From what I remember this outputs analogue signal, and Rpi has no ADC, so you would need an intermediate circuitry, how precise and complex is up to you. Here is an article that can help: www.raspberrypi.org/forums/viewtopic.php?t=13883
+Dmitry Art there is no lens, it's just a surround to minimise stray light coming in. LED's are usual warm white, not IR. And the distance is small. You have to be close to get good readings. The further away, the worst it gets. I found it best just about 5mm away from the plastic surround, so maybe 10-12mm from the sensor. But bear in mind, if you remove the plastic the readings will go really bad - LED's will swamp the sensor with light. Cheers!
Tried this sensor (a bit different looking board, less advanced than this) a year ago, had high hopes for it, but it was very disappointing. Because of poor performance it has no real practical purpose.
Thanks. I do agree, the results were a bit inconsistent, i.e. one green was different to another green, quite a lot. So as far as precision this sensor is a no-go. But I think the key is to have controlled conditions, like no external light during measurement, and most importantly to have the micro 'learn' objects it will be measuring in the same controlled conditions of course. I think it is possible to make it work somewhat reliably this way. Anyways, this is the plan, so stay tuned ;)
🎯 Key points for quick navigation:
00:00 *Color sensor demonstration*
01:08 *Color to frequency*
02:03 *TCS3200 photo diodes*
03:38 *Sensor LED setup*
05:39 *Background affects readings*
07:11 *Reflective surfaces benefit*
08:21 *Controlled conditions crucial*
09:03 *Photo diode array*
10:21 *Possible camera use*
Made with HARPA AI
hi, can you share the code if you dont mind
NIce Video, thanks for sharing, a quick question, how can I modify the program to identify the colors of the rainbow?
Does it work with raspberry pi?
Don't see why not, you just have to implement it on gpio (probably there is already a library for that somewhere on the googles)
alright that helps a lot, thanks :D
From what I remember this outputs analogue signal, and Rpi has no ADC, so you would need an intermediate circuitry, how precise and complex is up to you. Here is an article that can help: www.raspberrypi.org/forums/viewtopic.php?t=13883
Or on second thought, just use Arduino nano as ADC and talk with Rpi to Arduino. Nano costs like 1 £/€/$/whatever :)
What max distance with lense? 4 leds, is it IR led or usual?
+Dmitry Art there is no lens, it's just a surround to minimise stray light coming in. LED's are usual warm white, not IR. And the distance is small. You have to be close to get good readings. The further away, the worst it gets. I found it best just about 5mm away from the plastic surround, so maybe 10-12mm from the sensor. But bear in mind, if you remove the plastic the readings will go really bad - LED's will swamp the sensor with light. Cheers!
Tried this sensor (a bit different looking board, less advanced than this) a year ago, had high hopes for it, but it was very disappointing. Because of poor performance it has no real practical purpose.
Thanks. I do agree, the results were a bit inconsistent, i.e. one green was different to another green, quite a lot. So as far as precision this sensor is a no-go. But I think the key is to have controlled conditions, like no external light during measurement, and most importantly to have the micro 'learn' objects it will be measuring in the same controlled conditions of course. I think it is possible to make it work somewhat reliably this way. Anyways, this is the plan, so stay tuned ;)
can i get the code please?
In the video description here: ua-cam.com/video/GPFqbrSzc68/v-deo.html
there is a link for pastebin with the code I used. Hope this helps.
yess thank you :DD