Do brighter colours save LCD power?

[dreamlayers]

Do LCDs use more power to display dark pixels?  Pixels are normally transparent, and a voltage needs to be applied to make them less transparent.  This presumably uses some power.  I remember reading about tests of PC monitor power consumption, which found LCD monitors use more power when displaying a black screen.

Update: Measurements taken via FS#9728 and FireWire input power show that on my 5th generation 30 GB iPod, a white screen uses about 2 mA less than a black screen.

[Chronon]

Interesting, since I have been led to believe there shouldn't be any current flow in a TFT display element (not that I expected no power to be consumed, of course).  Is consumption higher while switching -- e.g. for a flickering screen?  I suppose that will take extra work to test. . .

[soap]

I don't know if this tool is sharp enough to see the (I assume) teeny tiny difference we are talking about, or if repeated carefully controlled runtime tests will be needed.

[dreamlayers]

Interesting, since I have been led to believe there shouldn't be any current flow in a TFT display element (not that I expected no power to be consumed, of course).

LCD elements are driven by AC (because DC would degrade the LCD via electrochemical changes).  Each element has some capacitance, making it like a capacitor connected to AC when darkened, and using power.  There may be a some very slight leakage current also.

Is consumption higher while switching -- e.g. for a flickering screen?  I suppose that will take extra work to test. . .

I know that would use power because new data needs to be generated and sent to the display.  I don't think the display elements themselves would use more power because they must be refreshed constantly anyways.

I don't know if this tool is sharp enough to see the (I assume) teeny tiny difference we are talking about, or if repeated carefully controlled runtime tests will be needed.

I have seen repeatable results both with FS#9728 and by measuring FireWire current with a multimeter.

Here's a study of this on PC monitors.  (I'm linking to an archived page because the site seems down.)

[Chronon]

LCD elements are driven by AC (because DC would degrade the LCD via electrochemical changes).  Each element has some capacitance, making it like a capacitor connected to AC when darkened, and using power.  There may be a some very slight leakage current also.

I was not aware of that.  I think you mean electromechanical, but I take your point.

[dreamlayers]

LCD elements are driven by AC (because DC would degrade the LCD via electrochemical changes).  Each element has some capacitance, making it like a capacitor connected to AC when darkened, and using power.  There may be a some very slight leakage current also.

I was not aware of that.  I think you mean electromechanical, but I take your point.

I really mean electrochemical.  A short period of DC moves ions toward oppositely charged electrodes, and a longer period can lead to irreversible chemical reactions due to the applied voltage.

BTW this is also why LCDs must be shut down properly.  With some displays, just cutting power can leave a charge on some elements (eg. a faded image or some dark lines).

[Chronon]

Ah, okay.  The brief reading I had done suggested that AC was used to prevent a persistent dipole moment from arising in the liquid crystal molecules.  I'm certainly not an expert on LCDs.