[Tinyos-help] I need some info on Power Management support

Santosh Kumar kumars at cis.ohio-state.edu
Sun Jan 11 11:13:45 PST 2004


This was great piece of information. Thank for you very much.

5. On the issue of using the power down mode in place of power save mode,
does snooze() work with power down mode? If it does not work with the power
down mode (may be because the power down mode does not have an asynchronous
timer running) now, do you know if there is a plan to use power down mode in
snooze() in the next 6 months?

6. Is there another primitive or function (other than snooze()) that can be
used to put the mote in power down mode and then wake it up after some time?

7. One short question: Am I correct in assuming that maximum time that a
mote can be put in sleep at a time using the snooze() operation is 34
minutes (2^16/32 = 2048 seconds)?

Thanks again for your prompt and insightful response.

Have a nice day,
Santosh Kumar,
Ohio State University

----- Original Message ----- 
From: <kwright at EECS.Berkeley.EDU>
To: "Santosh Kumar" <kumars at cis.ohio-state.edu>
Cc: <tinyos-help at Millennium.Berkeley.EDU>
Sent: Saturday, January 10, 2004 5:38 PM
Subject: Re: [Tinyos-help] I need some info on Power Management support


>
> (3) The GDI (Great Duck Island -- field research that the TinyOS group has
done in conjunction with the College of the Atlantic for the past two
summers -- see the related work page) application achieved about 10uA in the
field in the power save mode. The code is in contrib/ucb/TestLabApp. Talking
to the GDI folks, they said that a crucial component was disabling the JTAG
interface which seemed to consum about 50uA regardless of the power mode.
>
> A note about computing your power consumption: the timer process ends up
taking on average about 20uA when you take into account that the timer
interrupt is serviced every 220ms (using TimerM).
>
> Now back to (1) and (2). Short answer: the wakeup latency is about 4.5 ms.
I've attached a graph of the the power consumption during a timer interrupt.
As you can see, the baseline is about 10uA -- that's the mote in power save
mode. The timer interrupt happens at time 0. For approximately 4 ms, the
mote consumes about 600 uA while the oscillators stabilize. Once stabilized,
the mote resumes normal operation. In this case, the timer interrupt is
serviced and then the mote goes back into power save mode.
>
> (4) No, you can't actually power the mote off completely. Looking at the
datasheet, all the accessible modes leave at least something on so that the
mote can be awakened. Now, somebody here told me that power down mode gets
down to < 10uA, maybe even 1uA -- comparable to self-discharge rates of AA
batteries. Maybe that will work for you. As a bonus, you can wake it up :) .
>
> I hope this helps.
>
> -kw
>
>
>
>
> (3)
>


----------------------------------------------------------------------------
----


> _______________________________________________
> Tinyos-help mailing list
> Tinyos-help at Millennium.Berkeley.EDU
> http://mail.Millennium.Berkeley.EDU/mailman/listinfo/tinyos-help
>



More information about the Tinyos-help mailing list