Reported by Nikolaus Schaller, Sep 6, 2014
should be 2.8V... To check: 1. rfkill unblock gps (basically enable LNA power) 2. cat </dev/ttyO1 (power up GPS receiver and LNA) 3. on separate shell (ssh into the device): cd /sys/bus/platform/devices/48070000.i2c:twl@48:regulator-vsim/regulato r/regulator.* cat state microvolts min_microvolts max_microvolts cat /sys/class/extcon/gps_antenna/state The result is that the VSIM regulator is enabled, min and max microvolts are as defined in omap3-gta04.dtsi - but the microvolts is just 1.8V (can also be measured on the ext antenna pin). gps_antenna state reports "internal" because 1.8V is not enough to power and switch to the external antenna. Consistent with these findings, dmesg tells: [ 2.353149] VSIM: 2800 <--> 3000 mV at 1800 mV So the question is why VSIM isn't outside of the instructed voltage range of 2800 ... 3000 mV.
Comment 1 by Nikolaus Schaller, Oct 22, 2014
Here is an explanation: if a TWL regulator minvoltage and maxvoltage specifies voltages that the LDO does not support at all, both are ignored and the LDO remains in some default voltage. We did have: min = 2800000uV max = 3150000uV This is not possible. But the LDO did not simply take 2.8V for min and max, but remains with 1.8V Here is the fix: http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=aebcd83c9b4fd 4441758aad955fe3fe9e8ab166b
Comment 2 by Nikolaus Schaller, Oct 24, 2015