suggetion for line follow sensor

[aruna1]

guys can you suggest good emitter detector pair which has minimum effect on ambient light? (one that will work on evan bright sunlight with some shielding?

LDR-LED pair

IR reciver-transmitter pair?

LED(photovoltaic) - LED pair?


or any other?

[galannthegreat]

Ones like these will work very nicely:
http://www.solarbotics.com/products/qrd1114/ (make sure you put good shielding around the area though)

Also, for some more info on line followers look at David Cook's site:
http://robotroom.com/Jet.html(very fast line follower, great info here)
http://robotroom.com/Sweet.html(basic line follower, good info aswell)

[Joker94]

And have a look at this one to

http://www.societyofrobots.com/robot_mobot_2007.shtml

[Trumpkin]

Ones like these will work very nicely:
http://www.solarbotics.com/products/qrd1114/

I second that, these are very nice sensors.

[galannthegreat]

They use them on their Sumovore kit, and man do they work very well. It's pretty much an IR emitter and reciever in a convenient package.

[aruna1]

i was thinking of making something my own

[galannthegreat]

If that's what you want to do, definately, try it out. These will still be there if they're needed.

[Soeren]

Hi,

Making your own will enable you to use a greater distance as well - 6mm max. isn't much.

[galannthegreat]

Yea, that is very true, if you make your own the distance can be increased in which your sensors lie away from the floor/ground.

[paulstreats]

I have some of those sensors (from sparkfun though). I can vouch that they work at greater distance than 6mm. (I have used them as proximity detectors for distances around 30cm. running them through a decent transistor will produce a better output range)

[galannthegreat]

Could you show us a circuit of how you set that up?

[Soeren]

Hi,

I have some of those sensors (from sparkfun though). I can vouch that they work at greater distance than 6mm. (I have used them as proximity detectors for distances around 30cm. running them through a decent transistor will produce a better output range)

Exciting, not the least since Fairchild Semi disagrees (read page 3 of the datasheet). I never got any particular range out of them either and to be absolutely honest... I'd like to see the setup for a 30 cm detection range with a QRD1114 as well, so please post what you did.

[paulstreats]

I did it quite simple really, didnt really have to boost the output at all(it took a bit of trial and error to get the right resistors and diodes but it was the only IR sensor available when I really needed one so I persisted). I still have the circuit built so heres the schematic:-



Obviously it wont produce a digital high. But out of my 10bit analog input I get a reading of 997 for a distance of 2mm ranging to a reading of ~61 at just over 30cm. Any further and its too noisy to stretch.

The diode really helped reduce the noise to the point that it really made the sensor work. (its not just a fluke either, I have 3 sensors in a row that all produce a similar reading. Around a 4% error between sensors, but I ut that down mainly to noise).

The distance reading isnt linear, so its not good for map making but obstacle avoidance its okay for.

[galannthegreat]

Nice, now I got to try it out myself.

[Soeren]

Hi,

Obviously it wont produce a digital high. But out of my 10bit analog input I get a reading of 997 for a distance of 2mm ranging to a reading of ~61 at just over 30cm. Any further and its too noisy to stretch.

Cannot be a QRD1114 then!


Look at the above graph (directly out of the datasheet).
QRD1114 peaks at 25 mils (0.635mm).
At 2 mm (78.74 mils), the signal is down to around 53% of the peak output
61 is 6.1% of 997, so that would be 6.1% of 53% of the peak value = 3.24% of peak
Give or take a little to counter rounding errors, at your count of 61 you get a
max. distance of 270..280 mils, or in other words  6.858mm to 7.112mm
That doesn't play well with a claim of 30 cm (300 mm) IMO
Perhaps you ment 300 mils, which is attainable?

The diode really helped reduce the noise to the point that it really made the sensor work.

How would a diode in that place reduce noise?

The value 61 on a 10 bit 5V system represents 298mV. To get that low you'll have to remove the connection to the phototransistor, as any bias would make the output (to "port A") at least ~650mV due to the 1N4148 and that's not even counting the effect of the resistors (speaking of which... The 1.5k resistor doesn't really do much as it's less than ½%).

Don't wave dead chickens over your circuits - sacrificing virgins are so much more fun

[paulstreats]

Its definately a QRD1114.

The 1.5k resistor was to make it up to 331.5K (which when I originally used a POT produced the best results. When I swapped out the POT, a 330k and 1.5k were the best combinations that I had).

How would a diode in that place reduce noise?

Without it, it wouldnt produce a stable reading whatsoever. I thought that with the motors, scanning servos and other sensors that were involved then it might have supressed ground noise?

sacrificing virgins are so much more fun

Thats too messy for me.. Cant be doing with all the blood