How to calculate PSU load right?

Message boards : Number crunching : How to calculate PSU load right?
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Treasurer

Send message
Joined: 13 Dec 05
Posts: 109
Credit: 1,569,762
RAC: 0
Germany
Message 1172597 - Posted: 20 Nov 2011, 23:03:01 UTC

The optimal PSU load is between 50% and 80%(or slightly different), but % of what? Of the 12V rail (example: if there are 10A @12V so the optimal load is between 5A and 8A) or of the total 400W the PSU is capable of. And if there is a separate 12V rail for the 4pin and/or 8pin CPU plug, is this power used for the CPU only or is the board also using it for other parts?
ID: 1172597 · Report as offensive
Profile ML1
Volunteer moderator
Volunteer tester

Send message
Joined: 25 Nov 01
Posts: 20147
Credit: 7,508,002
RAC: 20
United Kingdom
Message 1172620 - Posted: 21 Nov 2011, 2:05:45 UTC - in response to Message 1172597.  

The optimal PSU load is between 50% and 80%(or slightly different), but % of what? ...

That is for each individual supply rail from the PSU.


The usual construction is that you have separate switch-mode units, one per supply rail, that are constructed side-by-side on the same circuit board but work independently. They just happen to be all in the same box and are fed by the same mains feed.

Also beware the "split rail" PSUs that have for example two 12V rails, each with an individual switch-mode unit. They can be difficult to plug in your peripherals so that you balance the load between the two rails nicely.


Happy crunchin',
Martin

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)
ID: 1172620 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1172675 - Posted: 21 Nov 2011, 10:30:57 UTC - in response to Message 1172597.  

The optimal PSU load is between 50% and 80%(or slightly different), but % of what? Of the 12V rail (example: if there are 10A @12V so the optimal load is between 5A and 8A) or of the total 400W the PSU is capable of. And if there is a separate 12V rail for the 4pin and/or 8pin CPU plug, is this power used for the CPU only or is the board also using it for other parts?

The ratings for efficiency are given for the input of the PSU. At either 120v or 240v. These ratings are under optimal conditions. So your mileage may vary. Loading sections of the PSU more than others could produce results other than the ratings given by the manufacture.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1172675 · Report as offensive
Treasurer

Send message
Joined: 13 Dec 05
Posts: 109
Credit: 1,569,762
RAC: 0
Germany
Message 1172778 - Posted: 21 Nov 2011, 20:09:44 UTC - in response to Message 1172675.  

Maybe thats my insufficient English but,

That is for each individual supply rail from the PSU


The ratings for efficiency are given for the input of the PSU


are that not the two possibilities i was confused about? What is true?
ID: 1172778 · Report as offensive
Horacio

Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,967,266
RAC: 0
Argentina
Message 1172801 - Posted: 21 Nov 2011, 21:14:08 UTC - in response to Message 1172778.  

Maybe thats my insufficient English but,

That is for each individual supply rail from the PSU


The ratings for efficiency are given for the input of the PSU


are that not the two possibilities i was confused about? What is true?


In fact both things are true with some extra details:

The Input and the output are ussually very close to been directly proportional, so a certain percentage applied to the imput will give you the same results that applied to the output.

When a PSU states its 850W (or whatever) thats the total power the PSU will be able to handle. As part of that power will be lost as heat then thats why you need to use the efficiency factor.

But also, you cant use the whole "efficient" power in just one rail, you need to find the specifications of the PSU to see how much power can supply on each of the rails and to be safe you should not try to get more than an 80% of the specs from each rail

(80% if it is good PSU, may be less if not so good. Anyway, there is no a general consensus on where the safe point is, as it varies even for different PSUs of the same brand/model...)




ID: 1172801 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1172807 - Posted: 21 Nov 2011, 21:38:50 UTC - in response to Message 1172778.  
Last modified: 21 Nov 2011, 21:41:11 UTC

Maybe thats my insufficient English but,

That is for each individual supply rail from the PSU


The ratings for efficiency are given for the input of the PSU


are that not the two possibilities i was confused about? What is true?


