[SOLVED] Pandora Lost Charge Overnight

[God Ginrai]

^ It's always best to charge when the Pandora is already switched on and toggle the power switch once to go into low power mode for fast charging.

Actually, from all the problems people seem to be having with low power mode, I would suggest against it for charging. Instead, Toggle WiFi off, make sure no apps are open, be at standard clock or maybe 100 MHz lower, and close the screen. It works fine for me.

-God Ginrai

 

[EmuGuy]

Thanks for the replies! I'm with God Ginrai on this one...getting some flaky results with the power switch, best just to slow everything down myself to charge it up. I'm sure this feature will improve with time.

 

[WizardStan]

Hi there. That must have been my mistake last night. I had run my battery down to about 20%, so I shutdown my Pandora and plugged it in to charge overnight. This morning I unplugged the charger, and tried booting it up, but when it got to the desktop, it immediately shut down again, due to extremely low battery. :huh:

So, I guess the Pandora has to be awake in order to charge...? Or, I plugged in the charger before shutdown was complete...to be honest, I don't remember exactly what I did in what order, which doesn't help troubleshooting. Anyway, hope this helps someone else.

If you plugged in the charger and it didn't immediately boot up, then something strange indeed was going on. I wonder if plugging in the charger at just the wrong time is doing something to the shutdown procedure. Some experimentation is required. Who's got a Pandora that wouldn't mind seeing what they can do to trigger this?

 

[Mithrildor]

Just tested my battery after about 36 hours and it's still at 100%. So folks; make sure to unplug your charger before initiating shutdown or you might wake up to an empty battery!

Thats actually quite normal. Having a charger in the Pandora will boot it up.

 

[conso]

off topic anecdote: I met up with a friend a few weeks ago who wanted to show me something on his laptop. He turned it on, only to be met with the "shutting down" screen. He had closed the lid on his laptop before it had finished shutting down, prompting it to go to sleep WHILE it was shutting down. When he turned it back on, it simply woke up and finished the shut down procedure. It was hilarious.

KDE4 did that for me several times, that's one of many reasons I switched back to lxde.

 

[Pleng]

Actually, from all the problems people seem to be having with low power mode, I would suggest against it for charging. Instead, Toggle WiFi off, make sure no apps are open, be at standard clock or maybe 100 MHz lower, and close the screen. It works fine for me.

-God Ginrai

Has anybody done any conclusive testing to see if underclocking actually has any affect on battery performance? Because the first thing I do when I turn my Pandora on these days is turn my CPU up to 800Mhz, and it the battery appears to last just as long under standard desktop use as it does at 500Mhz.

 

[God Ginrai]

Actually, from all the problems people seem to be having with low power mode, I would suggest against it for charging. Instead, Toggle WiFi off, make sure no apps are open, be at standard clock or maybe 100 MHz lower, and close the screen. It works fine for me.

-God Ginrai

Has anybody done any conclusive testing to see if underclocking actually has any affect on battery performance? Because the first thing I do when I turn my Pandora on these days is turn my CPU up to 800Mhz, and it the battery appears to last just as long under standard desktop use as it does at 500Mhz.

My battery seems to lose less power at lower clockspeeds, but not some overly significant amount. 800 definitely drains faster for me than 500.

-God Ginrai

 

[WizardStan]

My battery seems to lose less power at lower clockspeeds, but not some overly significant amount. 800 definitely drains faster for me than 500.

It was noted that 14Mhz for some reason appeared to actually be draining more battery than the standard 500, so such vast underclocking may actually be bad for it, but no one has yet to do a proper scientific analysis, so it could have been a mistake, or coincidence.

 

[God Ginrai]

My battery seems to lose less power at lower clockspeeds, but not some overly significant amount. 800 definitely drains faster for me than 500.

It was noted that 14Mhz for some reason appeared to actually be draining more battery than the standard 500, so such vast underclocking may actually be bad for it, but no one has yet to do a proper scientific analysis, so it could have been a mistake, or coincidence.

That's why I suggested only a slight underclock if you intended to underclock it.

-God Ginrai

 

[WizardStan]

That's why I suggested only a slight underclock if you intended to underclock it.

But that doesn't actually answer the question: does underclocking it to 14Mhz actually increase battery usage? Where is the sweet-spot of CPU clock that actually maximizes the battery?

