asterix: scrub watchdog and PMIC settings [FIRM-82] #182
Conversation
| { BUCK_BASE, BUCKENCTRL, 0 }, | ||
| { BUCK_BASE, BUCKVRETCTRL, 0 }, | ||
| { BUCK_BASE, BUCKPWMCTRL, 0 }, | ||
| { BUCK_BASE, BUCKSWCTRLSET, BUCKSWCTRLSET_BUCK1SWCTRLSET | BUCKSWCTRLSET_BUCK2SWCTRLSET /* use registers rather than resistor settings */ }, |
There was a problem hiding this comment.
note: vset resistors will always force nominal voltages (1.8 and 3.0V), so it may be unsafe to allow sw controlled values here.
Also, as a general comment: holding the back button for 10s will reset the PMIC, so we have a known path to go back to a safe state in case things go terribly wrong.
There was a problem hiding this comment.
Well, it will reset the PMIC as long as LPRESETCFG is set properly!
But I would like to explicitly set these values, because 1) analog straps always kind of scare me, and 2) when we set them explicitly, the behavior changes in at least some way (!) vs. when we leave them implicitly set from Rset.
| { ADC_BASE, ADCNTCRSEL, ADCNTCRSEL_10K }, | ||
| { CHARGER_BASE, BCHGVTERM, BCHGVTERMNORM_4V20 }, | ||
| { CHARGER_BASE, BCHGVTERMR, BCHGVTERMREDUCED_4V00 }, | ||
| { CHARGER_BASE, BCHGISETMSB, 38 }, |
There was a problem hiding this comment.
0.5C may still be a safer limit here, considering discharge can be adjusted independently
There was a problem hiding this comment.
Really, 1C will be fine. (Actually, 1C will be too slow for most users and we will want to turn it up even further in firmware.) But 1C is definitely safe for the bootloader to do.
| // GPIO0 is nPM_RST input (from debug header). Sadly for us, that's not | ||
| // how nPM1300 works: you can't wire GPIO0 to be a reset, you can only | ||
| // wire it to be part of a two-finger reset. So we just set it up to | ||
| // generate an event on falling edge, and I guess nRF can poll for it, | ||
| // assuming it doesn't otherwise get reset on its own. | ||
| { GPIOS_BASE, GPIOMODE0, GPIOMODE_GPIEVENTFALL }, | ||
| { GPIOS_BASE, GPIOPUEN0, 1 }, | ||
|
|
||
| // GPIO1 is nPM interrupt output to nRF (but nRF must enable PU if it wants it, to save on Iddq) | ||
| { GPIOS_BASE, GPIOMODE1, GPIOMODE_GPOIRQ }, | ||
| { GPIOS_BASE, GPIOOPENDRAIN0, 1 }, |
There was a problem hiding this comment.
probably irrelevant as bootloader does not handle any PMIC event. firmware can take care of this
There was a problem hiding this comment.
Agreed, left everything as an input.
| { BUCK_BASE, BUCK1NORMVOUT, BUCKVOUT_1V8 }, | ||
| { BUCK_BASE, BUCK1RETVOUT, BUCKVOUT_1V8 }, | ||
| { BUCK_BASE, BUCK2NORMVOUT, BUCKVOUT_3V0 }, | ||
| { BUCK_BASE, BUCK2RETVOUT, BUCKVOUT_3V0 }, |
jwise
force-pushed
the
jwise/asterix-nrf-watchdog-scrub
branch
from
2f3526e to
1cff5a2
Compare
|
@gmarull Updated some -- have a look? |
jwise
force-pushed
the
jwise/asterix-nrf-watchdog-scrub
branch
from
1cff5a2 to
f20e50e
Compare
|
blurgh that gitlint is really cramping my style. Updated again to make it happier. |
|
Yep I will do that in the morning, if everything else is ok with you you can approve-pending and I won't merge until I do |
jwise
force-pushed
the
jwise/asterix-nrf-watchdog-scrub
branch
from
f20e50e to
0974ee4
Compare
Signed-off-by: Joshua Wise <[email protected]>
Signed-off-by: Joshua Wise <[email protected]>
Signed-off-by: Joshua Wise <[email protected]>
jwise
force-pushed
the
jwise/asterix-nrf-watchdog-scrub
branch
from
0974ee4 to
bc9a831
Compare
No description provided.