If we take a 650w PSU that is specified to be 90% efficient at 50% load. Then when the load on the output is 325w the input would be about 389w in their test conditions.

If you were to have a system. With say an Atom D525 and a NVIDIA 590 that drew 325w from the PSU I mentioned before. Then the rails would probably not be very evenly loaded. With the 5v & 3.3v sections with very little load and the 12v with most of the load.
With the 12v section being the greatest amount of load and having a larger load. Where its efficiency started to drop. I would think the overall efficiency would be <90% as the manufacture rated it.

Depending on how the PSU is actually designed the outcome could be different.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1172807 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22160
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1172821 - Posted: 21 Nov 2011, 22:12:56 UTC

As PSU manufacturers rate their power supplies in so many different ways there is no "right" answer.

One wrecked PSU I have in front of me just now claims to be rated at a maximum of 300W. And this is clearly not right because if you add up the individual power rails you come to something very different (a shade over 400W).
For the same PSU if I take the input power I come to a figure of 920W.
So what was the rating of this PSU?
Well I can tell you that it failed when I did a load test on it, at about 310W total, with the 12V rail running at 160W and the 5V rails at about 150W, the rest being a small loads on the -12V and the 3.3V rails.
Just before the failure the input was running slightly over 400VA, which gives an "efficiency" of about 77%.

I do love the smell of burning power supplies in the morning...
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1172821 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1172822 - Posted: 21 Nov 2011, 22:13:37 UTC

This consideration of efficiency is actually the sticking point that finally solved Steve's ( http://setiathome.berkeley.edu/show_user.php?userid=202207 ) issues getting 2 x GTX 480's stable while running the CPUs full bore. IIRC he had something like a 1000W PSU, which 'should' have been enough, but factoring in efficiencies appeared very much 'borderline'. We actually worked out individual theoretical RAIL currents & the input current capability came up short. Apparently updating to a high efficiency 'overkill' unit solved that round of issues.

It's a good idea to further add about 30% (or even more) headroom allowing for component degradation over time, if you want to last at least a few years... That does tend to move DIY machine PSU selection into the enthusiast category very quickly even with a relatively low power machine.

Jason
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1172822 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1172852 - Posted: 21 Nov 2011, 23:46:24 UTC - in response to Message 1172822.  

This consideration of efficiency is actually the sticking point that finally solved Steve's ( http://setiathome.berkeley.edu/show_user.php?userid=202207 ) issues getting 2 x GTX 480's stable while running the CPUs full bore. IIRC he had something like a 1000W PSU, which 'should' have been enough, but factoring in efficiencies appeared very much 'borderline'. We actually worked out individual theoretical RAIL currents & the input current capability came up short. Apparently updating to a high efficiency 'overkill' unit solved that round of issues.

It's a good idea to further add about 30% (or even more) headroom allowing for component degradation over time, if you want to last at least a few years... That does tend to move DIY machine PSU selection into the enthusiast category very quickly even with a relatively low power machine.

Jason

Thankfully, If I wanted to, I could do cpu work on 2 cpu cores out of 4, 1 is for Cuda AP and 1 is to load Cuda MB, But If You want overkill, I don't know what floats Steve's boat, But for the moment I'm using a Silverstone ST1500 1500w psu, It's more than enough psu for My present PC, It was what I had on hand, the 1050w Enermax needs some repairs to the 24 pin main cable, otherwise It runs, 2-12v connections got toasted on the 1050w.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1172852 · Report as offensive
Profile Lint trap

Send message
Joined: 30 May 03
Posts: 871
Credit: 28,092,319
RAC: 0
United States
Message 1172999 - Posted: 22 Nov 2011, 13:08:52 UTC
Last modified: 22 Nov 2011, 13:27:18 UTC

One wrecked PSU I have in front of me just now claims to be rated at a maximum of 300W. And this is clearly not right because if you add up the individual power rails you come to something very different (a shade over 400W).


