I got my stuff today and spent several hours watching the meter and compiling a text file as I went. Here is a copy and paste.
I got my kill a watt p3 model p4400.01 in today so I ran a few tests. To quickly test its accuracy I plugged it into my 30 amp computer wall outlet and checked VAC and frequency compared to the readings I get from my Fluke 83 DMM. The Fluke is quite a bit more expensive.
kill a watt fluke
VAC 123.4 124.4
FREQ 59.9-60.0 59.98
Computer readings. Everything is plugged into a brand new Cyberpower model 1080 surge protector/ power strip. So normal readings include the power used by the mobo, cpu, ram, Hauppauge tv tuner,1 WD hdd (another one not connected), Sony dvdrw, Antec Earthwatts 430 watt 80+ psu (it is rather old), Motorola SB5101 cable modem, Viewsonic VX 2250 - A 22" LED monitor ( Viewsonic states it uses 28 watts at full brightness), my keyboard and mouse and a webcam that should not be running at all because I have not yet reinstalled guvcview since the KDE upgrade. That's it. That is what is running and possibly using power.
Everything off, power stripp off 0 watts
power stripp on- feeding power to modem and monitor 7.4 watts
watts jumps to 30 while monitor displays Viewsonic logo ets and then
goes to idle due to no signal
Boot 82 - 104 watts
Desktop (monitor set to 100% brightness) 83 watts
Desktop ( monitor adjusted, again, to 50% brightness) 73 watts
I have adjust the monitor onscreen brightness down to 50 % before.
I must screw things up while cleaning and running a cloth over the
touch sensitive buttons on the frame.
50% setting is not really half way from 100 to 0, more like 80% to me.
Desktop, monitor turned off 56 watts
Viewsonic numbers look very believeable
Firefox open to http://www.wunderground.com--
my weather page 74 watts
Firefox open to Tom's Hardware 76 watts
All web pages are about the same unless I run across a whacky one
that is probably trying something a little unusual with advertising?
Justin tv tuned in to stargate sg1 86 watts
Veetle tv tuned in to Futurama 76 watts
Back when I was running my old Pentium D, I noticed my cpu use was a lot
higher at Justin, they I believe use flash. Veetle uses a VLC plugin.
Very noticable then, somewhat noticable now.
You-tube Amazing Nature 4:58 Flash 1080p 82 watts
You-tube Space Shuttle Launch 3:53 Flash 1080p 81 watts
You-tube Planet Earth- Angelfalls 2:06 .webm WebM/vp8 html compatible 1080p 82 watts
From Hard drive: (video- just some random stuff)
some 1080p .wmv opened with VLC 79 watts
Dr Who episode .mp4 at 624 x 352 76 watts
Mars Attacks .mp4 1916 x 796 79 watts
In all video there is just no noticable difference from playing in a
window to going to full screen. With You-tube, I actually thought it
was maybe 1 watt less in full screen
Playing video from DVD:
Get Smart episode recorded at 575 x 432 .mp4 81.5 watts
Blue-ray rip of something recorded at 1920 x 1080 H264 .mpg4 at 59.94 fps 92 watts
That's it for video. I conclude that all video playback on Ivy Bridge
with the onchip HD4000 graphics, whether in window or fullscreen is pretty
darned close. The numbers I see from this Kill-a Watt meter correlate very
nicely to the numbers reported by my Conky display.
I loaded a default/preset and rendered that just to see what the outcome
was. I know all 4 cores jump to max. 133 watts
Phoronix Test Suite:
Only have the one Xonotic .6 test installed. First run was after I had run
the above video tests etc. No reboot etc. ultimate settings. 17.57 fps 110 watts
Rebooted and upped DRAM voltage from 1.565 to 1.580 (yeah, I overclocked
my ram from 1600 to 1866-- still playing) strange hesitations while
running above test
New test- after the reboot and dram change 18.96 fps 98 watts
This needs more time. I don't know.
Power use with sound system plugged in and set to 0 volume 95 watts
Remember, bare desktop is 73 watts. 20 watts doing nothing
Volume set to 1 turn type button goes 0 to 10. 102 watts
volume set to 2 123 watts
volume set to 2 1/2 130 watts
volume set to 3 145 watts
volume set to 4 170 watts
beyond 4 went over 200 watts- too loud- 4 is ok though.
That is it so far for my first day with this power meter. Everything is pretty darned close to what I expected, except for the Phoronix benchmark difference. Have fun with it.
Side-note: Formating on a copy/ paste sucks. My nice neat columns just aren't anymore.
i5 3570k, ASRock z77 Extreme 4, 8 Gb Ripjaws 1600, Antec 430w psu, HVR 1600 tv tuner, custom case- marble top, oak face. Carver & DCM Time Window sound system. Mint 14 KDE.