Hi Miklos,

thanks for pointing this out. I did not simulate errata and ignored them
therefore - Until now.
I will have to think about how i can handle this very special behaviour
in my simulation. Silicon bugs suck.

Regards,
Markus

Am 21.03.2012 23:28, schrieb Miklos Maroti:
> Hi Markus,
> 
> On page 89 of the datasheet there is a workaround that shows that you
> can switch between TRX_CMD=TX_START the to TRX_CMD=RX_ON and then to
> TRX_CMD=TX_ARET_ON. So even in the manual they use this (at lease in
> one of the published workarounds of an errata). I do not know how
> slower the "proper" way would be. Yes, this might be "undocumented",
> from all info I saw RX_AACK_ON is pretty much the same as RX_ON and
> TX_ARET_ON is pretty much the same as PLL_ON, and between RX_ON and
> PLL_ON you can switch, so this was a natural port of that behavior and
> it worked. If it makes you more confortable, I can happy to add a
> comment there :)
> 
> Miklos
> 
> On Wed, Mar 21, 2012 at 5:11 PM, Markus Jung
> <[email protected]> wrote:
>> Hello,
>>
>> hardware-acknowledgements have been reported to work on real devices
>> (AFAIK), but while trying to simulate a TOS application using HWACK with
>> LPL, i stumbled upon following behaviour:
>> Phase A: The driver puts the radio into RX_AACK_ON mode. All is fine
>> until
>> Phase B: A transmission gets started. The driver issues a direct state
>> transition from RX_AACK_ON to TX_ARET_ON, which is not defined either by
>> the datasheet or the software programming manual. It is quite
>> interesting to observe real device accepting this (undocumented) state
>> transition.
>>
>> This behaviour has been observed in simulation protocols and can be
>> confirmed by the code in RF230DriverHwAckP (RadioSend.send).
>>
>> However: AFAIK, there is no official source which mentions a transition
>> RX_AACK_ON<->TX_ARET_ON. The only documented way is to use PLL_ON (or,
>> not recommended according to Atmels software programming manual,
>> TRX_OFF) as a intermediate-state.
>> For stability and compatibility reasons, the driver should not use
>> undocumented state transitions.
>>
>> Regards,
>> Markus
>> _______________________________________________
>> Tinyos-help mailing list
>> [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