IR sensors range and noise

Hi, I am using IR sensors(transmitter and receiver pair) to detect an object.The object to be detected is a plain white colored. Basically I am having two problems with using my module viz 1) The range of IR sensors is very small about 2cm only(I want to increase it to at least 5cm). 2)The output of the receiver is noisy and affected a lot by ambient light. I am running the transmitter on 4Khz clock and 5V and the output of receiver is 0-2.1V in the range 0-2cm(I want to make this 5cm). I am using a high pass filter of 50Hz to reduce the effect of ambient light. Can anyone please suggest me on how to increase the range and reduce the noise. Please let me know if some other filters using opamp can be more effective in reducing the noise. Thanks

I take it that this is a reflection detection sensor, not a beam interrupt sensor?

If you are chopping the transmitter at 4KHz, then why filter the receiver only at 50Hz? The amount of reflection of the transmitted beam is indicated by the peak to peak level of the 4KHz received signal.

In this case, you not only know the frequency but the phase of the transmitted signal. That should allow you to do a synchronous detector, which will be more effective (tighter band) than a simple passive low pass filter.

In fact there is nothing magic about 4KHz. Putting both the transmitting and detecting algorithms in the same micro should greatly help eliminate external noise. The micro would turn on the transmitter, wait a few microcseconds for the receiver to stabalize, take a reading, turn off the transmitter, wait, etc. It would subtract the off reading from the on reading, then low pass that result a bit.

You should also be able to extend range with the right optics. If the transmitter sends out a narrow beam and the receiver looks only along the same narrow beam, then more of what the receiver sees will be a function of the transmitted beam, which increases signal to noise ratio.

Thanks for your reply.
Just let me know if the technique for noise reduction is correctly interpreted by me
1) Turn the transmitter via micro.
2) Give a delay a some usec and then read the output of the reciver i.e receiver reading should be taken after some time after the transmitter transmits. This should reduce the noise right!

That's half of it. The other half is to do the same thing but turn the transmitter off. The signal you are looking for is the difference between the on and off readings. I would also low pass filter this difference. In other words, take multiple differences into account when deciding what the signal level is.

The theory is that the ambient noise is low frequency and not synchronous to your turning the transmitter on and off. For any short interval, you assume the ambient light level is constant, so the difference between what you measure with the transmitter on and off represents the the transmitter's light seen by the receiver.

Another way to look at this is that the off reading is the ambient light level and the on reading the ambient plus reflected light level. You only care about the reflected light level, which is what you get by subtracting the two.

Thanks, got it now what are you saying

is there physical room to mechanically 'filter' the sensor? a tube extending from the sensor to reduce ambient light?

Yes I have just wrapped the sensor on sides with a black rubber.

Extremely happy the Micromouse 'bug' is still alive in you!! Good luck and call me if you need any help

You should also use an InfraRed filter over your IR sensor to block ambient visible light that can cause DC shifts. If you increase your transmitter frequency you can add a bandpass filter centered at that frequency with smaller components and this will filter out unwanted signals. Stay away from 40KHz to 100KHz range as this is where many IR remotes operate that could interfere with your readings. And be aware that your target color in visible light is not what your receiver sees in IR light. For instance, some blacks may be bright some whites may be dark in IR. The color and material of your target will affect your effective range unless it is controlled or you take other measures to remove this difference.

Yes, there can be quite a difference between reflectance or transmission in visible versus IR.

I found out the hard way once that the common black electrical tape is rather transmissive in IR. Even three layers of it in a IR photo interrupter slot didn't trip the interrupter, even though that blocked all but a tiny fraction of visible light.

That is surprising. I would have thought electrical tape would have blocked IR. I remember being surprised when I first learned that glass blocks IR.

If you are not too rigid about using IR sensors, you might want to try Ultrasonic detection, the range is longer than its IR counter parts and it is less interference prone than them, it uses the howlround or feedback effect experienced between mic and speakers. But in this case your speaker is an ultrasonic emitter and your mic an ultrasonic receiver.