Backlight power consumption analysis.

John Johansen john.johansen at canonical.com
Tue Dec 13 09:01:48 UTC 2011


On 12/12/2011 01:03 PM, Colin Ian King wrote:
> Hi there,
> 
> More power consumption analysis... this time I investigated the current drawn when cycling through all the back-light levels on various machines.   The results are interesting as they show non-linearity in power consumption.  A brief summary of my findings is:
> 
> All machines showed up non-linear current increase as backlight levels increasse.
> At the highest backlight setting the machines show a reduction of battery
> life from ~12-22%.  The Lenovo ThinkPad X220i shows a remarkable increase in
> power consumption as we approach the highest brightness levels, and so does
> the HPMini-100 to a lesser degree.
> 
> The general rule of thumb therefore is try not to exceed more than 2/3 the
> way up the brightness scale.  The top 1/3 of the brightness scale impacts
> of power consumption disproportionately.
> 
This matches with what I would expect, LEDs efficacy suffers as you increase
the current.  This is known as droop and it is what sets the limit of a LED's
brightness.

> If you want to see the data and some interesting graphs, I've put a LibreOffice spreadsheet in
> with a brief write-up of the results in:
> 
> http://zinc.canonical.com/~cking/power-benchmarking/backlight-non-linearity/
> 
So out of curiosity did you test with the back light off to get a measure of
what the draw was for just turning it on?



More information about the kernel-team mailing list