Efficiency of power bricks?

[jvroig]

I'm interested in learning how efficient power bricks are (such as in laptops or mini-ITX rigs) in general or, if possible, how to find out the efficiency of specific power bricks (this is doubtful as I tried the manufacturers' site - HP and ASUS - and of course speccing the efficiency of the power brick their laptops come with is not a priority).

Anyhoo, I thought I'd hit up the AT PSU mega-brains for input.

Are these generally 70%, 80%, 90% efficient? Because when I'm reading, for example, 40W or 7W using a kill-a-watt while it's plugged in, I'd like a more accurate estimate of what the electronics (SOC, wi-fi, integrated peripherals, etc) actually draw.

Nothing too important, but if one of you guys have excellent info, I'm all ears. I don't imagine we have so much literature about this, otherwise it would have turned up in google very easily, but no harm in asking.


Regards.

 

[nForce2]

I went through a similar exercise a while back, when I was trying to find a power supply for an ultra-low-power PC104 system.

I eventually got it down to 6W total power usage (including the power brick - measured at the wall with a Kill-A-Watt), but when I started, it was at 15 watts! In my case, the power brick alone was taking more power than the whole computing system!


I did a whole bunch of research, and experimentation with a dozen or so different power bricks, and these were my conclusions:

1) There *ARE* ultra-high-efficiency power bricks out there, and you can tell which ones they are because the sellers will be advertising their efficiency.

2) These efficient power bricks cost so much, that even after years and years of 24/7 use, you won't see enough of a return from your electricity bill savings to pay off the high cost of purchase (3-5x the cost of a "regular" power brick). :\ (I suspect this high cost is why you don't see very many of these for sale...)

3) Generally you get the best efficiency when the rating of your power brick is relatively closely matched to the needs of your load. But when I tested a bunch of my power bricks, my lowest-capacity power bricks were actually beaten by one of the middle-capacity bricks. So from an efficiency standpoint, the power supply design is more important than the output ratings.

4) Pay attention to the current draw when off, as that could make a huge difference in your *overall* power usage. In my case, I had some bricks that were drawing 4+ watts when plugged in on their own (no computer attached at the other end)... and eventually found one that used so little power that my Kill-A-Watt didn't register anything when it was plugged in. :thumbsup:



For finding more information, try adding "picoPSU" to your search terms. That will narrow down the results, and give a lot more relevant hits from the people who are really squeezing out efficiency wherever they can. There were a few sites (can't remember them now, sorry) where people were specifically rating different power bricks for efficiency, when paired with a picoPSU and a Mini-ITX system. :thumbsup:

 

[jvroig]

Interesting, thanks for the info :thumbsup:

The lowest I read using a kill-a-watt is 6W from this HP power brick that came with the laptop. That's with the laptop at idle, and the screen off (lid closed, but I set it to just lock the screen & turn off the LCD, instead of suspending/hibernating). Since there is some inefficiency from the power brick, the kill-a-watt certainly isn't reporting the actual idle power draw of the laptop, and if the brick is only 80% efficient or less, that would peg the actual idle power draw a little lower (6W * .80 = 4.8W), but if it was even less efficient like only 60%, that would peg the actual idle power draw to much lower (6W * .6 = 3.6W).

I'll do a more thorough reading on picoPSU as you suggested. Thanks for the tip!