Hi all,

I just bought a Kill A Watt P4400 from RadioShack and I thought that I would share my very unscientific findings with you about my N800's power consumption profile. I hope that someone might find this data useful, as there's a dearth of information out there on the tablets' wattage and because I'm sure that there are other people like me who prefer to use the tablets partly because of their low energy usage.

Boot: Generally steady at 8 W with dips to 0 W and 6 W

Idle: With one MicroB window and one AbiWord window open while connected to a Bluetooth keyboard and the Internet using my school's crazy security (I forget the details, but it's a WPA2 w/ PEAP setup), the meter reports 0 W with rare spikes to 3 W.

"Stress test": To prepare for this, I opened up two Tear windows, a Modest mail inbox window, a MicroB window, and a six-page document in AbiWord. Then, I observed the meter while opening OpenOffice (using easy deb). The meter generally stayed steady at 3 W with occasional spikes to 4 W and dips to 0 W. I was impressed, especially considering that I'm using a 48 MHz SD bus enabled kernel and swap file usage was at 40 MB.

General usage 1: With OpenOffice open in the background without any document loaded while typing this post in Tear, the meter read 0 W with occasional spikes to 3 W.

General usage 2: After finding that both Tear and Midori were still crashing on me as I tried to type this post for the third time (and after trying the horrible experience of using talk.maemo.org with MicroB--did you know that pressing backspace while typing a post here will navigate back one page??), I fired up Epiphany using easy deb. OpenOffice is open in the background as is the Modest mail Inbox window and one Maemo X terminal window, and the meter is consistently reading 3 W with seldom spikes to 4 W. (Side note: I was pleasantly surprised by how fast Epiphany is despite using the Gecko engine. It certainly works better than the other options. Why doesn't someone make a Hildon port and port the engine package?)

Fennec: Earlier today I had tried to load CNN.com in Fennec just to see what would happen. Load time was slow, and I thought that perhaps power usage would spike significantly, as well. While trying to load CNN.com again with the meter attached and one window of Epiphany open, however, power consumption only spiked to 4 W. Swap usage was at 35 MB.

Just Epiphany: I began to suspect that Epiphany had a big power draw. I've closed all other windows and am currently writing this note with my Bluetooth keyboard while connected to the Internet, and the meter is reading 3 W steady. Swap usage is at 35 MB.

So some interesting findings. I will continue to investigate power usage while using Epiphany, as I really like this browser.

Interesting. Could you try with the internet radio? The Nokia one, not vagalume. In the past I found that nothing sucked the battery as fast as streaming internet radio, and the N800 got slightly warm too. (Vagalume was less demanding, for some reason.)

If you can, try Angband-sdk. I notice that even when the device is locked, it sucks battery faster with Angband open than without. As an example, I went half a day with my N810 locked without Angband running, and still had 4 bars of battery. With Angband over the same period, I would be sitting at 2 bars.

Again, completely unscientific findings.

--vr

iirc, most browsers do that, if the input area is unfocused. I really do wonder what browser first displayed this behavior, as it makes basically zero sense, when one already have gui buttons and, when using a keyboard, keycombos that do the same, and is less likely to trigger when doing something else...