Color sensor distance problem

Hi,

I want to detect three colors(RGB) by using a circuit.I tried LDR and three leds(red,green,blue) but they are not sensitive.For example I can not detect colors more than 5 cm away.It should detect at least 10 cm away from LDR.

What do you suggest?Should I use other thing?

 

Another commercial detector option: http://www.taosinc.com/...look at Color Sensors

To increase your range, three options come to mind:

1. Increase the amplification of your detector circuit.
2. Add a lens system to concentrate the light on your detector.
3. Increase the amount of light shining on your colored surface.

Ken

 

Without details I don't know if this is useful, but you could use optical filters in front of each color detector. Also, you could modulate the light sources at different frequencies, and make detector circuits that look for those frequencies.

 

this is my circuit.I used two blue leds one for red one for green.Black thing in the middle is LDR.How can I increase the sensibility(range of measurement)????

 

You forgot to attach your schematic. Attach it here, not over at ImageVenue.
The schematic will show the supply voltage, the LED driver and the series current-limiting resistor so we can see how much current there is in the LEDs.

The schematic will show if something is loading down the variable resistance of the LDR's.

 

That is not a circuit - it is the arrangement of the parts. Can you post up a schematic? Something about 4 times larger than the fuzzy image, please.

By the way, KMoffett's link to the TAOS devices is a good one. A LED is very insensitive by comparison. You would need color filters, though.

 

Finally,I make some comparison and show the color...

Here is what you want...

 

I would think that to use a LED as a color sensor you would have to bias each LED to a point where is was just barely conducting and do the sensing between the resistor and the LED.

That can't be your schematic, by the way. Where is the CDS cell? What function does it provide? The uC can't possibly pull up the LED's and measure conduction at the same time.

 

I think he's lighting the colored LED's one at a time, and measuring the voltage at the resistor/LDR voltage-divider, on the left of the µC. That gives him one value for each of three colors.

Ken

 

yeah thats whAt I do...

I dont know how I increase the range...

 

From my earlier suggestions:

#1...Add an opamp with gain and adjustable offset between the voltage divider and the PIC's ADC input.

#2...Judging from the image of your LED/LDR arrangement this might be possible. A lens over the LDR, and tilting all three LEDs to aim at a single point on the colored surface.

#3...If you drive the LEDs with more current you get more light and more response from the LDR. You can briefly drive LEDs with much more current than their continuous max rating of usually 20mA. This would probably require an external drive BJT transistor or MOSFET so as not the exceed the PIC's output current rating.

Ken

 

Why don't you try using 3 sensors and 3 colour filters - red green and blue? Read the output from each sensor. The highest value will indicate the colour to the lamp that is lit.

 

I think he's trying to determine the color of the surface that is reflecting light from the LEDs back to the LDR. Maybe 3 LDRs with colored filters, and a white light source. I don't know if one, or more, white LEDs would work for him.

Ken

 

Color Sensors are commercial products not a textbook topic, so nobody will 'share' his commercial activity/ technique/ etc. with others. People will provide you awareness but not "turn key solutions".

 

Sorry raufjaleel, you are 2 years late on this thread. But, It may help someone that is googleing for this info.