Pandora Won't Turn On

[javaJake]

I'm really worried right now. I did a search of the forums and can't find an answer.

So, my Pandora won't turn on. I left it, lid closed, running on a blank XFCE desktop. I opened it up, and the screen was on but black. The only LED on is the power LED. I removed the battery, since Pandora+Power wasn't resetting it. When I put the battery back in, it wouldn't turn on. I plugged it into the charger, and it still wouldn't turn on.

I had to leave to meet up with friends, and left my Pandora on the charge cable. I came back three hours or so later and the Pandora had turned itself on. The charge and power LEDs were on now but the screen was still black. I let it keep charging for a while longer. After another two or so hours, the screen got two vertical white racing-like stripes down at about 1/3 across the screen from the left.

Thinking the Pandora must've had enough power to boot up and corrupt the display, I removed the battery and put it back in (since, again, Pandora+Power wasn't resetting). Still wouldn't turn back on. I tried all kinds of key combinations, and there was still nothing.

I left the battery out for five minutes, and came back. I put the battery in. Still nothing.

It's sitting on the charger now. I figured I'd let it charge for another 8 hours and come back tomorrow morning. That'll amount to about 15 or so hours charging in this state.

Interestingly enough, the first time I left it to charge, when I came back, it felt warm, like it usually does when it's powered on.

Some possibly related facts: it's been 90 degrees (F) today, 60% humid. I didn't leave it in the sun.

Thanks for reading.

Edit: using links2 for the moment, and it won't let me edit the subtitle of this post. I shouldn't say "prior" since it really happened "after" "afaict".

 

[Pleng]

Have you tried Fn-I? It's possible your display brightness has been set to zero. Although you'd still expect to see the initial Angstrom boot screen before after a reset before it goes blank if this is the issue.

 

[mali]

You should add more info, for example which OS version is installed(original, Hotfix 1,2, 3 beta). Sometimes increasing brightness via keyboard shortcut helps, but a corrupted display showing some lines indicates that the backlight is working. I suspect a corrupted firmware.

An important note: Never plug in the battery while the charger cable is attached.

 

[javaJake]

Oh my gosh.

Well, it turned on and booted by itself, like it should when charged. I used the Pandora for a bit to Google around for VPN plugins.

And then, all of a sudden, it crashed. The Wifi light died, but the charge and power LED's remained. The screen immediately went into a pink-black-red-etc pixel stripe pattern, then slowly faded (literally) to black and white, then when I unplugged the charging cable it faded back to the pink coloring. I pressed the reset combination (Pandora+Power) before I unplugged the cable, but it didn't actually reboot until a good 1-2 seconds after the cable came out.

Then I got to see it boot, and I think I know why the screen was black before. When it boots now, it goes to the Pandora logo, then black, and that's it. I'd say a simple reflash would do the trick, but it also does really odd crashes, like the one I described. Is this a corrupted rootfs issue?

Goshdarnit, I really hope I don't have to send this overseas for repairs. :/

 

[javaJake]

Thanks for the tips. I'll try reflashing (if I get that far).

What happens if I plug the battery in while the charge cable is attached? I've never actually done this, I don't think, but I'm curious.

 

[javaJake]

Oh, forgot to add: running Hotfix 2 + Codec. Brightness is set to lowest, but is definitely on.

 

[mali]

When no battery is found on AC adapter insertion, the unit goes into constant voltage mode which provides a constant 4.0V on the battery net. It doesn't like it if you put a battery in at this point because the TI chip doesn't appear to be designed to jump back to normal mode without a power cycle. It could do weird things like dump ridiculous amounts of current into the battery without software control. I liked the idea of letting customers power without a battery so we left this feature in but be aware of this restriction:

IF YOU POWERED THE SYSTEM WITHOUT A BATTERY, SHUT IT OFF BEFORE STICKING THE BATTERY BACK IN. (not an angry sentence, just making it clear.)

When the battery is inserted first, the system will power up using the normal mode. When the AC adapter is plugged in after this, it will power the system and charge the battery up to the limits set in software. Currently, Notaz and I have chosen 1A on the AC adapter and 600mA from the USB input. It can be changed by sending commands to the chip to alter these set numbers but that is a software lesson better learned from someone else. With the LCD off (lid closed perhaps) idling on the desktop, something like 750mA will go to the battery for a 6 hour charge time or so. With the system doing stuff and demanding more power, less current will go to the battery and it will charge slower. 1A was also chosen to consider thermal dissipation because the regulator of the TI chip is linear and wastes energy dropping the voltage from the 5V down to just above the cell voltage. You may notice the system feels warmer when the battery is more discharged than when it is full which is normal.

I haven't had much luck hot swapping batteries while the system is live but I think it worked for ED once. For me, as soon as I pull the battery, the system resets. The power management chip seems to rely on the status of the thermistor in the battery and it if disappears, it can trigger events. Maybe there are software work arounds but I don't know.

