> > *Had I designed it, it would have had a graceful shutdown* > * procedure.....* >
Technically this is already in place. You just need a $2USD battery. But anyway, I've given this a lot of though, as my buddy forcing me to talk about this in great detail several times. Until he understood what I could do with an external MCU in software. For a long time, he though he would be able to get away with using an external watch dog. And then he though that we could do all this solely using the PMIC + BBB interrupts. Which is partially true. But the "cant reset because I'm stuck in an odd power state" issue never occurred to him. Something that I've personally experienced first hand, several times. Now, after we've talked A LOT, and about the various potential situations, we think we've got it 100% bullet proof covered. My own personal preference would be to have an external battery powering the 5V power in, and an external power supply charging that. With again, and MCU ( MSP430G2553 ) communicating back to the BBB. When it's time to shut down. Really similar to UPS. but my buddy does not like that, he says the cost is too hight. And he is right, it would be. On Mon, May 16, 2016 at 6:13 PM, Harvey White <ma...@dragonworks.info> wrote: > On Mon, 16 May 2016 16:34:28 -0700, you wrote: > > >> > >> *The bottom line seems to be that the BBB was not designed for this* > >> * kind of situation or application, and making it fit this requires* > >> * additional resources of some sort. Now the question comes down to* > >> * cost, utility, percentage of applications needing this, elegance of* > >> * design, and whether or not the hardware platform can cooperate in > this* > >> * or whether or not it simply lives in its own world.* > >> > >> > >> * Harvey* > >> > > > >I think the real bottom line is that the BBB *could* have been designed to > >do all this and more. At additional costs. As Gerald has stated many times > >on this group. Which I can completely understand. > > > > Exactly. Now this demonstrates the difference between a hobby project > (where cost is secondary) and a commercial product, where cost is a > goal, if not a limit. > > Had I designed it, it would have had a graceful shutdown > procedure..... > > 1) had I thought of it > 2) had I had the need for it > 3) had I figured out how to do it (after X tries, depending) > 4) had I decided it was important enough to do (bearing in mind 1 > through 3 above) > 5) fill in other limits as needed.... > > Second guessing a design is easy. However, it does show that any > design (like plans of war) rarely survives contact with the enemy..... > > Harvey > > > > > > > > > >On Mon, May 16, 2016 at 4:28 PM, Harvey White <ma...@dragonworks.info> > >wrote: > > > >> On Mon, 16 May 2016 15:45:14 -0700, you wrote: > >> > >> >You do not need anything connected to the beaglebone for any reason. > The > >> >beaglebone has an on die ADC that can detect if the AC mains is > powered or > >> >not. In which case, after a preset time period the Beaglebone could > shut > >> >its self down. > >> > >> True enough. The prevailing wisdom was going with an external device > >> having all the smarts about power failure, while the BBB was being > >> held up by batteries. > >> > >> The requirement that you propose is that the BBB have, somewhere, > >> access to power long enough to do a graceful shutdown. > >> > >> How this is done is left as an exercise for the student. > >> > >> > >> > > >> >Meanwhile, an external "device" can just switch off the input 5V to the > >> >beaglebone after a preset amount of time. Then once you have AC power > >> back, > >> >the "Device" simply turns the 5V back on. > >> > >> Yep, and with the same requirements of powering from either a battery, > >> a supercapacitor, or something more exotic. > >> > >> The bottom line seems to be that the BBB was not designed for this > >> kind of situation or application, and making it fit this requires > >> additional resources of some sort. Now the question comes down to > >> cost, utility, percentage of applications needing this, elegance of > >> design, and whether or not the hardware platform can cooperate in this > >> or whether or not it simply lives in its own world. > >> > >> > >> Harvey > >> > >> > >> > > >> >On Mon, May 16, 2016 at 3:09 PM, Harvey White <ma...@dragonworks.info> > >> >wrote: > >> > > >> >> On Mon, 16 May 2016 11:35:54 -0700 (PDT), you wrote: > >> >> > >> >> >Looks like nut been ported to Debian for the BBB. > >> >> > > >> >> >It and a smart UPS might be the easiest solution. > >> >> > > >> >> >I'm thinking along these lines, but haven't done anything with it > yet. > >> >> The > >> >> >nut client getting a signal over the network from my desktop is > kind of > >> >> >what I'm thinking. I've my BBW IOT app, router, and ISP interface > on > >> >> >a separate UPS that I want running as long as the battery lasts, > but a > >> >> >controlled shutdown of the BBW is something I'd like to add > eventually. > >> >> > > >> >> >The "shutdown if the power outage lasts longer than X" is pretty > easy, > >> >> >robust automatic start-up when the power returns might require a > >> smarter > >> >> >than the average UPS. > >> >> > >> >> I'd say that you want one that does automatic battery tests as well. > >> >> The one that I knew of at one time was a sine wave inverter. > >> >> > >> >> To summarize the types of inverters, there are two schemes. > >> >> > >> >> 1) keep a battery charged at all times. When power fails, detect the > >> >> loss of AC at the output. Start the inverter and switch that power > to > >> >> the output of the inverter. What happens is that power drops out for > >> >> the output with a power failure, and your equipment is supposed to > >> >> stay "up" for a certain amount of time (that the UPS takes to switch > >> >> on). Then the UPS takes up the load and life is good. > >> >> > >> >> 2) keep a battery charged at all times. Power the inverter from the > >> >> battery at all times. When the power fails, the battery charger > >> >> simply shuts down. > >> >> > >> >> The second one is the one I'd think you'd want to get. > >> >> > >> >> An opto isolator, driven by an AC bridge (or an AC style > optoisolator) > >> >> would give you a power failure indication within a half cycle. > >> >> > >> >> Harvey > >> >> > >> >> > >> >> > > >> >> >I'd be interested in success stories, but my experience with brand > name > >> >> >(APC) and off-brand UPS with desktop system is while they are better > >> than > >> >> >nothing, they aren't good at reporting battery issues and > ultimately I > >> >> end > >> >> >up with a power failure and "pull the plug" type shutdown because > the > >> UPS > >> >> >batteries can't support the switch over. We get a lot of 0.5 - 15 > >> minute > >> >> >power failures from thunderstorms here, so I'm sure the USP has > saved > >> me, > >> >> >but they are not foolproof. > >> >> > > >> >> > > >> >> >Ultimately I'm trying to sell the wife on a "whole house" natural > gas > >> >> >powered backup system so that a dumb UPS or battery with only a few > >> >> minutes > >> >> >run time to let the generator come on and switch over would be > needed. > >> >> >She was excited about it after Hurricane Ike, but now that its been > >> ~eight > >> >> >years, selective memory has her thinking we don't need it. > >> >> > >> >> -- > >> >> 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 beagleboard+unsubscr...@googlegroups.com. > >> >> To view this discussion on the web visit > >> >> > >> > https://groups.google.com/d/msgid/beagleboard/apfkjbtvsk9h94q0qvfpke40ofhbc7c22h%404ax.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 beagleboard+unsubscr...@googlegroups.com. > >> To view this discussion on the web visit > >> > https://groups.google.com/d/msgid/beagleboard/silkjbt8elht0qlo1536jduujmhaekcobc%404ax.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 beagleboard+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/beagleboard/mmrkjbh6nci8f7av55dfb68c36fb1c8r53%404ax.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 beagleboard+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/CALHSORoCcteqA5SPc-PgrQxg698Jw6u08Xo2JM8LfOjTvkhjnA%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
