Not bad English, just didn't know to what you were referring...
I would place a small wager that you are seeing the MAC protocol
backoff, although 20 micro-sec seems a little short...
Search this list for MAC related stuff, I think there is an
approved way to change the behavior.

MS

Carlos Gil Soriano wrote:
> Hi Michael,
> 
> 
> I am sorry because of my bad bad English ;D.  What I like to mean it's 
> that after sending a message, CC2420 automatically goes to a reception 
> state (according to FSM in data sheet). After this, -in the case that 
> you sending a message (so you're using continuos sending)- it appears a 
> variable timing according to oscilloscope readings.
> 
> In the programs that I have made, I have tried to avoid any kind of 
> access to another devices (for ex. Flash memory). So, in between TX 
> state and RX, it should be happening that SPI bus is busy (at least for 
> filling TXFIFO memory), but another strange variable delay appears 
> (that's what I called it "spare variable timing"). That variable delay 
> is something unknow for me.
> 
> If you need anykind of further info don't hesitate to mail me.
> 
> 
> Regards,
> Carlos
> 
> 
> P.S.: in my programs, if I try to send a message in imote2 with a 
> payload greater than 80 bytes it fails. It is NEVER been send, so I 
> think that this is a big problem. I don't know why it is happening. If 
> someone knows why it is happening I would appreciate that they report 
> its experience by posting it here.
> 
> 2009/6/6 Michael Schippling <[email protected] <mailto:[email protected]>>
> 
>     I'm not sure what "spare variable timing" means, but the protocol
>     does have a random backoff 'feature' for CDMA (or whatever the
>     correct Multi-Letter-Acronym is). Another delay might be due
>     to something sampling the SPI bus for other activity. Can you
>     verify that no other SPI access is being attempted?
> 
> 
>     MS
> 
>     Carlos Gil Soriano wrote:
> 
>         Hi Michael,
> 
>         You're right regarding to SPI problems. I don't expect to find
>         problems with SPI in micaz. I've been testing continuous sending
>         in cc2420 in micaz and, after carefully read data sheet and all
>         code (line by line), there's one spare "variable timing". The
>         value of this timing moves from 200 us to 2000 us.
> 
>         I have requested further information of 2420 to Texas so as to
>         discern if that timing belongs to radio transceiver or not. I'm
>         waiting to hear from them. Anyway, I get the impression that
>         this timing is excessive for 2420, so here it comes SPI bus.
> 
>         Another issue comes with CC2420 in imote2. I've found out that
>         when you try continuous sending in imote2 with a payload greater
>         than 80 bytes it fails. I do know why but it does so (at least
>         in a couple of programs that I've made). I feel that the source
>         of that kind of problem could be SPI bus, too (I haven't check
>         timings, but I suppose that the "variable timing" should be
>         considerably higher).
> 
>         I'm a newbie in SPI studying. So I would appreciate any kind of
>         help.
> 
>         Regards,
>         Carlos
> 
>         2009/6/5 Michael Schippling <[email protected]
>         <mailto:[email protected]> <mailto:[email protected]
>         <mailto:[email protected]>>>
> 
> 
>            I think it's one of those little lapses that makes TOS so
>         intriguing.
>            Xbow seems to have chosen not to publish a full schematic,
>         only the
>            bits that changed from the mica2 when they added the
>         different radio.
>            So what's on that 2420 radio page is probably the best you'll
>         find.
>            I do remember something about the SPI lines needing to have
>         pull-ups
>            added in order to work correctly with external devices...
> 
>            I'm just about to embark on my SPI learning curve so I can't
>            pontificate,
>            but I'd be surprised if there were timing problems due to
>         hardware that
>            behaves according to spec.
> 
> 
>            MS
> 
> 
>            Carlos Gil Soriano wrote:
> 
>                Thank you very much Michael,
> 
>                what I am looking for it is the way in which all the SPI
>         modules
>                are interconnected. I am studying timings in CC2420
>                architectures and I have found one timing that should be
>         caused
>                by SPI factors.
> 
>                Before starting to read carefully all the code which involves
>                SPI bus I would like to know micaz schematic to clear up my
>                mind. I would like something similar to the mica schematics
>                found at http://www.tinyos.net/scoop/special/hardware.
> 
>                Thanks for your help,
>                Carlos
> 
>                2009/6/2 Michael Schippling <[email protected]
>         <mailto:[email protected]>
>                <mailto:[email protected] <mailto:[email protected]>>
>         <mailto:[email protected] <mailto:[email protected]>
> 
>                <mailto:[email protected] <mailto:[email protected]>>>>
> 
> 
>                   I thought it was in the User's Manual:
>                        
>         
> http://www.xbow.com/Support/Support_pdf_files/MPR-MIB_Series_Users_Manual.pdf
>                   but you need to piece it together using the mica2 bits as
>                well. I have a
>                   nice mica2 schematic here:
>                   http://www.etantdonnes.com/Motes/mica2_sch.pdf
>                   which I can't find at:
>                http://www.tinyos.net/scoop/special/hardware
>                   anymore. Maybe that will help a bit.
> 
>                   MS
> 
>                   Carlos Gil Soriano wrote:
> 
>                       Hi everyone,
> 
>                       I was looking for the schematics of MICAz but I've
>         found
>                       nothing. Do you know where I can find them? I've
>         taken a
>                look to
>                       xbow site but it only appears a few parts of the
>                schematic, not
>                       the whole one.
> 
>                       Regards,
>                       Carlos
> 
> 
>                            
>         
> ------------------------------------------------------------------------
> 
>                       _______________________________________________
>                       Tinyos-help mailing list
>                       [email protected]
>         <mailto:[email protected]>
>                <mailto:[email protected]
>         <mailto:[email protected]>>
>                       <mailto:[email protected]
>         <mailto:[email protected]>
>                <mailto:[email protected]
>         <mailto:[email protected]>>>
> 
>                            
>         
> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
> 
> 
> 
> 
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to