >from as little parts as possible.
>>
>> 2 parts:
>>
>> C1 C2
>> touchhere
>> ------------------- copper
>> =========================
>====================== thin PCB
>> ------- -------- copper
>> | |
>> | |
>> |osc |
>> -------- |
>> out | |comp |
>> <----| PIC |----- |
>> | | |
>> ------- R
>> | |
>> /// ///
>>
>> This is a capacitivetouchswitch.
>> The advantage is that there is no physical contact with the circuit if you=
>touchit.
>> The PIC runs on the internal osc.
>> It can be a small 8 pin PIC, one with a comparator or ADC.
>> The internal osc is output at 'osc', and via the capacitance C1 and C2,
>> between top and bottom PCB, interrupts the PIC constantly.
>> The interrupt routine resets a counter that is incremented in main.
>> When one touches the PCB top side, then the RF clock is shorted,
>> and the interrupts stop, the counter reaches a preset point,
>> and the PIC will flip 'out'.
>>
>> I have not tried this cicuit, it is just a minimal parts idea.
>> There are of coursetouchswitchchips, with more then one switches in
>> a package.
>
>
>
>hi
>thanks for these suggestions!

You are welcome.

>i am sorry but i'm not very experienced in electronics and have
>problems understanding the diagram and the description.

Well, the 'PIC" is a Microchip micro controller.
As you perhaps have no experience with those, it would
be better to look for an other solution.

>if you could clarify what you mean (for stupid) that would be great!

Not you fault, the terms refer to programming techniques for those
micro controllers.
So you'd have to program one....

>the object which this touch switch needs to go into has got a metal
>plug-socket anyway so i would not mind using this as the touch plate
>but if it would work without it ...that would be fine as well.
>one more important thing is that it works with three AAA rechargeable
>batteries, so 3.6v.

Oh, PIC will work on 3 V upwards.

Here is a more generic description, avoiding micro controllers:

optionally
your finger
|
oscillator ---- capacitor ----- capacitor ---- detector ---- D flipflop ---- load
A

The oscillator generates a high frequency (MHz), that is normally
passed via the 2 capacitors in series to the detector, causing a voltage
at the detector.
As the capacitors are very small in value, but sufficiently big
to pass the RF, touching point 'A' short the RF via the much bigger capacitance
of your body to ground.
The detector voltage will drop, and the D flip flop will trigger, switching
the load on or off.

Up to you now to design the circuit :-)
The oscillator could be a Schmit trigger NAND gate, the D flipflop a 74HC74,
detector a diode + cap + resistor (time constant).
The capacitors can of course be small capacitors of a few pF with some kV
voltage spec, does not need to be PCB oriented.

Hope this helps.

10 May 2008 12:06:50

Re: modifying a circuit design for a touch switch

On May 10, 6:46=A0pm, Jan Panteltje <pNaonStpealm...@yahoo.com > wrote:
> On a sunny day (Sat, 10 May 2008 10:26:27 -0700 (PDT)) it happened
> markus.kay...@friendstore.com wrote in
> <d187fb6e-9e2d-411b-8077-59c1162ab...@27g2000hsf.googlegroups.com>:
>
>
>
> >> >There also is a space restriction to theswitchso it should be made
> >> >from as little parts as possible.
>
> >> 2 parts:
>
> >> =A0 =A0 =A0 =A0 C1 =A0 =A0 =A0 =A0 =A0 C2 =A0 =A0 =A0 =A0
> >> =A0 =A0 =A0 =A0 =A0 =A0touchhere
> >> =A0 =A0 =A0 =A0------------------- copper
> >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D
> >=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D thin P=
CB
> >> =A0 =A0 =A0 =A0------- =A0 =A0-------- copper
> >> =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0|
> >> =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0|
> >> =A0 =A0 =A0 =A0 =A0|osc =A0 =A0 =A0 =A0 |
> >> =A0 =A0 =A0 =A0 -------- =A0 =A0 =A0| =A0 =A0 =A0
> >> out =A0 =A0| =A0 =A0 =A0 |comp =A0|
> >> =A0 <----| PIC =A0 |----- |
> >> =A0 =A0 =A0 =A0| =A0 =A0 =A0 | =A0 =A0 =A0|
> >> =A0 =A0 =A0 =A0 ------- =A0 =A0 =A0 R
> >> =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0 =A0 |
> >> =A0 =A0 =A0 =A0 =A0/// =A0 =A0 =A0 =A0 ///
>
> >> This is a capacitivetouchswitch.
> >> The advantage is that there is no physical contact with the circuit if =
you=3D
> >touchit.
> >> The PIC runs on the internal osc.
> >> It can be a small 8 pin PIC, one with a comparator or ADC.
> >> The internal osc is output at 'osc', and via the capacitance C1 and C2,=

