Why is the IrDA Baudrate limited to 115.2kBd?

[flyer31]

Posted on May 13, 2013 at 20:56

Hi, in STM32F4 RM 26.3.12 it says, that ''USART supports only bit rates up to 115.2kBps for the SIR ENDEC''. Can anyone tell me the reason? (115.2kBd must be an increibly low data rate for a 168MHz processor).

In fact I tried the USART_Printf example in STM32F4xx_DSP_StdPeriph_Lib_V1.0.0 and enabled IrDA there, and it was no problem at all to increase the baudrate to 1MBd, the chip created nice 3/16 output pulses as expected. I did not try yet the reception part neither the IrDA transmitter. But as now there are IrDA transceivers available until 4MBd and even 16 MBd, I would expect that there is problem on the transceiver side? So did any other guys here perhaps already try to use IrDA at some higher baudrate than 115 kBd and could possibly share their experiences? #irda #sir

[Andrew Neil]

Posted on May 13, 2013 at 22:54

''Can anyone tell me the reason?''

  Because that's what the IRDA

SIR

  Specification says:

https://en.wikipedia.org/wiki/Infrared_Data_Association#IrPHY

http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0183g/I31065.html

http://msdn.microsoft.com/en-us/library/windows/desktop/ms691781%28v=vs.85%29.aspx

''there are IrDA transceivers available until 4MBd and even 16 MBd''

SIR

  is a very simple modulation/coding scheme, only specified to 115k2. Other schemes are used for higher rates

https://en.wikipedia.org/wiki/Infrared_Data_Association#IrPHY

[flyer31]

Posted on May 13, 2013 at 23:26

Ups - very interesting (you did not get the lamp from me - but I would given you one if not somebody else would have come first).

As we want to use the IrDA communication only for some internal STM32-STM32 processor communication (NOT to communicate with any ''Standard IrDA'' devices) - do you think it would make sense to try to use the 3/16 RZI scheme up to 1-1.5MBd (only for relatively low distance, range 20-80cm would be fine - we need it for STM32F1 chips)? Or would we be running into large problems there for sure? PS: As you seem to be such an expert in optical transmission - do you have an idea, which encoding scheme is used for the Photolink/TOSLINK fiber transfer (red laser light, as it is currently getting very fashionable in Audio equipment, also for many CD players, ...)? Do they use any encoding at all, or perhaps just standard inverted UART signals?

[Andrew Neil]

Posted on May 14, 2013 at 09:18 Forum playing up again

[Andrew Neil]

Posted on May 14, 2013 at 09:20

''As we want to use the IrDA communication only for some internal STM32-STM32 processor communication (NOT to communicate with any ''Standard IrDA'' devices) - do you think it would make sense to try to use the 3/16 RZI scheme up to 1-1.5MBd (only for relatively low distance, range 20-80cm would be fine - we need it for STM32F1 chips)?''

  That's a very good question!

As you're not constrained by IRDA standards or interoperability, you're free to do whatever you like - but also responsible for the results!

I would guess that an (the?) important thing is the amount of energy in each pulse:  as you go faster, the pulses get narrower, so the energy in each pulse gets less - which is going to make the receiver's job harder. Particularly under ''adverse'' conditions.

''As you seem to be such an expert in optical transmission''

  I'm not - I just googled ''IRDA SIR''

I would suggest that you spend some time researching the issues; the ARM doc references an ''Agilent IrDA Data Link Design Guide''  and, of course, the IRDA's own site

http://infocenter.arm.com/help/topic/com.arm.doc.ddi0183g/DDI0183G_uart_pl011_r1p5_trm.pdf

I would also suggest that you try some experiments with the transceivers you intend to use - particularly how they behave at higher speeds under ''adverse'' conditions.

[Tesla DeLorean]

Posted on May 14, 2013 at 11:23

only for relatively low distance, range 20-80cm would be fine

In air, or down a fiber? The former has a whole bunch of multi-path issues, ie signals arriving via alternate paths, whose travel times are delayed by the increased distance of travel and C.