How do IR remote controls use frequency modulation to avoid interference?

[PhysEd]

Hi guys, my first post on these forums.

I am learning electronics and am still very new to this all.

My current tinker project is to learn how regular tv remote controls work.

So far I have understood that the remote sends blinking IR pulses through an 940 nm LED.
The remote uses frequency modulation to help avoid noise and interference.
The remote uses some protocol depending on which brand it is.
And the receiver uses an IR photo transistor with some demodulation and then to some microchip to do the work.

I have hooked up a very simple circuit, which has a 5v power supply, a resistor and a photo transistor that goes to my oscilloscope.

When I use the remote control to press a key close to the phototransitor, I can see the transmission on my oscilloscope just fine and I have even decoded the protocol so I understand everything going on there.

But my question is this, why am I seeing the protocol data on my oscilloscope instead of the frequency modulated signal?

If it helps, I am using the samsung protocol / remote and according to their specs, the carrier frequency is 38.9 khz.

Thanks.

 

[schip666!]

I think you are asking why you don't see the actual modulated signal pulses at 39K, yah?

If so I'd guess some capacitance in your receiver, or perhaps the photo detector you are using is a bit too slow. Try some different detectors and maybe decreasing the resistor value to get more current to over-come stray capacitance.

That said, I remember having trouble seeing the chopping frequency too...

 

[PhysEd]

Some very good tips.

But, I am suspecting I have something backwards here.

As far as my understanding, the oscilloscope should not even be able to separate the original data from the carrier signal, but I see no trace of the carrier signal, I only see the data.

If you look at this image:
http://img.zdnet.com/techDirectory/_FMMOD.GIF

I am expecting to see the "Modulated Result", but I am seeing the "Modulating Signal".

Maybe the remote controls frequency modulation differs from what I think it does?

 

[Mike_In_Plano]

Using a photo transistor is most likely your problem. They have gain, but are very slow. Chances are, you're seeing the envelope of the signal - and even that probably has ugly slopes.
Most wide band receivers use a PIN diode. You keep it reverse biased and the current that comes through is proportional to the number of photons. With a 5 volt source, and a 1K resistor, you'll be seeing the modulation in no time ;-)

- Mike

 

[PhysEd]

I'll will try that, right now, to ease the learning curve, I am trying to to ghetto-send the IR signals through a serial port connected to an IR LED.

Let's see if I have better luck on this end.

 

[schip666!]

Hmm, good point about photo-diode vs PIN diode... I was trying to communicate the signal envelope concept, but failed greatly in comparison to Plano-Mike.

Also I believe you should see Amplitude Modulation, not Frequency as in your FMMOD.GIF. Basically a bunch of pulses filled with the 39KHz wiggles alternating with empty space-time.

Will be interested to hear about the "ghetto-sending" idea. I suppose the generated signals are actually serial format data, so it might just work.

 

[PhysEd]

Ok here is a quick update (for anyone that cares :).

As for me not seeing the frequency modulation, I still think I got that part wrong.
From the responses here and some more reading, the "modulation" part in IR remotes is actually a 38.9 khz wiggle that turns on and off during specific time periods (lets say 0.56 ms for logical 0 etc).

It does not work the same way as normal radio FM, where you actually put the digital signal inside an FM carrier like my image suggested. (This might still be wrong, but at least it makes a bit sense now why my oscilloscope can see the perfect digital output).

As for my serial transmission project, I have made some progress.
It definitely looks like it might work, but me being a noob and all, I still have to learn how the serial port works etc. Basically I need to find a baudrate and the correct string to send to make the wave look like the remotes output wave. Once I figure that out, I should have my "ghetto-remote" :).

 

[Mike_In_Plano]

Sounds fun in any case.
IR links were looking to be an up and coming thing at one time. There were chips to support it, but then people got interested in RF...
As I recall, they would send either a narrow or wide IR pulse (simply on - not modulated) to represent marks and spaces from the serial port.

Since the serial sequence is what you're trying to control and the 39'odd carrier is just there to make it detectable, I'd be tempted to just use a 555 running at the carrier frequency and enable/disable it with your serial line.

-Mike

 

[PhysEd]

Mike, I think you just solved a problem I faced.

I was trying to make the digital waveform match the remotes output and I kept getting into problems with the RS232 stop bit. If it wasnt for the stopbit, I could just send x amount of 0's and then y amount of 1's to simulate the remote's wave.

But now that you mention it, it would be smarter to use a 555 timer and then just use the serial port to turn it on and off.

So much to test, so little time :)