Power consumption of a PSU????

[Confused_Chimp]

I'm looking for a bit of an explanation of power usage of PSUs.

I bought a Seasonic 600W PSU for my current computer. On the basis that it got some very good reviews and that it would provide enough power for anything that that I put inside my case.

Part of this decision was based on my understanding that the PSU will only be drawing what it requires from the mains electricity.

However I now have another friend telling me that the PSU will always be drawing the maximum amount from the mains, but may not use it all as the components may not need all the electricity.

So my question is this, if my computer could be run on say a 400W or a 600W power supply because each would be sufficient to power everything I am currently using, would the 600W PSU cost me more in terms of electricity than the 400W? Or would they both cost me the same to run, with the only difference being that the 600W would ahve more spare capacity to power more components?
And for this example I'm ignoring all the other factors, efficiency etc...

Please excuse me if I use the wrong terms: power, current, ampatude, watts etc. I could go an look them up, but I'm not that interested unless I need to. I'm not looking for an explanation of the relationship.

Just looking to find out if its costing me more or the same on the electricity bill? and if it is costing me more, is it worth trying to work out the maximum all my components could draw at any one time and get a power supply that is just capable of coping with that and no more?
Cheers for any help
CC

 

[doolittle]

the PSU will only be drawing what it requires from the mains electricity.

correct, with a slight adjustment for the effiency factor (loss in power of the internal components used in the PSU itself)

I now have another friend telling me that the PSU will always be drawing the maximum amount from the mains, but may not use it all as the components may not need all the electricity.

incorrect (but common assumption from the PSU's rated power capacity)

So my question is this, if my computer could be run on say a 400W or a 600W power supply because each would be sufficient to power everything I am currently using, would the 600W PSU cost me more in terms of electricity than the 400W? Or would they both cost me the same to run, with the only difference being that the 600W would ahve more spare capacity to power more components?

This is where effiency factors in to the equation, which varies on the load factor placed on the unit.

If the 400w unit had a load of 300w, and was 75% efficient at 75% load it is "better" than a 600w unit that has a 70% effiency at 50% load.

Now the 5% difference is so small and will take a long time to accumulate any visible savings, say running 24x7 for a few years it maybe worth it - if you never upgrade your hardware. Once you upgrade (say your GPU) and you outgrow your 400w unit, any savings go out the window since you are forced to buy a PSU.

Just looking to find out if its costing me more or the same on the electricity bill? and if it is costing me more, is it worth trying to work out the maximum all my components could draw at any one time and get a power supply that is just capable of coping with that and no more?

IMHO no, just stick to your guns - you made an excellent choice since that unit will have high effiency at all load percentages, and will last you a very long time. Unless power specs become very different but that is unlikely since power consumption is being factored in to the manucfacturing process - and should reduce the total wattage vs. increase it for the up and coming hardware (i.e. "conroe") and the quest for power at cost over watt consumption will hopefully come to an end. Well except for ATI cards

 

[jaguarskx]

I bought a Seasonic 600W PSU for my current computer. On the basis that it got some very good reviews and that it would provide enough power for anything that that I put inside my case.

The Seasonic S12 600 is an excellent PSU. It has been test for it's efficiency over at www.SilentPCReview.com. If I remember correctly, at it's worse the Seasonic S12 500/600 are only 77% efficient at low loads of 25% or lower of the max capacity. That would be 125w for the S12 500 and 150w for the S12 600. As the load increases, efficiency increases. I think the Seasonics are at thier most efficient between 50% - 90% of the max capacity. At that point the Seasonic S12 are 82% - 85% efficient. At higher loads the efficiency drops a little bit but not below 79% efficiency.

I have the Seasonic S12 500 and I did a lot of research on it before I bought it. I am glad I did because it is really quiet, efficient and reliable.

Good Choice!

 

[Mobius]

Your friend is a fool and his assertion defies all logic.

Why would you even begin to think he had a point???

