[Tinyos-devel] current draw in deep sleep for telosb

[Vlado Handziski]

Indeed. Some transient instability is to be expected, due to the missing of
pull-up and pull-down resistors for the clock and the chip select lines of
the flash during the MCU reset, like page 10 of the flash manual suggests,
but I am not sure why the pin is left floating on the flash side after PD.

If this is really what is happening, then the proper state for the MCU pin
after the flash is in PD would indeed be the output mode 0, as for all
unused (like in not wired) pins. We should do some more tests to see how the
consumption behaves with the pin left in output and then we can change the
Init if necessary.

Vlado

On 7/1/07, Prabal Dutta <prabal at cs.berkeley.edu> wrote:
>
> The line's probably floating and you're probably acting like an
> antenna and picking up 60 Hz AC.  This might be causing the input
> pin's FET to cycle, drawing power as the gate capacitance charges and
> discharges.  The solution is usually to either make it an output or
> pull it up or down.
>
> - Prabal
>
> On 6/30/07, Razvan Musaloiu-E. <razvanm at cs.jhu.edu> wrote:
> > Hi!
> >
> > On Sat, 30 Jun 2007, Vlado Handziski wrote:
> >
> > > I suspect a HW problem on your testing node. Have you tried with
> another
> > > one?
> >
> > I tried with several of the tmote skys I have around. The effect is most
> > visible (and persistent) on the ones which have the 10-pin connector
> > populated with pins. When I touch the I2C Clock pin the current changes.
> >
> > I was able to replicate the problem also on a mote without pins by
> > touching with my finger the 10-pin connector on the area of the I2C
> Clock
> > hole. The increase in this case was only temporary though. :P
> >
> > I just checked and the problem is also reproducible in XubunTOS using
> the
> > T2 version 2.0.1.
> >
> > > The content of MotePlatformC has been stable for a long time and Null
> > > consumes ~ 7-8uA on a sensor populated node without any modifications
> to
> > the
> > > code.
> >
> > I think it looks stable because on a mote without pins on the 10-pin
> > connector the situation is stable unless you start touching it.
> >
> > > In general, when not used, the pins on the msp430 are put in input to
> > > reduce the leakage currents.
> >
> > My hardware knowledge is limited so I cannot really argue why puling
> that
> > pin to zero helps. I just experimentally noticed that it does. :-)
> >
> > --
> > Razvan ME
> >
> > > On 6/30/07, Razvan Musaloiu-E. <razvanm at cs.jhu.edu> wrote:
> > >>
> > >> Hi!
> > >>
> > >> I'm running the latest CVS and today I noticed that the Null
> application
> > on
> > >> telosb draws an unusually high (and variable) amount current (around
> > >> 0.5-0.8mA). The problem seems to be in in Init.init function from
> > >> MotePlatformC.nc from tos/platforms/telosb/. At the end of it the
> > >> TOSH_FLASH_M25P_DP function the clock pin is put on input. By making
> it
> > >> output on zero the current stabilized to the normal low values
> (7-8uA). I
> > >> attached the patch with the changes. It can be aplied by running the
> > >> following commands:
> > >>         cd $TOSROOT
> > >>         patch -p0 < telosb-current.diff
> > >>
> > >> Can somebody confirm the problem and test this solution?
> > >>
> > >> Thank you!
> > >> Razvan ME
> > >> _______________________________________________
> > >> Tinyos-devel mailing list
> > >> Tinyos-devel at Millennium.Berkeley.EDU
> > >>
> >
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-devel
> > >>
> > >>
> > >>
> > >
> > _______________________________________________
> > Tinyos-devel mailing list
> > Tinyos-devel at Millennium.Berkeley.EDU
> >
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-devel
> >
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.millennium.berkeley.edu/pipermail/tinyos-devel/attachments/20070701/1c325da8/attachment.htm