A typical 600 Watt psu is a fairly simple example:




Reading from Bottom to Top the values shown determine the combined entire output possible above it.

In this example the +12v rails are rated to provide 420 Watts maximum, even though individually the maximum output would be higher, which means both rails cannot provide maximum output at the same time.

Different manufacturers rate differently. Occasionally different model lines are rated differently from the same manufacturer. Are the ratings for room temp (25c) or full load (50c) or somewhere in between? Sometimes the info is provided or is available if asked for, but not always.

The closer the overall load gets to maximum value, the more heat and less efficient the PSU will be. At 80% efficiency, a fully loaded 600 W supply would need 750 Watts from the wall!

The only constants are that heat and load are the biggest enemies of a PSU.

Lt

edited...
ID: 1172999 · Report as offensive
Profile ML1
Volunteer moderator
Volunteer tester

Send message
Joined: 25 Nov 01
Posts: 20147
Credit: 7,508,002
RAC: 20
United Kingdom
Message 1173013 - Posted: 22 Nov 2011, 14:40:32 UTC - in response to Message 1172999.  
Last modified: 22 Nov 2011, 14:50:35 UTC

One wrecked PSU I have in front of me just now claims to be rated at a maximum of 300W. And this is clearly not right because if you add up the individual power rails you come to something very different (a shade over 400W).


A typical 600 Watt psu is a fairly simple example:




Reading from Bottom to Top the values shown determine the combined entire output possible above it.

In this example the +12v rails are rated to provide 420 Watts maximum, even though individually the maximum output would be higher, which means both rails cannot provide maximum output at the same time.

Different manufacturers rate differently. ...

... At 80% efficiency, a fully loaded 600 W supply would need 750 Watts from the wall!

The only constants are that heat and load are the biggest enemies of a PSU. ...

That is an unusually detailed specification label for a PSU and quite interesting...

Note that the maximum rated input power is 1380 Watts[*]...!

The combined power ratings for multiple outputs suggest that either an intermediate voltage drop-down stage is used for those multiple outputs, or shared magnetic circuits or shared heatsinks are used.

An important point though is:

Do not confuse percentage efficiency with a de-rating percentage. You should be able to run any of those outputs at their maximum rated current. The less than 100% efficiency tells you what to expect for the extra input power required that is wasted as heat.

So... For running 600 Watts maximum output with 1380 Watts input, you have your PSU generating 1380 - 600 = 780 Watts as waste heat. (I'd hope that in reality it was not that badly inefficient!)


Note that I strongly NOT recommend running any one output at 100% rating. That gives you no margin for load fluctuation or for ageing or any other de-rating effects. A good rule of thumb is to try to keep loadings in the 50% to 80% region.


Happy cool crunchin',
Martin


* That would be the case if the maximum input current was taken for the nominal mains input voltage. Actually, the maximum input current can be expected only during 'brown-out' conditions when the mains input voltage is lower than nominal. The maximum input current can be expected for full output load and minimum input voltage.
See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)
ID: 1173013 · Report as offensive
Profile SciManStev Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Jun 99
Posts: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1173022 - Posted: 22 Nov 2011, 15:02:25 UTC

I just read this thread and decided to post a comment. The original PSU I was using was a BFG 1250 watt, with 28 Amp rails. As Jason noted, I was, and still am using all 6 cores from my CPU 980, non-hyperthreading, to crunch. By switching to a Corsair AX1200 single rail PSU, my overall wattage was reduced by 50 watts, but as I now have a single, 100.4 Amp 12 Volt rail, all of my power problems went away. Now I can pull all the power I need from the single rail, without ever running into shortages.

When the rail limit is exceeded, the only thing that can happen is the voltage drops, and the output can become very noisey. That can cause system crashses, or other wierdities, some of which are hard to find.

