Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
You have here 2 option. I should say 3 option but the third option will be to build a colour sensor by yourself is rather unrealistic.
So option 1: buy of-the shelf colour sensor, for example: https://www.lasercomponents.de/pdf/mazet/integral.pdf
Option2: use CCD camera fitted with a pinhole lens and RGB output and process these signals. If a camera doesn't have RGB output but only composite video you can always use external circuit which will extract R, G and B components from composite ..
In the previous post I told about LDR
but the LDR i am using is having a sensitivity of 3msec. That means time taken for output of the sensor to change when input light condition change.
So if you want a good sensitive one then go for
IR led+photo diode
This is having sensitivity in uS because u know that
LDR is resistor so the response is slow while photodiode is a semiconductor one that will be faster.
If you are using IR then there is also a good possibility of increasing range of device
I used to IR led+photo diode combination for my robot to detect edge of a table, In that i got the range about 5-7cm, but in LDR you won't get that much and also u have to apply an LED or any other form of light with it for good response.
Select which one is suitable for ur application
Remember that photo diode work only in reverse biased region. The configuration remains same
Photodiode(or LDR)+series resistor.
ADC connected across any of them as you wish
Just another thought, if range was not too great, or colour sensing does not need to be very precise. How about a tricolour led or even 3 discrete leds. You could then drive these with 3 MCU analogue ports and scan through the colour spectrum. You would then use a single photodiode to look for a peak response at any given point in the colour map. This is only a theory but may stand a good chance of working.
I´ve seen this approach on robotroom.com, but if you only need 3 colors it would be better to use 3 filtered photodiodes (with a plastic or film of the color you want to sense over the photodiode) and a matching led of the color.
Of course you can always use a sensor like the ones from TAOS inc (though they are hard to get and the module that is sold in some stores is too expensive)
You can make color sensor with green,red,blue led and ldr.Output voltage of ldr in different led light is changing.You can test this output voltage with mcu has adc like pic18f458.
Depending on how accurate your color sensor needs to be, you may be able to use a photo resistor, three LEDs (red, blue, yellow) and a microcontroller with an ADC
I built one for proof of concept for an M&M sorter that I never built.
The process was basically:
Shine each LED separately, wait for the photo resistor to even out, and get a reading
The LED is not only light emitting but also light detecting.
So for the desired three colors: red, green and yellow take
LEDs- The beauty behind: They include the filter function ;o)
Some OpAmps and A/D converter and that´s it.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
