Just wanted to share what I was playing around with tonight. While at work I started thinking about the paper towel dispensers that are "touchless" and began to just play around with some ideas tonight. I wanted to do this with a photoresistor (becasue that's what I have already) and it seemed like the tempsensor code would be a good place to start (again, since I had that already lol). Now having played with the LM34 aside from the tempsensor project I know that connected to +5v it outputs ~0.8v at my room temp. This told me that inorder to use a photoresistor for this project I'd need to bring it's voltage down to around that range. I just used some resistors for this (nothing fancy here).

Then I connected this "input" voltage to the ADC per the tempsensor project and watched the readings on the LCD change as I moved my hand closer and farther away. Then it was just a matter of setting one of the pins to output (which I just used with transistor and an LED to play with the concept) which could control a mosfet to open your ice bin, fill your cup with water, etc.

One of the things I noticed would be a problem with this is the ambient light in the room changing (if the light is dimmer/brighter than usual that's going to affect the ADC reading which could affect the functioning of the 'device'). Seems like the quick and dirty way around this was to take a reading from the ADC every so often to get a "baseline" level of the light in the room. Now you can set up a conditional statement evaluating something like "if reading is 20% lower than baseline then do something". This method could possibly be used to adjust how sensitive the device is as well.

I didn't actually make anything useful with this (other than an LED turn off and on), this is just what resulted from my "playing" tonight and I just thought I'd share. Anyone else done something similar and have any ideas to share?

Medic -

Cool concept. If you want to see something similar, look no farther than the NK tutorial on the "Halloween Capacitive Touch Sensor" HERE. It uses the exact same logic (even the "baseline" terminology) only it uses electrical capacitance instead of ambient light changes.

One thing I would explore with your project would be a second light sensor (pointing in a different direction maybe?). The second sensor would measure the general light conditions and the first would measure local disturbances. That way if a drastic light change occurred (like someone turning off the overhead light) you wouldn't trigger an event since both sensors would detect the change.

The two sensor scenario is what is used in automatic dimming car rear view mirrors. There is a front facing sensor that just checks ambient light and a rear facing sensor that detects the headlights of the car behind you. I hadn't really thought much about using a CDS cell for near proximity motion detection, but I bet it would work quite well. Good call on the two sensor thought pcbolt.

Rick

Darn, just what I need something else to play around with.

I like the concept and then with the two sensors it really ought to work nicely.

Ralph

I made a capacitive touch sensor as well using a little different hardware but what was nice about it was that all i had to do was increase the resistance and then i could use a hand swipe ( like what they used on star trek to open and close doors lol ) to show the readings and then based on what numbers i gave it the led would turn on or off on the output side of the controller. Here is a link to get you started, no sensors required just a little metal tape and some wire and a controller with some cheap and large Resistors.

http://www.pjrc.com/teensy/td_libs_CapSense.html