Isn't Wifi supposed to be off by default?

Werner Almesberger werner at openmoko.org
Thu Jan 15 21:37:06 CET 2009


Andy Green wrote:
> That'd be worthwhile, great.  I think someone looked at this in .tw
> about a year ago but I never really read about what they found.

Didn't you do some suspend current measurements early last year ?
Do you remember how far down you could go ?

> Glamo is also permanently powered anyway, but it doesn't take 145 mA :-)

Naw, when I throw a crowbar in its general direction, I get up to
40mA. (The crowbar being  poke 0x48000000 0x22119190  to ignore
nWAIT, then  gpio j5=0  to hold the Glamo in reset.)

> I would guess it is either WLAN or GSM, maybe pop the SIM if it is still
> in there.

No SIM. There's a uSD card, but that one seems to behave. WLAN
should be peacefully asleep, but I'll pop it, just to be sure.
My main suspect is GSM. We'll see.

> Balaji had auditing regulators that were still up on his list.

Good ! With IDBG, I should be able to get a PMU register dump
while suspended. That should be useful to check if the real
state is the same as expected. Let me try ...

Here's what I got:

Operation	Output			Current (min avg max stddev)
---------------	-----------------------	-----------------------------------
# hard reset
					 -0.14mA  -0.10mA  -0.07mA  0.021mA
# cut USB power
pmu 0x49=3
					300.72mA 300.89mA 300.98mA  0.075mA
# screen blanker
echo 1 >/sys/class/graphics/fb0/blank
					224.57mA 224.96mA 225.70mA  0.300mA
# HCLD: uSD
pmu 0x3a 0x3a=0
		0x00
					224.64mA 225.00mA 225.82mA  0.353mA
# LDO1: accel
pmu 0x2e 0x2e=0
		0x02
					225.66mA 226.02mA 226.67mA  0.278mA
# LDO2: codec
pmu 0x30 0x30=0
		0x02
					226.34mA 226.70mA 227.58mA  0.339mA
# LDO4: BT
echo 0 >/sys/class/rfkill/rfkill1/state
pmu 0x34
		0x00
					218.87mA 219.30mA 220.15mA  0.359mA
# LDO5: RF (GPS)
pmu 0x36 0x36=0
		0x01
					176.70mA 176.77mA 176.98mA  0.079mA
# LDO6: LCM
pmu 0x38 0x38=0
		0x01
					165.14mA 165.38mA 165.63mA  0.145mA
# LEDOUT
pmu 0x29 0x29=0
		0x00
					164.98mA 165.27mA 165.84mA  0.230mA
# Glamo
poke 0x48000000 0x22119190
gpio j2 j2=0
		>1
					165.28mA 165.39mA 165.58mA  0.103mA
# WLAN reset
gpio j12 j12=0
		>1
					165.19mA 165.59mA 166.28mA  0.294mA
# GSM AMP
pmu 0x15 0x15=0
		0x00
					165.34mA 165.50mA 165.80mA  0.144mA
# GSM reset
gpio b5 b5=1
					165.24mA 165.40mA 165.55mA  0.102mA
# suspend
echo mem >/sys/power/state
					103.22mA 103.26mA 103.31mA  0.024mA


Note that some things are no-ops. Also, blanking the screen seems to
have shut down the Glamo pretty decently. Not sure if GSM reset does
anything in this particular machine. For comparison:

# hard reset
# cut USB power
pmu 0x49=3
# suspend
echo mem >/sys/power/state
					143.09mA 143.42mA 143.90mA  0.283mA
# disconnect debug board
					143.05mA 143.61mA 144.44mA  0.480mA

The utilities I used (gpio, pmu, poke) live here:
http://svn.openmoko.org/trunk/src/target/gpio/
http://svn.openmoko.org/developers/werner/pmu/
http://svn.openmoko.org/developers/werner/poke/

- Werner



More information about the openmoko-kernel mailing list