White line sensor circuit, determening currents

[rusty009]

Hey, I have this QTI sensor ( circuit is shown, the left schematic ) and I am trying to get it to detect a white line only, no red,blue or green just a white line. 5 volts is applied to the pin called White(top pin). Now when the IR light is emited from the IR LED will the Ie, emiter current from the photo transistor, be greater or smaller than it would be if there was a black surface underneath it ? Thanks.

 

[berkeman]

Hey, I have this QTI sensor ( circuit is shown, the left schematic ) and I am trying to get it to detect a white line only, no red,blue or green just a white line. 5 volts is applied to the pin called White(top pin). Now when the IR light is emited from the IR LED will the Ie, emiter current from the photo transistor, be greater or smaller than it would be if there was a black surface underneath it ? Thanks.

You want to detect a WHITE line with an IR emitter and detector? What's wrong with that hope?

 

[rusty009]

I'm going to measure the reflectivity of the surface by determining the current through the transistor and then hopefully find the colour. What is wrong with it ?

 

[berkeman]

I'm going to measure the reflectivity of the surface by determining the current through the transistor and then hopefully find the colour. What is wrong with it ?

What is color? What consititutes "white" light? What is the emissivity curve of the IR source you are using? What is the sensitivity curve of the IR pickup you are using?

What is the fundamental difference between IR (what does IR stand for?) and white light?

If you want to detect a white line, and distinguish it from other color lines, what do you need to use as a light source? What kind and how many detectors do you need to use?

 

[rusty009]

White light is red,green and blue light together.. infrared light has a higher wavelength than visible light. I have the sensitivity curve of the sensor do you want me to upload it ? I would of thought I only need to use one sensor as the amount of IR light reflected should determine the colour...

 

[berkeman]

White light is red,green and blue light together.. infrared light has a higher wavelength than visible light. I have the sensitivity curve of the sensor do you want me to upload it ? I would of thought I only need to use one sensor as the amount of IR light reflected should determine the colour...

Nope. If you want to distinguish a white line from other color lines, you will need a light source with what colors in it? And _____ detectors, each tuned to what colors?

 

[rusty009]

do you mean a blue green and red light source so three detectors ? I really don't understand that :S... but do you understand my idea of the reflectivity of the surface by measuring the current through the transistor ? And if so why is it a bad idea...

 

[mheslep]

White light is red,green and blue light together.. infrared light has a higher wavelength than visible light. ...

longer wavelength, greater frequency

 

[mgb_phys]

Imagine the IR detector as a black and white camera.
In a black and white movie can you tell the difference between a white object and a red, green or blue one?

 

[rusty009]

Imagine the IR detector as a black and white camera.
In a black and white movie can you tell the difference between a white object and a red, green or blue one?

Ok, I get what your saying. But... If i send IR beam on a blue surface I do not believe the reflectivity will be as high as that of a white surface... You can't see blue green or red inblack and white but blue green and red do not show as white , they are a shade between black and white.

 

[mgb_phys]

Correct but you would have to carefully calibrate it for the incident light and the detector. You also have to be careful that a lot of colored materials have different reflectivity in the infrared depending on the dyes used.

You could use an IR source and an IR detector and set a limit that only the white will trigger. Reflective silver tape will give an even stronger response of course.
If you are fussy about detecting white you need a color sensor ie measure red,green,blue and pick the line where the signal is the same in all bands.
The easiest way is to have a weak IR source and a cover over the source+detector so you are sure that the light is always the same and pick the line that gives the strongest response.

 

[rusty009]

You could use an IR source and an IR detector and set a limit that only the white will trigger.

Ok, I set the limit by using a schottky diode ? But my question is, is there anyway to calculate the current through the transistor from the reflection of the white line without actually measuring it with a multimeter ?

If you are fussy about detecting white you need a color sensor ie measure red,green,blue and pick the line where the signal is the same in all bands.
The easiest way is to have a weak IR source and a cover over the source+detector so you are sure that the light is always the same and pick the line that gives the strongest response.

Unfortunately I have to make do with the QTI circuit shown above, but concerning the cover, what do you mean by a cover ?

 

[mgb_phys]

but concerning the cover, what do you mean by a cover ?

Presumably you are doing this in a lit room?
If you just rely on the room lights it will be difficult to control the reflected light level.
So you need some sort of light proof cover over the IR emitter, the IR detector and the line. Black cardboard will probably do.

 

[rusty009]

Black cardboard will probably do.

Is that a joke :S ? Back to my first question, can I calculate the current through the transistor ?

 

[mgb_phys]

No it's not a joke.
If you just have the IR detector pointing forward you will get the reflection from the white line multiplied by the continually changing daylight, shadows from people walkign around etc.
If you make a black tube from the detector to the floor then you will only receive the light coming directly from the line. And something simialir from a stable IR source LED.
You should also arrange so that light from the LED can't get into the detector directly.
Most optical instruments and experiments consist largely of black cardboard.