Phototransistor matrix/strip project and questions

Forum 2005-2010 (read only) Hardware Interfacing
1

Hello everyone !

I recently built a ir phototransistor matrix/strip with this general layout for each smallest element (2*2):

This worked really well, , i even expanded to 2*3 , got analogic multiplexing working (with a 4051) etc.

Now the problem is , i am trying to build a bigger matrix (at least 5*5 for starters), and the wiring of the whole thing is a mess !

-in the current setup , i used a common emitter amplifier configuration (the only one i knew of until recently) and so for EACH phototransistor i need :

  • one 10k ohm resistor
  • one wire to vcc
  • one wire to gnd
  • one wired to the multiplexer

All of this is way to much , and way to unpractical : imagine : for a 5*5 matrix i would have to have 25 wires just for the analog out, without any way to simplify it as far as i can see...

-So i recently found this article : http://hades.mech.northwestern.edu/wiki/index.php/Photodiodes_and_Phototransistors
and after a bit of thinking and studying the possibilities i came up with this configuration: (base on the common collector amplifier configuration)

-i reasoned that since , when using multiplexing you can only access the data from the transistors sequentially, one at a time, and that the analogRead is really sufficiently fast, i could:

  • use a digital multiplexer instead of an analog one, and not READ the analog data sequentially, but POWER the phototransistors sequentially
  • everytime a phototransistor gets powered, i just have to read the analog data through a wire that is the SAME for all the transistors
  • since only one of them gets powered at the time , i would not even need a 10k resistor for each , but just one (as seen in the diagram)
  • So i would then have LESS wiring, less components needed, more clarity , and same efficiency IF i got it right of course :slight_smile:

The questions are:

  • am i completely wrong here or would this really work?
  • Would the constant on/off switching of the phototransistors shorten their life span considerably?
  • Did i overlook something?
  • Would there be a similar way to simplify the wiring but still use analog multiplexers? (i have 25 off them just begging to be used)

In case you wonder what the hell i will be using this for :
a small video i did last week end: "Ir_Multitouch Test"

A tad more info on my site , as always : www.kaosat.net

Thanks in advance for any feedback, criticism and suggestions!

2

Yes that look OK but:-
Have a look at the way I have done it here:-

http://www.thebox.myzen.co.uk/Hardware/The_Sound_Square.html

3

Thanks grumpy_mike i actually already had a look at your project : quite impressive i must say!
-did you notice any shorted lifespan of your leds/ir transistors since you switch them on/off at high frequency as well ?
-i also though about using a setup similar to yours, unfortunately having the sensors on the side of the screen this way can be a bit problematic for multi touch technology , since there are some cases where multiple aligned fingers/ object on an axis would be hidden, but i might use a similar configuration for the ir emitters themselves

-any other feedback or pointing out of potential problems is more than welcome !

4

did you notice any shorted lifespan of your leds

I didn't notice any life shorting but then I wouldn't. There is no problem with high speed switching of LEDs and reliability, there is no known mechanism of how this could occur and it is not a consideration in "professional" estimations of life time or reliability.

can be a bit problematic for multi touch technology

Yes agreed, did you mention it was multi touch, if you did sorry I didn't notice that.

I am considering using a modulated IR beam and a sensor with built in amplifier to overcome the stray light problem. At the Maker Fair I had problems with the light coming through the top of the tent swamping the sound square sensors. Where as the IR harp has it's sensors pointing down and uses modulated IR at 38KHz.