We are porting the cc2520 to a new mote and also we are modifing it to use
a range extender which requires a change to the h/w interface which ripples
to the s/w interface as well.

So at some point I will need to dig into it.   The 2520 driver is in the
trunk.   I know the cortex folks (t. schmidt et al.) are using it.

On Fri, Feb 17, 2012 at 4:22 PM, Sergio Gonzalez <[email protected]> wrote:

>  Yes, that is what the explanations in the corresponding interface files
> suggest.
>
> Are you developing a driver for the CC2520? My understanding is that it
> was done a few years ago as part of the Blaze project in tinyos-contrib.
>
> Sergio
>
>
> On 02/17/2012 01:10 PM, Eric Decker wrote:
>
>
>
> On Fri, Feb 17, 2012 at 3:48 PM, Sergio Gonzalez <[email protected]>wrote:
>
>>  Hi Eric,
>>
>> I don't know whether the radio drivers for the respective chips were
>> written so that if sendDone() returns a success value, then it means that
>> the packet's transmission is indeed complete.
>>
>
> I know explicitly what sendDone means is that all of the packet has been
> handed off to the radio's fifo.   It is the signal that says you can send
> the next packet.
>
> It does NOT mean the packet has left the hardware and has been completed
> as far as the transmission is concerned.
>
> I'm not sure what exactly sync does.   I haven't looked at what that
> does.   Will probably get to it in the next month or so.   At least for the
> 2520 chip.
>
>
>
>> Hopefully someone more experienced on transceiver details can help.
>> Either way, the available interfaces do not allow to specify the
>> transmission channel on a per-packet basis, so for the time being, using
>> sync() and syncDone() seems to address Ramakrishnan's need.
>>
>> Sergio
>>
>>
>> On 02/17/2012 11:38 AM, Eric Decker wrote:
>>
>>
>> That still doesn't help with seeing when the end of the packet has
>> actually gone out.
>>
>> On Wed, Feb 15, 2012 at 5:44 PM, Sergio Gonzalez <[email protected]>wrote:
>>
>>>  Hi Ramakrishna,
>>>
>>> Before sending a packet, do:
>>>
>>> ...
>>> call CC2420Config.setChannel(<channel>); // 11 or 26, etc.
>>> call CC2420Config.sync();
>>> ...
>>>
>>> Then, send the packet inside the synchDone callback afterwards:
>>>
>>> // Called after radio has tuned into a new channel
>>> event void CC2420Config.syncDone(error_t error)
>>> {
>>>     if(error == SUCCESS) call AMSender.send(..);
>>>     ...
>>> }
>>>
>>>
>>> Sergio Gonzalez
>>>
>>>
>>>  > Hi , This is Ramakrishna, working on tinyos-2.1.1. I am experimenting 
>>> on> dynamic channel switching for telosb mote. I am trying to send 2 
>>> packets> one in 26th channel and second in 11th channel. When i change the 
>>> channel> from 26 to 11 after sending the first packet in my program, first 
>>> packet> also being sent on 11th channel but not on 26th. why it is so? I 
>>> am> modifying the channel only after getting 'sendDone' event of the first> 
>>> packet. how it is overwriting the channel register with 11 before> 
>>> transmitting the first packet? Please help me in this regard.>
>>>
>>> It is my understanding that sendDone gets generated when the last byte has
>>> been transfered to the h/w.   It doesn't mean that the packet has acutally
>>> transmitted.
>>>
>>> So if you are changing the channel at that point you will probably cause
>>> problems for the back part of the packet.
>>>
>>> I'm not aware of an event being available from the h/w to indicate the
>>> complete TX has finished.  Not sure if  current h/w  makes that
>>> available....   You will need to stare at the radio docs.
>>>
>>>
>>> You could try putting a delay of somekind to see if that helps.
>>>
>>> As far as the channel register changing, I have no clue.
>>>
>>>
>>>
>>> --
>>> Eric B. Decker
>>> Senior (over 50 :-) Researcher
>>>
>>>
>>>
>>> _______________________________________________
>>> Tinyos-help mailing list
>>> [email protected]
>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>>>
>>
>>
>>
>> --
>> Eric B. Decker
>> Senior (over 50 :-) Researcher
>>
>>
>>
>
>
> --
> Eric B. Decker
> Senior (over 50 :-) Researcher
>
>
>


-- 
Eric B. Decker
Senior (over 50 :-) Researcher
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to