Battery completely dead?

[mcobit]

I left my Pandora in suspend mode on SZ Beta4 and a half day later it shut down. Now, 2 days later, I put it at the usbcharger and it didn't come up again directly, but needed maybe 10 mins, before it booted. Then I was asked, if i want to keep 500 Mhz etc, but as soon as i click on OK, it shuts down completely again.


The charginglight does not light.


Maybe it doesn't charge at all and just runs on usbpower?


Do I need a new battery? The one I have in there was fine before and I could get a lot of runtime out of it.


Maybe the suspend mode did suck the battery completely empty without shutting down at the warning level and destroyed it this way?


If it is like this there should be a warning at least, that stuff like this can happen in suspend mode. I think it was never mentioned.


Edit: OK, after waiting for it to boot up and booting into the old kernel, with SD cards removed, wifi disabled and screen at lowest brightness, I was able to fully boot it and shortly after, the charging led came up, too.


The batteryindicater shows 0%.


Hope it will charge now. I'll leave it in low power mode for today to get some charge.


But I still think that suspendmode sucking the Battery completely dry is a problem.

 

[notaz]

But I still think that suspendmode sucking the Battery completely dry is a problem.

..which can't be fixed since CPU is off while suspended, so it can't check the battery and do full shutdown when needed.

 

[double7]

Is there another full discharge protection circuit?

 

[Binky]

I thought there was one in the battery itself

 

[Zink]

So if, by an accident, Pandora is left uncharged for a long time, and the protection circuit disconnects the battery, how can one charge the battery again (if someone has a Panodra that doesn't work on AC alone)? I think I've read once, that when put on AC and turned off Pandora will charge the battery, but very slowly. Is this true? If yes, that could be a solution in such situations.

 

[EvilDragon]

There is a circuit in the battery that prevents the total discharge.


As soon as the voltage goes below a critical level, it switches off (which is why the battery seems dead).


The Pandora needs a certain voltage to start up (more power is needed when you switch it on than when it's idling) which is why the Pandora won't boot up.


Charging while the Pandora is switched off works, but is very slow. For that reason, it might take 10-20 minutes before the Pandora comes back to life.


After that, the battery works fine again.


A completely dead battery won't start anymore, regardless how long you charge it.

 

[Zink]

Thanks for the explanation. And for clarification: by "completely dead" you mean broken, not just switched off by internal circuit?

 

[WizardStan]

You will almost certainly have to charge through the AC plug to get it past critical. 500ma is good enough for a start, and idle charging, but I don't think it'll cover it in actual usage, and if the battery is really so dead it takes minutes to start up then

..which can't be fixed since CPU is off while suspended, so it can't check the battery and do full shutdown when needed.

The TWL has a timer interrupt that can trigger a power on. I and someone else (who I'm forgetting the name of, sorry whoever you are) experimented with it months ago as a kind of alarm to turn on the Pandora at a specific time. In theory we could catch this interrupt as part of the un-suspend process and use it as a signal that it's just turning on to check the battery and then should go back to sleep, every hour or so should do it.

 

[mcobit]

We talked about that in irc and it would be a possibility to schedule the batteryckeck with the estimated runtime when going to suspend.

 

[gunni]

I seem to have a similar problem with my battery, I haven't used suspend at all though... my pandora powers on to the openpandora logo and seems to reset itself blanking the screen then returning to the logo. as with mcobit's problem the charging light does not light whether I try to use the charger or charge by usb. I tried leaving the pandora plugged in overnight last night but still the same problem. I had noticed previously that the charging light would go out before the battery was fully charged (last time was around the 60% mark)

 

[mcobit]

Try holding the right shoulderbutton, when itcomes on and select to boot into the 2.x kernel. Then turn backlight low and remove sdcards. It started and charged again forme.

 

[gunni]

Try holding the right shoulderbutton, when itcomes on and select to boot into the 2.x kernel. Then turn backlight low and remove sdcards. It started and charged again forme.

nope, can't even do that

 

[gunni]

http://youtu.be/FFdWVbtJza0


So this is what I've got....


sorry for the poor quality video.

 

[Mr_Loon]

@ gunni : Will the Pandora boot with the battery removed & the unit powered via the charger alone?

 

[gunni]

no, nothing happens at all plugging in the charger without the battery

 

[double7]

I know the vertical coloured stripes as signs of low power on my gp2x, especially visible at booting. That is strange. Next thing i would try is to prepare a firmware on SD card and see how that do. Maybe it's software related.

 

[mcobit]

That might as well be another hardwaredefekt.

 

[wdt]

Have you actually measured the voltage?


With a DVM? (digital voltmeter), on battery terminals


Otherwise, you are just guessing.


dave

 

[gunni]

thanks Dave I overlooked the obvious, battery voltage is fine according to my multimeter.


was gonna try reflashing an older firmware but it just resets itself before it can begin.


looks like my pandora is knackered... seriously regretting "upgrading" my preorder to get my no-wifi unit, its already been to ED once for a bust shoulder button