It's best to reflash and install hotfix 2 to have a clean base for further observations. I suggest you should run it without charger attached for a while then and check battery drain behaviour.

You should also power it off completely(from menu) when not in use for a longer period of time and you should put it in low power mode while charging(flip the power switch).

 

[tsh]

Another thing to do is when it is running OK, and the battery is more than 20% charged, un-plug the charger, and shutdown then turn it on again. That might give some clues.

 

[javaJake]

Hmm, whenever I reflash, I get a "bad CRC" message on boot.scr.

Running fsck on the card to see what's up.

Edit: nope, everything checks out. The Pandora refuses to get by that CRC error. :/

 

[Lobo]

Did you try re-downloading the hotfix? Maybe it's the actual file that's corrupted, not the SD card.

 

[javaJake]

I reformatted as FAT32, and it liked that. For some reason ext2 wouldn't work.

So, I reflashed fine. I rebooted it, and it turned on. I let it sit there, lid open, to do its first boot. I walked away and did other things for an hour or so and came back, only to find out it had crashed again. I can't Pandora+Power reset, and the charge and power LEDs remain on even after I unplugged the cable.

Some other patterns: if I force it off by removing the battery, and then put the battery back in, and try to turn it on, it won't turn on. But if I let it sit for a while, at some random point in time in the future it turns itself on. While it's on, it sometimes does random reboots before it reaches the Angstrom OS, and sometimes on the reset there's graphical display glitch on the OpenPandora logo, where the logo is outlined in some areas with white pixels. If it does reach the Angstrom OS, and I open and close the lid, or I log in and try to do anything, the screen immediately blanks (but remains on) as I had described in my other posts.

Should I call tomorrow to get an RMA number?

(This all started when I (think I) let it run out of battery life, turned on, lid closed. Has anyone else done the same thing and been fine?)

 

[mali]

^ One last thing you could try, is powering it from mains alone without battery attached. I remember someone else had a similar problem and I will search for that thread.

edit:
http://www.gp32x.de/board/index.php?/topic/54376-openpandora-wont-boot-after-battery-discharge/

 

[javaJake]

Alright, this is odd. A family member picked it up, opened the lid, and maybe hit the power switch, and it turned right back on. Now it's running fine. It did the first boot wizard, I'm in, it's puttering along happily.

:blink:

I don't get it... at all...

 

[javaJake]

^ One last thing you could try, is powering it from mains alone without battery attached. I remember someone else had a similar problem and I will search for that thread.

edit:
http://www.gp32x.de...tery-discharge/

Hmm, sounds exactly like what happened to me. Hardware bug perhaps?

Oh, and while I was writing this, the unit rebooted itself again. Edit: I didn't touch it, so it idled at the login screen. It rebooted yet again. Oh, and now it did it again in the middle of booting Angstrom.Another edit: And again... I'm going to stop counting now, and just remove the battery. I'm done watching this thing writhe on my desk.

 

[mali]

^ Try running it without battery then. Maybe the battery or its gauge is damaged, who knows.

 

[javaJake]

Since I removed the battery, the unit is off now. I put the battery in its special case, and replaced the battery cover.

It's plugged in now, and left the lid open. I tried to turn it on, first via the power switch, then Pandora+Power, but, as usual, it seems to want to boot itself whenever it pleases. I'm going to let it sit for a bit and see if it turns itself on.

 

[mali]

^ Well, if it's still doing this it's rather hopeless, I think. A very last thing you could try is unplugging it, leave the battery out and let it sit for a few hours, maybe press some buttons like reset (power + pandora button) for a while to empty elcos that eventually still hold charge and retry the next day or so. You'll likely have to send it back, I think.

 

[javaJake]

I saw in the Tricks thread, you reposted MWeston's post, where he said that not all boards are able to boot purely on A/C power. In fact, mine never has (but then again I've never tried it before this whole mess began).

It hasn't turned on. I'll try removing all sources of power for 12 hours and see what that does.

 

[mali]

^ Then I misunderstood. It's still possible that the battery is the culprit, but you barely have a chance to find out yourself. You could insert the battery and leave the compartment open. If you carefully move the battery around a bit and it turns on or off then, the battery contacts might not be aligned properly.

 

[javaJake]

^ Then I misunderstood. It's still possible that the battery is the culprit, but you barely have a chance to find out yourself. You could insert the battery and leave the compartment open. If you carefully move the battery around a bit and it turns on or off then, the battery contacts might not be aligned properly.

If I leave the compartment open, the unit shuts down. It appears that the battery cover is what keeps the battery connected. I'm not sure whether or not this is normal.

Edit: I just remembered... my BeagleBoard liked to hard-reset whenever I under-powered it. Maybe it is a bad battery!

Another edit: Alright, so if that's true, that means the battery needs to charge a bit before it can support the unit. If I close the lid and put the power switch into low power mode, will that be effective even while the unit is off?