> >> between top and bottom PCB, interrupts the PIC constantly.
> >> The interrupt routine resets a counter that is incremented in main.
> >> When one touches the PCB top side, then the RF clock is shorted,
> >> and the interrupts stop, the counter reaches a preset point,
> >> and the PIC will flip 'out'.
>
> >> I have not tried this cicuit, it is just a minimal parts idea.
> >> There are of coursetouchswitchchips, with more then one switches in
> >> a package.
>
> >hi
> >thanks for these suggestions!
>
> You are welcome.
>
> >i am sorry but i'm not very experienced in electronics and have
> >problems understanding the diagram and the description.
>
> Well, the 'PIC" is a Microchip micro controller.
> As you perhaps have no experience with those, it would
> be better to look for an other solution.
>
> >if you could clarify what you mean (for stupid) that would be great!
>
> Not you fault, the terms refer to programming techniques for those
> micro controllers.
> So you'd have to program one....
>
> >the object which thistouchswitchneeds to go into has got a metal
> >plug-socket anyway so i would not mind using this as thetouchplate
> >but if it would work without it ...that would be fine as well.
> >one more important thing is that it works with three AAA rechargeable
> >batteries, so 3.6v.
>
> Oh, PIC will work on 3 V upwards.
>
> Here is a more generic description, avoiding micro controllers:
>
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 optionally
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 your finger
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 |
> oscillator ---- capacitor ----- capacitor ---- detector ---- D flipflop --=
-- load
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 A
>
> The oscillator generates a high frequency (MHz), that is normally
> passed via the 2 capacitors in series to the detector, causing a voltage
> at the detector.
> As the capacitors are very small in value, but sufficiently big
> to pass the RF, touching point 'A' short the RF via the much bigger capaci=
tance
> of your body to ground.
> The detector voltage will drop, and the D flip flop will trigger, switchin=
g
> the load on or off.
>
> Up to you now to design the circuit :-)
> The oscillator could be a Schmit trigger NAND gate, the D flipflop a 74HC7=
4,
> detector a diode + cap + resistor (time constant).
> The capacitors can of course be small capacitors of a few pF with some kV
> voltage spec, does not need to be PCB oriented.
>
> Hope this helps.

Great!
Thanks a lot
I'll try to work it out from here... although i'm sure i'll come up
with more questions.
Thanks again for helping me out on a sunny day like this...
m

10 May 2008 21:41:59

Frank Buss

Re: modifying a circuit design for a touch switch

Jan Panteltje wrote:

> Here is a more generic description, avoiding micro controllers:
>
> optionally
> your finger
> |
> oscillator ---- capacitor ----- capacitor ---- detector ---- D flipflop ---- load
> A
>
> The oscillator generates a high frequency (MHz), that is normally
> passed via the 2 capacitors in series to the detector, causing a voltage
> at the detector.
> As the capacitors are very small in value, but sufficiently big
> to pass the RF, touching point 'A' short the RF via the much bigger capacitance
> of your body to ground.

But your idea depends how good you are grounded? But even if you are
grounded, I don't think that the influence of the finger is sufficient for
grounding the RF. You have at least adjust the sensitivity of the detector
and adjusting it if the ambient changes (humidity, temperature).

I've found a nice description how a touch sensor could be built:

http://www.planetanalog.com/showArticle.jhtml?articleID=181401898

Take a look at the chapter "Finger Capacitance". This works with floating
fingers, too.

I've build one of these sensors:

http://www.frank-buss.de/TouchPad/

Looks like the capacitance changes about 10% when touching the sensor, so a
very simple sigma-delta ADC should work. With the integrated analog
comparator of some microcontroller you don't need any more active external
components. The microcontroller can filter ambient changes and can detect
impulses from touching the sensor.

A more expensive solution (but $3 is not that expensive) is available from
Analog Devices:

http://www.analog.com/library/analogDialogue/archives/40-10/cap_sensors.html

The AD7142 can measure 14 capacitors. With some clever software and PCB
layout, you can even implement sliders, wheels and fine resolution
touchpads.

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de,http://www.it4-systems.de

10 May 2008 20:06:29

Jan Panteltje

Re: modifying a circuit design for a touch switch

On a sunny day (Sat, 10 May 2008 21:41:59 +0200) it happened Frank Buss
<fb@frank-buss.de > wrote in <1m82xva49entc$.15sm3fw6okq6y$.dlg@40tude.net>:

>But your idea depends how good you are grounded?

No it does not.
You capacitance to the 'environment' is easily a few hundred pF, or more.

>But even if you are
>grounded, I don't think that the influence of the finger is sufficient for
>grounding the RF.

Well, I have seen this system used in TV sets.

>You have at least adjust the sensitivity of the detector
>and adjusting it if the ambient changes (humidity, temperature).

No, very wide operating range.

>I've found a nice description how a touch sensor could be built:
>
>http://www.planetanalog.com/showArticle.jhtml?articleID=181401898

Well, that is a slightly different circuit.
And 10mm thick glass is not what I was talking about.

>Take a look at the chapter "Finger Capacitance". This works with floating
>fingers, too.
>
>I've build one of these sensors:
>
>http://www.frank-buss.de/TouchPad/

Yes, different setup.

>Looks like the capacitance changes about 10% when touching the sensor, so a
>very simple sigma-delta ADC should work. With the integrated analog
>comparator of some microcontroller you don't need any more active external
>components. The microcontroller can filter ambient changes and can detect
>impulses from touching the sensor.

The circuit I am referring to uses the body - environment capacitance to _short_ out
a RF signal.
So if the series caps are for example 22 pF, then 100 pF body - environment capacitance
will significantly attenuate the output of the detector.
Anything bigger is OK too.

>A more expensive solution (but $3 is not that expensive) is available from
>Analog Devices:
>
>http://www.analog.com/library/analogDialogue/archives/40-10/cap_sensors.html
>
>The AD7142 can measure 14 capacitors. With some clever software and PCB
>layout, you can even implement sliders, wheels and fine resolution
>touchpads.

Sure.