Does battery estimation take in account power settings

[questions2000]

iPod / Linux
I apologize I know this is an issue with iPods but not sure if other devices experience this as well.

I have read in the RB iPod manual that RB developers have mentioned there have been difficulties figuring out how to accurately display the battery percentage, so that at the moment battery percentage is based on an estimation until a better method is figured out.
I look at the work being done by RB developers as amazing, so I can imagine that the battery must really be something difficult to accurately read/access.  I wonder how the Apple firmware accesses the battery to generate a value.  Although I do not remember them showing an actual percentage, just a picture of a battery so hard to say how accurate they were either?

Estimation also makes it a little difficult to get an idea of how efficient RB is compared to Apple firmware as far as power consumption.

Which brings me to my question about the estimation.
It can also be a little difficult to know how much battery life is really available, so I was wondering if anyone knew if any of the other power settings are taken into account when it comes to the estimation.

So as far as I know, the estimation is based on what "Battery Capacity" setting is choosen, for example 550mAh.

Does anyone know if any of the following settings are taken into account by the battery estimation?
-Backlight = Off
-Display = Sleep when Backlight is Off
-Brightness = Lower values
-Accessory Power Supply = Off
etc.

Meaning for example does turning "off" backlight completely alter the battery estimation shown in the battery meter?
Or is the estaimation a static setting based only on the size battery choosen in the "Battery Capacity" setting section?

Thank you for any information offered.

[saratoga]

Are you asking about the battery state of charge or the runtime estimation (hours remaining)? Battery state of charge is by measuring the battery voltage. Runtime estimation is the battery state of charge times battery capacity in mah divided by the run current for the device.

[Bilgus]

We have a way to import battery levels for the end user now

First time you run battery bench the current battery levels will be exported as

Code: [Select]

battery_levels.default
Set your device up as you normally use it and set a playlist to play on repeat with the same formats of the songs you generally listen to (mp3 flac etc)

now run battery bench till the device dies (maybe do this several times and average the results to get really precise)

now alter

Code: [Select]

battery_levels.default with your new voltage values and rename it to

Code: [Select]

battery_levels.cfg
https://www.rockbox.org/wiki/PluginBatteryBenchmark.html

an example of the defaults from my clip zip:

Code: [Select]

# Rename to /.rockbox/battery_levels.cfg
# Sandisk Sansa Zip Battery Levels (f93982349bM-250112)

# Battery voltage (millivolt) lower than this won't access the disk to write
disksafe = 3400

# Battery voltage (millivolt) lower than this player will shutdown
shutoff = 3300

# Battery voltage (millivolt) of 0%, 10%, ... 100% when charging disabled
discharge = { 3300, 3653, 3701, 3735, 3768, 3790, 3833, 3900, 3966, 4056, 4140 }

# Battery voltage (millivolt) of 0%, 10%, ... 100% when charging enabled
charge = { 3427, 3786, 3842, 3877, 3896, 3924, 3971, 4028, 4084, 4161, 4190 }

I wouldn't mess with the disksafe level if you are NOT using solid state storage

Also you shouldn't change shutoff unless you know what you are doing and the underlying battery chemistry,
as most devices will still shutoff anyway and the ones that don't will likely damage the battery if you set it too low.


basically you will take your battery bench results and map them to the 0-100 of the battery level to get a more accurate display of battery capacity
https://download.rockbox.org/daily/manual/rockbox-ipod6g/rockbox-buildch12.html#x19-33600012.4.2

[questions2000]

Thank you guys for these replies.
I am still interested in the question, just need a minute.
I typed out a long response and then realized there was a new reply