| Forum Home | ||||
| Press F1 | ||||
| Thread ID: 51056 | 2004-11-10 08:48:00 | PSUs - Why so utterly misleading? | george12 (7) | Press F1 |
| Post ID | Timestamp | Content | User | ||
| 290085 | 2004-11-10 08:48:00 | Why are Power Supply makers allowed to make their wattage figures so misleading? And where do they even get them from? Let's take the 3.3V rail for example (important as it powers the CPU). My PC has a 400W Olympic power supply. It came with my case. The current for the 3.3V rail is 20A. That means the wattage is 3.3 x 20 = 66W for that line. If a power supply could do 100W for that line, the current would be higher as the voltage would be the same. Therefore, even a dirt cheap crap as hell power supply will provide what its WATTS rating is. If we add all the other rails up: 3.3V @ 20A = 66W 5V @ 20A = 100W 12V @ 10A = 120W -12V @ 0.5A = 6W 5VSB @ 2A = 10W -5V @ 0.5A = 2.5W -------------------- 304.5W Now that's interesting, isn't it? So the output of my 400W power supply is pretty much 300W... I'm sure there is a reason but anyway, moving on: Now let's look at the 400W power supplies I sell in my shop. They're made by Leaf. 3.3V @ 30A = 99W 12V @ 16A = 192W -12V @ 0.5A = 6W 5V @ 32A = 160W -5V @ 0.5A = 2.5W 5VSB @ 2A = 10W --------------------------- 469.5W The PSU (bottom) by far outperforms the one at the top, by an astonishing degree - but they are both 400W. But the strangest thing I found is that the actual output of the power supplies seemed to have little relevance to their rating. So where do they get these wattage ratings from? George |
george12 (7) | ||
| 290086 | 2004-11-10 08:52:00 | Obvious - some are input rating - others output. Bye |
Peter H (220) | ||
| 290087 | 2004-11-10 08:54:00 | isnt this from the other thread ? | ApeNz (4220) | ||
| 290088 | 2004-11-10 09:07:00 | i think he didnt wanna hijack the other tread | Prescott (11) | ||
| 290089 | 2004-11-10 09:21:00 | Good post George. It was getting lost in the other thread. My CPU (stats from EHE): Maximum Power 68.3 - 76.8 W (depending on clock speed) My Enermax 350W EG365P-VE (www.enermax.com.tw) Which has 32amps on the 3.3v rail 3.3v x 32a = 105.6W So that's got enough juice for the CPU it looks like :) It can also pull 32A on the +5v rail and 26A on the +12v rail. Respectable? |
alphazulusixeightniner (185) | ||
| 290090 | 2004-11-10 09:31:00 | +3.3V @ 32A = 105.6W +5V @ 32A = 160W +12V @ 26A = 312W -5V @ 1.0A = 5W -12V @ 1.0A = 12W +5Vsb @ 2.2A = 11W = 605.6W That's using the maximum values for each rail. See I don't think the PSU could pull the maximum for each rail all at once... It must use a combination of all of them up to a total of 350W? |
alphazulusixeightniner (185) | ||
| 290091 | 2004-11-10 10:09:00 | "Which has 32amps on the 3.3v rail 3.3v x 32a = 105.6W So that's got enough juice for the CPU it looks like " If you've got recent technology it's more likely to be 12V feeding the CPU regulator on the MB. |
PaulD (232) | ||
| 290092 | 2004-11-10 10:14:00 | "Let's take the 3.3V rail for example (important as it powers the CPU)." Only on old comps, went to +5 then +12 (Any MB with P4 connector uses +12 for CPU) Make yourself less dangerous by reading this www.quepublishing.com |
PaulD (232) | ||
| 290093 | 2004-11-10 11:55:00 | Yep I'm using Athlon XP which uses the 4plug ATX connector on the motherboard. That makes a bit more sense, cheers Paul. | alphazulusixeightniner (185) | ||
| 290094 | 2004-11-10 23:49:00 | Your 400W supply is probably a 400VA supply. That's (rms volts in times rms amps in times the power factor) plus the losses in the regulated DC side. That's how you only get 300W out. I |
Graham L (2) | ||
| 1 | |||||