 

[JonathanDeane]

I have a high effeciency PS its rated at 85% at 50C wich brings up another point you also have to check what temps they test at when they list effeciency some companies like to test there PS's at something like 20C wich would be just above freezing.... LOL

 

[jaguarskx]

I have a high effeciency PS its rated at 85% at 50C wich brings up another point you also have to check what temps they test at when they list effeciency some companies like to test there PS's at something like 20C wich would be just above freezing.... LOL

The ambient temperature was 22C - 24C. Of course inside the PC the temp would be higher.

I wouldn't even attempt to turn on a PC if the ambient temperature was 50C.

 

[JonathanDeane]

The ambient temperature was 22C - 24C. Of course inside the PC the temp would be higher.

I wouldn't even attempt to turn on a PC if the ambient temperature was 50C.

Yes but the point is that if your PSU is tested at lower temps it will provide much better numbers then if its tested at realistic temps. I for one am glad they test at 50C so I can get a good idea of the quality

 

[jaguarskx]

Yes but the point is that if your PSU is tested at lower temps it will provide much better numbers then if its tested at realistic temps. I for one am glad they test at 50C so I can get a good idea of the quality

Got a link to that review? I would like to check out the test environment.

 

[JonathanDeane]

Sure thing

Google has about 135,000 hits on "Antec NEO HE Review" so ill just pick one at random lol (well not too random ill pick the exact same PSU I am using)

http://www.thinkcomputers.org/v2/index.php?x=reviews&id=308

Acording to this sometimes it even hits 99% effeciency... maybe another review site is in order

Maybe this one...

http://www.driverheaven.net/reviews/neoHE550%20review/

This ones from the company so I take it with a grain of salt...

http://www.antec.com/NeoHEawards.html

 

[jaguarskx]

Well none of the tests were actually done in a 50C test environment. It was only stated as:

The Total Power is listed as 550Watts for the available continuous output at 50C ambient temperature.

While that is a nice statement, I would like to see actual test results that backs up that claim.

Sorry, but if I can't see it being tested, then I can't believe such a bold statement.

 

[Datman]

like 20C wich would be just above freezing.... LOL

20'C is a nice cool and comfortable temp. 8) 0'C is freezing.

20'F is below freezing. 32'F is freezing.

http://www.straightdope.com/classics/a891215.html

 

[jaguarskx]

The ambient temperature was 22C - 24C. Of course inside the PC the temp would be higher.

I wouldn't even attempt to turn on a PC if the ambient temperature was 50C.

Yes but the point is that if your PSU is tested at lower temps it will provide much better numbers then if its tested at realistic temps. I for one am glad they test at 50C so I can get a good idea of the quality

By the way 20 degrees Celcius is equal to 68 degrees Fahrenheit.

50 degrees Celcius is equal to to 122 degrees Fahrenheit.

That is why I find Antec's claim hard to believe. That is almost the average temperature in Death Valley. And is why I stated I would not even turn on my PC if the ambient temperature was 50C.

 

[Datman]

http://www.pcpowercooling.com/home/

http://www.pcpowercooling.com/technology/myths/
http://www.pcpowercooling.com/technology/terminology/
http://www.pcpowercooling.com/technology/optemps/

 

[JonathanDeane]

Well anyway its a nice quiet PSU and runs extreamly cool They have a good reputation so I doubt they would lie about testing lol and yes I know 0C is freezing (well alot of people would complain they where cold if it was 68F in there house as the comfort zone is idealy 72-74F) Of course none of this realy matters all I said was that different companies test there PSU's under different conditions in order to make sound better and thats a fact While I dont know if Antec actualy tests at 50C I realy dont think they would make that up

 

[doolittle]

An artist's rendition of the Antec PSU testing facility, and some engineers conducting tests:

 

[JonathanDeane]

An artist's rendition of the Antec PSU testing facility, and some engineers conducting tests:

HAHAHA !!! I like it !!!