Question about touch switch

Thread Starter

Tekk

Joined Apr 18, 2009
16
I am trying to make a simple electronic instrument with 12 touch switches as the keys.

Basically I want each switch to insert a different resistor into the circuit to change thr frequency of a 555 oscillator. The output of this will be sent to a 386 audio amplifier.

I've read into Darlington pairs as a possibility, but I'm confused on how to wire them so that when one is pressed, the resistance is changed, and when none are pressed, no sound is made at all.

Any suggestions?
 

R!f@@

Joined Apr 2, 2009
9,918
you want to make one or you want really make one.
What are the components you have if you want to make one

What does darling-ton got to do with touch switching.

By the way, what type of touch switch do you have?
 

Thread Starter

Tekk

Joined Apr 18, 2009
16
Any components I need can be bought, I just want to do everything in the best/most stable way.

As for what Darlington pairs have to do with touch switches, I was looking at this in particular:
http://talkingelectronics.com/projects/TouchSwitch/images/touchswtr.gif

By the way, what type of touch switch do you have?
I'm looking to make them from scratch using something like the picture linked above. The 'touch contacts' would be etched into a PCB or something.

Thanks!
 

R!f@@

Joined Apr 2, 2009
9,918
Oh tht!!!!

That's quite useless actually. It probably will switch on from noise.
When the base is left open in a darlington, the bast tends to pickup noise and sometimes noise will bias the transistors in to conduction.

You better first study about capacitive touch pads.
Try googling
 

retched

Joined Dec 5, 2009
5,207
No its not from noise. It uses your skin as a conductor to complete the circuit.

If you want, there is a 555 circuit by Forrest M. Mims III that is a touch switch.

If you wanted, you could use a 555 on each button, but that is kinda overkill.

I would look into the capacitive touch material overlays for GLCDs and such.

Mr. Mims circuit is in this Engineer's Mini Notebook Volume I 'Timer, Op Amp & Optoelectronic Circuits & Projects.

I highly recommend getting your hands on Volume I and IV.

You can get them here:
http://www.forrestmims.com/engineers_mini_notebook.html
USED WITH PERMISSION
 

Attachments

Last edited:

smartweb

Joined Jun 28, 2010
11
I'm not very knowledgeable in electronics, but I was working on something similar and thought I would share what I came up with just in case it is helpful at all. I have not tested this and have no idea if it would actually work, but I'll post it for you to look at, at your own risk. If anyone who knows something wants to critique it and let us know if it would work for his purposes, that would be great. Once again, this is just a modified version of something I came up with while working on a similar project and I'm not sure if it will actually work in practice.

I just did the diagram with 3 buttons but you should be able to get the idea for more. R3, R4, and R5 are the values you will pick (ie, 1000, 10K, 100K) or something along those lines. There is a nifty calculator on the guys website whose basic astable timer diagram I included and my example is linked to. Anyone think this will work for what the original poster is asking? Good luck either way, sounds like a fun project!

The "R2" in my diagram is not actually present. This is just labeled so that you can see where the wires hook up in the astable circuit shown to the right.


I apologize if this diagram isn't correct.

Here is the astable timer calculator:
http://home.cogeco.ca/~rpaisley4/LM555.html#3
 

Attachments

Thread Starter

Tekk

Joined Apr 18, 2009
16
Thanks everyone. I will look into all of the suggestions.

smartweb: What program did you use to make that schematic? I'd like to try it out if it's any good.
 

smartweb

Joined Jun 28, 2010
11
Thanks everyone. I will look into all of the suggestions.

smartweb: What program did you use to make that schematic? I'd like to try it out if it's any good.
I actually drew it in Adobe Flash. I just turned on the grid, snap to grid, created a few graphic symbols for the different components that I could reuse, then drew in the connecting lines. I'm sure there are probably better/easier programs to do it in, but not being an electrical engineer by trade or knowing what to use, I just used what I had available.

I have attached an internal schematic of the 555 which was also found on: http://home.cogeco.ca/~rpaisley4/LM555.html#3

The one major thing I'm iffy on with my idea posted earlier is whether or not having the transistors emitter flow to pin 6 is correct. The internal schematic is a little confusing, so I wasn't really sure if the idea would work or not.
 

Attachments

John P

Joined Oct 14, 2008
2,025
If I were making a touch switch based on conduction through the user's finger, I'd have an input to a 74HC14 hex inverter with a pullup to Vcc on it--3.3M Ohm sounds right--and the conductive finger would connect the input to Gnd.

But it would be way more cool to have only one connection to the user's body, namely the switch itself. You'd need to have an oscillator running, which could be in the form of a 74HC14 with the 3.3M Ohm resistor connected from output to input, with a small (100pF?) capacitor from the input to Gnd. Then when you touched it, the frequency would go down, maybe to about half the previous frequency. Trouble is, it's much easier to tell that an input is connected or not, versus telling that a frequency has changed. It would be a whole lot easier with a microcontroller.

Capacitive touch sensors can be tricky if you have several of them running at the same time. The oscillators can influence each other, though it's possible to set up a counter which only operates them one at a time. Yet another reason to have a processor in there.
 
Last edited:

Thread Starter

Tekk

Joined Apr 18, 2009
16
If I were making a touch switch based on conduction through the user's finger, I'd have an input to a 74HC14 hex inverter with a pullup to Vcc on it--3.3M Ohm sounds right--and the conductive finger would connect the input to Gnd.

But it would be way more cool to have only one connection to the user's body, namely the switch itself. You'd need to have an oscillator running, which could be in the form of a 74HC14 with the 3.3M Ohm resistor connected from output to input, with a small (100pF?) capacitor from the input to Gnd. Then when you touched it, the frequency would go down, maybe to about half the previous frequency. Trouble is, it's much easier to tell that an input is connected or not, versus telling that a frequency has changed. It would be a whole lot easier with a microcontroller.

Capacitive touch sensors can be tricky if you have several of them running at the same time. The oscillators can influence each other, though it's possible to set up a counter which only operates them one at a time. Yet another reason to have a processor in there.
Yes, it would be way cooler if there was only one touch electrode instead of 2. :D
How about those dedicated 'capacitative touch switch' ICs? They provide all the oscillation/detection circuitry and you just connect some outside supporting components and your touch point. Are there any of these that are 'popular' or are they all relatively obscure and hard to get?

How would frequency detection with a microcontroller work?

Also, the resistive touch with 74HC14 or even transistor pairs is always a fallback option if I can't get any of this to work.
 

John P

Joined Oct 14, 2008
2,025
Re touch switch IC's, I don't know. Look it up on the Digi-Key site?

It's a safe bet that the frequency in an oscillator circuit would be quite high, so a processor would have to count pulses over a fixed interval, rather than measure the length of a single pulse. To do this, you'd have to use a processor with a pin that can trigger an internal counter; luckily, it's a common feature.
 
Top