Hi Hugo, I see what you're saying. The oscillator, voltage regulator, etc. are off inside the radio chip, but the chip itself is only in a deep-sleep mode. The chip itself is technically still consuming a slight amount of power - on the order of a few microamps.
The on-time of the radio fluctuates a little bit based on the presence of other transmitters. When no other transmitters are present, your 2% duty cycle will really be approximately 2%. But if there are other transmitters, the radio may have to spend a little more time on to determine whether or not those packets are intended to be received. One way to ensure the radio is not on while performing ADC measurements is just to turn it off - call SplitControl.stop() and then the duty cycling stops until you tell it to turn back on. I haven't verified this type of usage before, so if you try to do that we'll need to ensure potentially rapid calls to start() and stop() do not actually increase the duty cycle of the radio. When transmitting an LPL packet to the broadcast address, the transmitter will continually retransmit the packet for the entire sleep duration of the receiver node(s). It will consume more power, but you can be most likely sure (RF propagation aside) that all receivers in the vicinity will have received the packet, even though their wake-up schedules don't align. -David -----Original Message----- From: Hugo Sousa [mailto:[EMAIL PROTECTED] Sent: Monday, June 04, 2007 2:59 PM To: David Moss; [email protected] Subject: Re: [Tinyos-help] Questions about low power listening in CC2420 But the MTS300 isn't independent from the radio, they share a pin (Port E6/FIFOP/Int2), which turns on the temp sensor. Reading the photo or temp sensor (which in turn share the same ADC line) does not provides accurate values while the radio is on. I tried using low power listening and the reading problem was still there that's why I asked if the radio was completely off or just some part of it. The duty cycle I used was just 2% (so 2% was also the probability that the reading was done with the radio on). Also, what is the period of the duty cycle? For example, if I set the duty cycle to 2%, is it on for 20ms and off for 980ms, or on for 10ms and off for 490ms, or something else? A final question, about the "synchronization": how does it work with Dissemination? If Dissemination uses broadcast to send messages, no ACKs used, there is no way to know if some node got the the message. Thanks in advance Hugo Sousa On 6/4/07, David Moss <[EMAIL PROTECTED]> wrote: > TEP 126: http://www.tinyos.net/tinyos-2.x/doc/txt/tep126.txt > > The current implementation is asynchronous low power listening - there's no > synchronization going on. The packets get retransmitted until the receiver > wakes up and detects the transmission. > > The MTS300 is independent of the radio. The radio goes into power down > mode, where that chip is off. Your node's microcontroller SPI bus and > everything else is still operational as needed by other components. > > -David > > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Hugo Sousa > Sent: Monday, June 04, 2007 5:54 AM > To: [email protected] > Subject: [Tinyos-help] Questions about low power listening in CC2420 > > Where can i read a bit about tinyos' implementation of the radio low > power listening mode? > > How does it handle the > packet-loss-while-sleeping/radio-wakeup-synchronization issue with the > other nodes? > > What part of the radio exactly does it turn off? Because it seems it > still interferes with the temp and photo sensor in the mts300 > sensorboard (because they share a pin) > > > thanks in advance > > Hugo Sousa > _______________________________________________ > Tinyos-help mailing list > [email protected] > https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > > > _______________________________________________ Tinyos-help mailing list [email protected] https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