 

[God Ginrai]

That's why I suggested only a slight underclock if you intended to underclock it.

But that doesn't actually answer the question: does underclocking it to 14Mhz actually increase battery usage? Where is the sweet-spot of CPU clock that actually maximizes the battery?

There was no question, merely a comment. And if you are referring to Pleng's question, that is not the same question. Because I did answer Pleng's question saying that, yes, underclocking does have an effect on battery performance.

-God Ginrai

 

[Pleng]

Actually, from all the problems people seem to be having with low power mode, I would suggest against it for charging. Instead, Toggle WiFi off, make sure no apps are open, be at standard clock or maybe 100 MHz lower, and close the screen. It works fine for me.

-God Ginrai

Has anybody done any conclusive testing to see if underclocking actually has any affect on battery performance? Because the first thing I do when I turn my Pandora on these days is turn my CPU up to 800Mhz, and it the battery appears to last just as long under standard desktop use as it does at 500Mhz.

Are you referring to idling or when running CPU intensive stuff? When I run emulators, the increased clock speed does tend to have an impact on battery life. But when idle or chatting on Pidgin I get the same 14-odd hours regardless of the CPU being set to 500 or 800

My battery seems to lose less power at lower clockspeeds, but not some overly significant amount. 800 definitely drains faster for me than 500.

-God Ginrai

 

[tsh]

Some quick current measurements, as reported by the fuel gauge device (from the terminal):
Idle, wifi on, 500 MHz: 328mA
Idle, wifi off, 500 MHz: 268mA (odd, can't repeat this)
Idle, wifi off, 50 MHz: 294mA
Idle, wifi off, 100 MHz: 293mA
Idle, wifi off, 780 MHz: 300mA

Now, I believe that the measurement is instantaneous, so this is measuring the current draw whilst reading a system device to get the result, and that is why there is no real change. The only good way to perform a test is by monitoring the battery capacity drop over an hour or so. It may be that there is a more significant increase in current over a certain cpu speed.

 

[mali]

^ Thanks much for testing this

 

[fusion_power]

Some quick current measurements, as reported by the fuel gauge device (from the terminal):
Idle, wifi on, 500 MHz: 328mA
Idle, wifi off, 500 MHz: 268mA (odd, can't repeat this)
Idle, wifi off, 50 MHz: 294mA
Idle, wifi off, 100 MHz: 293mA
Idle, wifi off, 780 MHz: 300mA

Now, I believe that the measurement is instantaneous, so this is measuring the current draw whilst reading a system device to get the result, and that is why there is no real change. The only good way to perform a test is by monitoring the battery capacity drop over an hour or so. It may be that there is a more significant increase in current over a certain cpu speed.

Interesting, indeed the numbers look like underclocking has not a huge effect on Battery life. The good thing is that overclocking also has no huge effect onto Battery life.

 

[mali]

^ Yep. being stuck at OPP3 either way is the main reason, I suppose. Proper clock and voltage scaling should improve battery life further

 

[tsh]

500 MHz, idle, wifi off, lid closed:
4h10m 81%->63% (~23 hours battery life)

Now at 50MHz, must remember to check before going to bed...

 

[mali]

^ Finally someone tests it, thanks much for that
It seems to prove what I mentioned several times in the past.
We could actually have a much improved low power mode with mp3 playing functionality and all.

 

[tsh]

4 hours, 62% -> 45% (50MHz)

So that is basically unchanged from 500 MHz. Low power state only gains at the moment by giving a quick way to disable wifi.
This suggests that the CPU is already either managing to down-clock itself, or stopping when it is idle. Either way, clocking below 500 MHz currently seems to have little benefit. There is the DVFS feature, which ought to be able to adjust the operating voltage on the fly - I have a feeling this isn't in the current kernel, but I'm not sure. Wonder where my undervolt hack lives....

Currently testing playing MP3...
Edit: 63 min, 6% (17.5 hours)

This might be interesting to play with...
http://processors.wiki.ti.com/index.php/Running_PowerTOP_on_OMAP35x_platform
Edit: Maybe needs a more up to date kernel, not sure...

 

[fusion_power]

Very nice and still very interesting, especialy mp3 playing.

But if underclocking has no effect, why did Pandora clock to 14MHz when "power saving" mode is enabled? ^^""
I hope, a new Kernel comes soon for the Pandora. I've read that there could be alot HW stuff enabled with, if we are lucky.