IR communication protocol/design for low latency/fast response - ideas?

[pwlps]

IRDA association did not pick phase modulation but PPM instead for a good reason.

No, you still don't get it. I give up.

[grifftech]

visable light, 1 wavelength per pair

[ogden]

IRDA association did not pick phase modulation but PPM instead for a good reason.

No, you still don't get it. I give up.

No. *You** don't get it - that it is pointless to double-modulate signal.

[pwlps]

No. *You** don't get it - that it is pointless to double-modulate signal.

I get it perfectly. It might be pointless for a single channel but the OP precised it has to handle independent communication channels. Buriedcode suggested using different carriers and I tried to develop the idea.
Here is the (basic) point you don't get yet. The direct OOK modulation such as used in IrDA (with various coding schemes, NRZ etc.) has a broad spectrum(fundamental defined by the bitrate+harmonics +possibly some subharmonics depending on the coding scheme) and it is not possible to define isolated channels by just using different modulation frequencies. The purpose of double modulation (carrier) is merely to shift the entire spectrum to some high frequency band, then choosing a large enough spacing between carriers you can have completely isolated channels.  I hope it was clear this time.

[ogden]

OP precised it has to handle independent communication channels.

Please be so kind and show quote where he said so.

Buriedcode suggested using different carriers and I tried to develop the idea.

You did not develop idea of different carriers. Different carriers in optical communications means different wavelengths. https://en.wikipedia.org/wiki/Wavelength-division_multiplexing

The purpose of double modulation (carrier) is merely to shift the entire spectrum to some high frequency band, then choosing a large enough spacing between carriers you can have completely isolated channels.  I hope it was clear this time.

LOL, no. You suggest to use same carrier frequency - same wavelength of light, same type of IR LED to transmit multiple channels? 

[Marco]

Wavelength division multiplexing is used because the individual bandwidths of the signals are already pushing technology limitations. Modulating a laser with a modulated 1 THz carrier to get 100 10 GHz bands is not technologically possible, with a MHz carrier and 100 10 kHz bands it is.

Now whether frequency division multiplexing is a better way to share the spectrum than time division here is debatable.

[pwlps]

LOL, no. You suggest to use same carrier frequency - same wavelength of light, same type of IR LED to transmit multiple channels? 

Of course it is possible to use the same wavelength of light and same type of IR LED to transmit multiple channels. That's what it all was about.

[ogden]

Now whether frequency division multiplexing is a better way to share the spectrum than time division here is debatable.

Of course time division of pulse-modulated signal is most effective way for said application - because available bandwidth is more than needed.

Of course it is possible to use the same wavelength of light and same type of IR LED to transmit multiple channels. That's what it all was about.

Well, then explain your phase modulation principles of IR LED light which does not allow phase (of the light) modulation. You BTW did not provide quote of OP asking  independent communication channels.

[Buriedcode]

Lads, the OP hasn't returned, can we at least keep on topic.

[ogden]

Lads, the OP hasn't returned, can we at least keep on topic.

Are you moderator? Which part of our discussion you find off-topic? We *are* talking about subject "IR communication protocol/design". You think we are not?

[pwlps]

Now whether frequency division multiplexing is a better way to share the spectrum than time division here is debatable.

Of course, but here we were not talking about how to optimize a standard for best spectral efficiency, the question was rather how to make a simple but robust system with a few channels, focusing on the ease and robustness of the actual implementation. TDMA will work with simpler electronics but will need a more robust software protocol (collision detection etc.), FDMA needs more analog stages but eliminates many problems too. Consider such an annex problem like the AGC training: in FDMA this is handled by the analog stage IC, but in a direct carrierless coding you don't have it and I don't even know how it is achieved, do we need to append long preambles ?

[pwlps]

Well, then explain your phase modulation principles of IR LED light which does not allow phase (of the light) modulation.

I don't have time to give a course on modulation now but note that there is nothing new here. Carrier modulation is used in IR remote control standards like Sony SIRC or Philips RC-5 : https://en.wikipedia.org/wiki/RC-5. In RC-5 "the command data is a Manchester coded bitstream modulating a 36 kHz carrier". As you see the carrier has nothing to do with the optical frequency. NB. In the standards I'm citing only amplitude modulation is used but there is no reason a phase modulation (BPSK, QPSK etc.) could not be used as well. Please try to understand all this before your next post.

[pwlps]

Lads, the OP hasn't returned, can we at least keep on topic.

As long as the discussion is entertaining who cares about the OP 

[ogden]

I don't have time to give a course on modulation

Don't even try please.

"the command data is a Manchester coded bitstream modulating a 36 kHz carrier". As you see the carrier has nothing to do with the optical frequency.

When you properly tell what actually is carrier - then sure. Look... I do not deny possibility of signal double-modulation, nor possibility to transmit QPSK-modulated baseband using amplitude modulation of IR light. What you propose is indeed possible to build, but it is pointless overengineering. It is already said multiple times here - pointless.

NB. In the standards I'm citing only amplitude modulation is used but there is no reason a phase modulation (BPSK, QPSK etc.) could not be used as well.

IR remotes do not use amplitude modulation but OOK (on/off keying), even for "modulation of the carrier". BTW IR receivers have pathetic selectivity - if you point interfering (transmitting) 56KHz remote to 36KHz receiver, it can't decode 36KHz transmission aimed at it. That "carrier modulation" for TV IR remotes does not create parallel transmisison channels, it is just smart balance between design complexity and protection against false triggering from ambient IR noise and other kind of TV IR remotes.