IR multiple frequency separation

[larsof54]

If I have two different frequencies of IR hitting a photodiode, can I use active filters to separate them? What I have is two parallel beams about 5.25 inches apart and need to isolate them from each other. The frequencies can be anywhere between 125 kHz and 400 kHz.

 

[keith1200rs]

Assuming the modulation frequencies are known and different then yes, you can separate them with active filters.

Keith

 

[larsof54]

so if one of the beams was at 125 kHz and the other at 350 khz, I could separate them without trouble?

 

[keith1200rs]

No problem. There are two main choices splitting with a high pass and low pass at around 200kHz or a bandpass at each frequency.

For designing high pass/low pass or bandpass, the TI Filterpro software is handy:

**broken link removed**

Keith.

 

[larsof54]

Thanks Keith. I will give it a shot soon.

 

[larsof54]

so why wouldn't the receiver only see a series of irregular pulses? Why would it "see" the individual frequencies of IR?

 

[keith1200rs]

I don't understand what you are asking. The IR receiver will see every frequency (within reason). If you have two separate data streams then you need to split them so there has to be something to differentiate them - in your case the modulation frequency.

Keith

 

[larsof54]

There is no data and no modulation. Just pulses of the 2 different frequencies

 

[keith1200rs]

... pulses of the 2 different frequencies

Well that's modulation i.e. the amplitude is 'modulated'.

Keith

 

[larsof54]

MODULATE: alter the amplitude or frequency of (an electromagnetic wave or other oscillation) in accordance with the variations of a second signal, typically one of a lower frequency : radio waves are modulated to carry the analog information of the voice.

There is no second signal that modulates the first..

 

[keith1200rs]

There is no second signal that modulates the first..

The light is the an electromagnetic wave modulated in amplitude by the pulses.

I don't know what point you are trying to make!

Keith

 

[FvM]

Refering to the original question, I see the problem that with pulse modulated light (square waveform), a highpass or bandpass for the 350 kHz signal also receives the 3rd harmonic of 125 kHz. So a narrow band pass or something like a LM567 tone decoder would be needed to separate both signals. It hasn't been said, what "separating" both beams actually means. Just detecting the presence or measuring an intensity? Also the intended switching speed would be important.

why wouldn't the receiver only see a series of irregular pulses?

The receiver would treat the input as analog signal rather than as a pulse series.

 

[larsof54]

Here's the whole picture.
I wish to measure the speed of an object using 2 InfraRed beams that are approx 5.25 inches apart. So, detecting the absence of a pulse train is my method of seeing whether a beam is broken or not.
For accuracy reasons I need to use frequencies of at least 125 kHz and because of the OP amp I wish to use, an upper frequency limit of 400 kHz is also there. I am concerned about the beams interfering with each other, so to isolate one from the other I thought that 2 different frequencies might work. That would be preferable (easier) to trying to isolate the 2 beams by physical means.
Hence my original question as to whether I could eliminate the presence of the "other" beam by using active filters. My uncertainty comes from using IR and how the transconductance amplifier would react.

 

[FvM]

I think, I basically answered the question. As long as the amplifier is operating in it's linear range, the signals don't interact in the amplifier. But if you supply rectangular pulses to the IR LED (either of 50% or lower duty cycle), the received signals are not clearly separated in frequency due to the harmonic content of the signal and can't be simply filtered out.

 

[Enlightenment]

SIMPLE SOLUTION to prevent interference: transmit each LED on the opposite cycle of a square wave, thus no interference. Technically there could be some interference at the point of the turn-on and shut-off but the point is make sure the LED's are not both turned-on at the same time.

 

[larsof54]

Unfortunately with battery operation I pretty much have to use an ultra low duty cycle with a high current to be able to receive at about 56 inches

 

[keith1200rs]

SIMPLE SOLUTION to prevent interference: transmit each LED on the opposite cycle of a square wave, thus no interference. Technically there could be some interference at the point of the turn-on and shut-off but the point is make sure the LED's are not both turned-on at the same time.

When you see light at the receiver, how would you know which LED it was from?

Keith

 

[FvM]

Unfortunately with battery operation I pretty much have to use an ultra low duty cycle with a high current to be able to receive at about 56 inches

Good to know about this point, because it changes a lot. Unfortunately, low- respectively band-pass filtering of the received signal cancels any signal-to-noise ratio benefit of low duty cycle signals. To utilize it, the receiver has to process the pulses with full bandwidth.

I fear, there's no easy solution in this case. Similar problems are present in signal processing of photoelectric sensors for industrial automation and have found various solutions, e.g. usage of "orthogonal" pseudo-random sequences.

 

[keith1200rs]

Why only one detector? With two detectors you could use different colour LEDs and optically filter. Or you could use a couple of well collimated sources, e.g. Laser diodes.

Keith

 

[larsof54]

There are 2 parallel beams and 2 receivers, separated by approx 5.25 inches and approx 56 inches between Tx and Rx. Color LEDs sound interesting to differentiate (distinguish) the 2 beams. from each other.