Open GoogleCodeExporter opened 9 years ago
Can we get the blessing of the administration for including this in core in v
17 if
some coders work on it?
Original comment by pbad...@verizon.net
on 13 Jun 2009 at 2:50
It would be cool if this function came in two forms:
* sleep(time) - sleep for a given period of time (seconds? milliseconds?)
* sleep() - sleep until an external interrupt (allows a deeper sleep mode)
Original comment by dmel...@gmail.com
on 13 Jun 2009 at 3:31
This would be awesome, one of the thing I would really like to see. This and
perhaps
the ability to run at 1mhz, for real low power use
Original comment by aristotl...@gmail.com
on 4 Aug 2009 at 8:47
Aristotle, is their an issue open here for the 1mhz thing? If not, consider
adding
one. :-) Better to track things like that individually.
Original comment by josiah.ritchie
on 4 Aug 2009 at 8:50
that would be a nice thing! hope it makes it into the next release :)
Original comment by designer2k2
on 7 Aug 2009 at 10:34
Aristotle... this changes clock to 1MHz at runtime.
CLKPR = (1<<CLKPCE);
CLKPR = B00000011;
taken from http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1163418637 although
syntax
for binary constants needed changing.
Original comment by pixelfre...@gmail.com
on 5 Sep 2009 at 12:36
[deleted comment]
Original comment by dmel...@gmail.com
on 6 May 2010 at 6:40
To get significant power savings (i.e., to go from 3 mA to less than 3 uA!),
you must use the "power-save" mode of the AVR. However, to get a timed wakeup
from power-save mode, you need timer/counter 2 to be clocked asynchronously
from a 32.768 kHz external crystal.
A millisecond-precision timed deep sleep function would be extremely helpful.
Unfortunately you can't use the external crystal as the system clock source
while also using a 32.768 kHz watch crystal for the low frequency timer/counter
2 asynchronous mode clock, so do serial communications with the UART is
difficult then, since the internal RC oscillator is not stable enough for
reliable UART communication (unless continuous runtime calibration is
performed).
How can significant power savings into the micro-amp range be achieved with the
ATmega328P's capabilities, and specifically the current Arduino hardware
platform that has no 32.768 kHz crystal?
Original comment by sky...@gmail.com
on 16 Jun 2010 at 12:08
For example an ATmega168 @16Mhz uses 10mA in active mode, and 2mA in idle mode.
So support for an idle sleep would be better than nothing.
Original comment by ldro...@gmail.com
on 28 Sep 2010 at 7:15
Original comment by dmel...@gmail.com
on 20 Dec 2010 at 3:25
I think some hardware tweak would be great in addition to sleep related
software funcions. A jumper or configuration to disconnect the USB in the
Arduino board circuitry if not needed, saving energy.
I've measured the current flowing in a simple sketch and got more than 100mA
with all pins in input mode.Even if the avr was in deep sleep, probably the
solution still would not be suitable for projects where low / ultra low
consumption are mandatoy, because of the extra circuitry drain.
With this tweak, we could do this kind of projet without installing the AVR in
a custom board.
Original comment by ampbra...@gmail.com
on 25 Jan 2012 at 8:49
I did something very similar in the design of the Mosquino board (an Arduino
variant specifically designed for low power operation) - the USB-serial
converter is powered from Vusb rather than Vcc. It's that easy :-) The board
uses the old FT232 converter, but the same should work on the Uno/etc.'s
USB-serial solution. With this change and a more efficient LDO, sleep power
consumption on the board is a few uA (most of that is due to a hardware RTC and
other onboard features).
The current schematic for this board is at
http://code.google.com/p/mosquino/source/browse/?repo=hardware#hg%2Fmain
CAVEAT: In the unpowered state (USB unplugged), driving the TX pin into the
FT232 high will still consume current. This is probably true with the Uno
converter too. For lowest power consumption with this fix, the user code should
disable Serial when there is no PC attached.
Original comment by drmn...@gmail.com
on 5 Apr 2012 at 2:51
Original issue reported on code.google.com by
josiah.ritchie
on 8 Jun 2009 at 2:41