Power supply amperage query

CurlyWhirly · 14 Jul 2009 at 20:00

When I read power supply specs like the one from my PSU (below) I get confused as the total amperage (Max Output Current) exceeds the fuse in the plug unless DC is not directly comparable to AC amperage :confused:

I don't know anything about electrics by the way

siumatfung · 14 Jul 2009 at 20:10

nah its fine

CurlyWhirly · 14 Jul 2009 at 20:14

siumatfung said:
nah its fine

That's good to hear but I still don't understand it

james.miller · 14 Jul 2009 at 20:16

whats the fuse in the plug then? remember its a fuse rated for 240v use. ie, 3A x 240v = 720w

Jokester · 14 Jul 2009 at 20:17

600W will draw approx 3A from the wall socket at 230V. P=IV, so down the voltage, up the current for the same power (hence why you have 36A on the 3.3V rail).

Mr-White · 14 Jul 2009 at 20:18

We used to have to use 5A in the plugs we sold

MW

CurlyWhirly · 14 Jul 2009 at 20:50

Jokester said:
600W will draw approx 3A from the wall socket at 230V.

I've just checked the PC plug and it's a 5A fuse so it should be fine.
I think that a 3A fuse would be pushing it!

Jokester said:
P=IV, so down the voltage, up the current for the same power (hence why you have 36A on the 3.3V rail).

I had to google "P=IV" and apparently it's to do with Ohms Law and I'll have to examine it further as it doesn't make much sense to me at the moment.

p.s. I have only upgraded my PSU a few days ago (from a 500W generic one to the OCZ 600W StealthXstream) and I don't seem to have messed anything up even though I have never carried it out before

I just wanted to try and understand power supply specs better hence my query

CurlyWhirly · 14 Jul 2009 at 20:55

james.miller said:
whats the fuse in the plug then? remember its a fuse rated for 240v use. ie, 3A x 240v = 720w

A 5A so it should provide up to 1200W according to your formula.

Perhaps I will change it to a 3A fuse just to be on the safe side although most power supplies come with over-current, over-voltage protection, etc so I might just leave it as it is.

I never paid any attention to fuse ratings before

james.miller · 14 Jul 2009 at 20:58

that is correct, more or less. remember a psu that is rated to output 600w (assuming it can actually do that) will need more than 600w input. assuming its say...85% efficiant at peak output, thats about 706w input. if it were 80% efficiant at peak, thats 750w input.

i wouldnt worry about changing the fuse in the plug, it'll be correct for the psu. if the psu is going to go bang, it'll go bang wether there's a 3a or a 5a fuse in there.

take a look at electric heaters, those things are monstorously power hungry - 2-3kw yet they run of a standard power socket, using a 13amp fuse which is good for about 3.1kw

CurlyWhirly · 14 Jul 2009 at 21:03

james.miller said:
that is correct, more or less. remember a psu that is rated to output 600w (assuming it can actually do that) will need more than 600w input. assuming its say...85% efficiant at peak output, thats about 706w input. if it were 80% efficiant at peak, thats 750w input.

i wouldnt worry about changing the fuse in the plug, it'll be correct for the psu. if the psu is going to go bang, it'll go bang wether there's a 3a or a 5a fuse in there.

Okay I'll take your advice and leave it as it is

Thanks for everyone's advice and sorry if it's such a noob question

JonJ678 · 14 Jul 2009 at 21:37

I'm going to be pedantic I think, because no one else has yet. P=IV is only applicable for direct current. Mains ac is not direct current, so you can't rely on this.

I've just tried to type up the equivalent for ac and discovered I can't remember it, shame on me. The conclusion would still be that 5A is fine, if you manage to blow it with your multigpu rig then 7A is the next step up. It's black, like 10A and 5A, and may be difficult to locate.

Pretty sure ac power assuming sinusoidal (resonable) is at most half what you'd expect from dc of the same voltage and current. It varies between this and zero depending on how out of phase the voltage and current are, engineers will have gone to considerable effort to keep the value closer to 1/2VI. There's more to it than this though. Whether this means drawing 460W at 230V needs a 2A fuse or a 4A I'm currently struggling with, as I don't really know how power draw at the wall is measured. The 460W figure at the wall is probably in terms of ac anyway, thus negating the entire above arguement. P=IV still isn't applicable, it just happens to work out ok. That's the current conclusion anyway *rambles on*

westom · 14 Jul 2009 at 22:37

JonJ678 said:
Pretty sure ac power assuming sinusoidal (resonable) is at most half what you'd expect from dc of the same voltage and current. It varies between this and zero depending on how out of phase the voltage and current are, engineers will have gone to considerable effort to keep the value closer to 1/2VI.

Parameters sought include harmonics and power factor. For example, most all computers consume less than 200 watts (despite the hype about wattage) and yet also demand something less than 300 volt amps. Volts times amps is watts. And then other factors apply. This specmanship game is also why some power supplies from mainstream computer manufacturers rated at 350 watts are equivalent to 500 watts supplies marketed to computer assemblers.

Just to make it more interesting. Many assume a computer consumes more power during power on. But most better supplies include a circuit to limit power consumption during power up. Computers love to be powered up during brownouts while motorized appliances hate it. A power supply will even create a brownout internally during power on.

westom · 14 Jul 2009 at 22:42

james.miller said:
whats the fuse in the plug then?

To understand how a fuse is selected, read the first sections in Wikipedia on electric fuses. Appreciate no step function between a fuse not blowing and when a fuse blows. Just because the fuse is 3 amps does not mean it blows at 4 amps.

JonJ678 · 14 Jul 2009 at 23:06

westom said:
Parameters sought include harmonics and power factor.

With regard to harmonics I'm chosing to stick my fingers in my ears and sing loudly. I did not get on well with fourier, there are no harmonics until further notice.

Which power factor would this be? It sounds suspiciously like a dimensionless coefficient to explain why the numbers don't match reality

Oh, naturally a plug fuse doesn't blow precisely. Its a piece of wire that fails by melting, sustained at 3A in high ambients would probably kill a 3A fuse eventually, and an instantaneous pulse at 30A probably wouldn't. I believe psus to be internally fused however, likely with tighter tolerances