SPCR

Got AC 36,4W of apparent power with EDAC AC-DC(235V>20V) + Winmate DC-DC(20V>ATX), idle. Apparent means ...there is even less, maybe 30W of active power I struggle now for getting one of a 230V Kill-a-Watt versions, then I'll tell precisely how much is consumed.

Configuration:
E6300 2.8GHz @ 1.0125V + 1W cooler
Biostar G31D-M7 (a bit modified)
Patriot PE 2x2Gb DDR2-800 (1.7V)
Samsung HD502HI 5400
+mouse/keyboard/flash drive
(w/o ODD)

AC ilde: 36,4W
AC load: 65,8W

...

My DC wattage is very low (see the pic bellow, measured w/o HDD&ODD). So I have to do something with power brick, cause it has 48% efficiency at such low idle. Maybe a 40W adapter of EEE PC will do, or I'll make something custom to feed DC Winmate board.



P.S. Add 3.2W to DC total in the table -- HDD do consume too

Ksanderash--

You don't mention how any of those power measurements are obtained -- AC or DC. Are they right? I don't believe the power brick would drop to 48% efficiency. It's rated at 85%!

MikeC wrote:You don't mention how any of those power measurements are obtained -- AC or DC. Are they right? I don't believe the power brick would drop to 48% efficiency. It's rated at 85%!

AC! )) There is written! I measure by the means of a digital multimeter U1242A Agilent, it's quite adequate thing, though it can not tell me how much real power is there, and I simply multiply average V on average A. So I guess the Kill-a-Watt readings will be significantly lower.

Yes, 90W power brick is rated at 85% (w/o range mentioning), but I have 17W load! And my multimeter shows 36W, then I loose 19W of energy in:
1) Edac AC brick,
2) Winmate DC board,
3) reactive power, that is added to active(i.e. real) power by Agilent.

...

As for the table, there are DC values. I took average readings. The DMM was connected as an amperemeter in the break of a supplying wire of each DC source from Winmate ATX connector. I wrote here abt it.

Ksanderash wrote:AC! )) There is written!

Not sure what you mean there, but when I wrote AC or DC, I meant that you did not explain how either of these measurements were obtained. And I don;t see it written anywhere in this thread, if that's what you mean.

But you do explain now:

I measure by the means of a digital multimeter U1242A Agilent, it's quite adequate thing, though it can not tell me how much real power is there, and I simply multiply average V on average A. So I guess the Kill-a-Watt readings will be significantly lower.

I'm not sure if it would show lower -- depends on good the Active PFC in the EDAC is -- it apparently has APFC.

As for the table, there are DC values. I took average readings. The DMM was connected as an amperemeter in the break of a supplying wire of each DC source from Winmate ATX connector. I wrote here abt it.

OK, you might be right, you might be getting less than 50% efficiency.

But then you mention that the DC power numbers exclude the power taken by the HDD. What about the AC power? If the AC power includes the HDD, this would not be fair, but you don't mention it -- the AC power figures are given after you describe your config, which lists the HDD.

If the HDD power is excluded for DC and included for AC, then you'd have to make adjustments:

I measured the Samsung HD502HI idle power at 3.2W (shown in my review). So adding this 3.2W, we get ~20W DC for your system at idle and ~48W at peak. At the higher load, PF of the EDAC might be 0.9, which means real AC wattage would be 60W -- this means 80% eff. For the idle load, the EDAC PF is probably lower, maybe 0.8. If this is so, the real power is ~29W, and the eff. is 69%, not as bad as you think.

MikeC wrote:I'm not sure if it would show lower -- depends on good the Active PFC in the EDAC is -- it apparently has APFC.

Yes, you right.

But then you mention that the DC power numbers exclude the power taken by the HDD. What about the AC power? If the AC power includes the HDD

AC includes HDD, and DC doesn't. I just was lazy to measure the HDD DC

I measured the Samsung HD502HI idle power at 3.2W (shown in my review). So adding this 3.2W, we get ~20W DC for your system at idle and ~48W at peak. At the higher load, PF of the EDAC might be 0.9, which means real AC wattage would be 60W -- this means 80% eff. For the idle load, the EDAC PF is probably lower, maybe 0.8. If this is so, the real power is ~29W, and the eff. is 69%, not as bad as you think.

Mike, you're right absolutely. I was too pessimistic naming such low efficiency. Forgot about HDD power.

hmsrolst wrote:Thanks for linking that thread, which I hadn't seen. Given the small difference at idle between G31 and G45, it seems likely that other differences in motherboards are likely to be more significant.

Just from looking at the two 785g MB SPCR review I can tell you there can be big differences power consumption among motherboards of the same chipset. This was a big surprise for me as well.

BTW I am a little disappointed they didn't try the Winmate with a 80W brick for a more direct comparison with the pico psu. Given my Winmate + ~80W NEC brick + Athlon 240 + 785g system idles at 32W from the outlet I suspect there are some gains to be had here.