Temporarily pressing the power button (or pulling the PWR_BUT line to
ground) does shutdown the system but it seems to take much longer, often
more than a minute. I have put together a small PIC and a super-capacitor
setup; the PIC monitors the main power line and triggers the shutdown and
the capacitors can power the BBB for about 15 to 20 seconds so shutting
down fast is rather important. Once the system is down (as evident by the
VDD_5V going to 0V), the PIC turns off the main regulator and it then waits
for main power to come up and the capacitors to charge before it enables
the regulator again which will boot the BBB next time power arrives.
Using the service to shutdown or issuing a shutdown -h now' takes 8 seconds
to power down most of the time. Sometimes it takes long, but rarely does
it seem to take more than 15 seconds. Pressing the power button
consistently takes over 45 seconds to over a minute. I have not
experimented enough to feel extremely confident on all of this, but the
shutdown command appears to consistently be faster than the power button.
Maybe there is a way to make that shutdown faster, but I have not found
anything obvious.
Regarding talking to the power management chip, I am resisting the urge for
a couple of reasons - it will take time away from working on the design of
my system and even if I were to invest the time, it would make my design
more BBB specific; right now, the setup I'm using can work with just about
any linux system with a single input pin and an exposed power rail which
can be used as feedback for when the shutdown is done.
If there is a straightforward way to do the same without using a service
process, I would welcome the instructions, but I just don't know enough
about it right now.
On Thursday, June 2, 2016 at 7:59:23 PM UTC-4, Wulf Man wrote:
>
> any reason you cant just use the pwr_butt input and use the apci driver ?
>
>
>
> On 6/2/2016 3:28 PM, YP66 wrote:
>
> I am working on an automatic shutdown based on an external signal (through
> a digital IO line), on a BB black, running Debian Jessie.
>
> I build a program that monitors the digital line and when that line goes
> high, the program calls
> system("/sbin/shutdown -h now");
>
> When I run that program interactively, it takes anywhere between 6 to 10
> seconds to shutdown following the digital line flip. I use an LED to tell
> me when the process detects the bit flip so I can reliably count the time
> to shutdown. The actual shutdown is evident because all the lights go off.
> At that point I have to recycle power to start again, which is exactly
> what I want.
>
> However, when I put the same program to run as a service (by adding a
> service in /lib/systemd/ and using systemctl to start it), the system takes
> over 30 seconds to shutdown, sometimes a minute or two. I know the program
> is running as a service because I can see the LED I am using as feedback.
> The LED changes blink rate when the 'service' detects the GPIO change so I
> can tell the program is getting to the point of making the system call, it
> just takes so much longer. In fact, initially I thought it did not work,
> only to have the system shutdown on my suddenly a couple of minutes later.
>
> So the situation is that a program running directly goes through a
> shutdown almost instantaneously but when run through a service, it takes
> significantly longer. Any idea why?
>
> By the way, I have also tried
>
> sync();
> reboot(LINUX_REBOOT_CMD_POWER_OFF);
>
> instad of system("/sbin/shutdown -h now'), and I get more or less the same
> behavior.
>
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups
> "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected] <javascript:>.
> To view this discussion on the web visit
> <https://groups.google.com/d/msgid/beagleboard/501fb3a7-bdae-4b3e-96ad-3df2a8c67d0e%40googlegroups.com?utm_medium=email&utm_source=footer>
> https://groups.google.com/d/msgid/beagleboard/501fb3a7-bdae-4b3e-96ad-3df2a8c67d0e%40googlegroups.com
> .
> For more options, visit https://groups.google.com/d/optout.
>
>
>
--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/beagleboard/9e02c110-f5bd-4a13-9b4c-d18d06835166%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