The Silverstone 1500 is a very powerful supply, but it has 8, 25 Amp rails. If any of thoise individual rails get near the limit, the PSU will act as I described above. As long as each device connected to it draws some proportion less current than 25 Amps, leaving plenty of headroom of course, it should be fine. My 480's, being overclocked as they are would push that beyond the limit.

Before I had that BFG 1250 watt PSU, I ran an 850 watt from a vendor I don't remember. That PSU had noise on the lines that affected my stability with my overclocks. That was before Fermi GPU's were introduced, and I had a GTX 295. It turned out that due to the noise the PSU was putting out on it's outputs, every time I used a Killa-Watt meter, my system Blue Screened. Without the Killawatt meter is was fine. This was only a problem when overclocking the CPU.

At any rate, I do reccomend a single rail power supply with enough headroom to not be maxed out. It does make a difference.

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1173022 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1173023 - Posted: 22 Nov 2011, 15:07:14 UTC - in response to Message 1173013.  

One wrecked PSU I have in front of me just now claims to be rated at a maximum of 300W. And this is clearly not right because if you add up the individual power rails you come to something very different (a shade over 400W).


A typical 600 Watt psu is a fairly simple example:




Reading from Bottom to Top the values shown determine the combined entire output possible above it.

In this example the +12v rails are rated to provide 420 Watts maximum, even though individually the maximum output would be higher, which means both rails cannot provide maximum output at the same time.

Different manufacturers rate differently. ...

... At 80% efficiency, a fully loaded 600 W supply would need 750 Watts from the wall!

The only constants are that heat and load are the biggest enemies of a PSU. ...

That is an unusually detailed specification label for a PSU and quite interesting...

Note that the maximum rated input power is 1380 Watts[*]...!

The combined power ratings for multiple outputs suggest that either an intermediate voltage drop-down stage is used for those multiple outputs, or shared magnetic circuits or shared heatsinks are used.

An important point though is:

Do not confuse percentage efficiency with a de-rating percentage. You should be able to run any of those outputs at their maximum rated current. The less than 100% efficiency tells you what to expect for the extra input power required that is wasted as heat.

So... For running 600 Watts maximum output with 1380 Watts input, you have your PSU generating 1380 - 600 = 780 Watts as waste heat. (I'd hope that in reality it was not that badly inefficient!)


Note that I strongly NOT recommend running any one output at 100% rating. That gives you no margin for load fluctuation or for ageing or any other de-rating effects. A good rule of thumb is to try to keep loadings in the 50% to 80% region.


Happy cool crunchin',
Martin


* That would be the case if the maximum input current was taken for the nominal mains input voltage. Actually, the maximum input current can be expected only during 'brown-out' conditions when the mains input voltage is lower than nominal. The maximum input current can be expected for full output load and minimum input voltage.


The current rating you see for input is normally a value called "max inrush current". Which will be seen for microseconds when the unit is first turned on. It is not an "input power" rating, and should not be confused with any such thing.

A standard 60w 230v light bulb may have an inrush current of 4 amp, but it isn't sitting there consuming 920w.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1173023 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1173081 - Posted: 22 Nov 2011, 22:46:19 UTC - in response to Message 1173022.  
Last modified: 22 Nov 2011, 22:59:03 UTC

I just read this thread and decided to post a comment. The original PSU I was using was a BFG 1250 watt, with 28 Amp rails. As Jason noted, I was, and still am using all 6 cores from my CPU 980, non-hyperthreading, to crunch. By switching to a Corsair AX1200 single rail PSU, my overall wattage was reduced by 50 watts, but as I now have a single, 100.4 Amp 12 Volt rail, all of my power problems went away. Now I can pull all the power I need from the single rail, without ever running into shortages.

When the rail limit is exceeded, the only thing that can happen is the voltage drops, and the output can become very noisey. That can cause system crashses, or other wierdities, some of which are hard to find.

The Silverstone 1500 is a very powerful supply, but it has 8, 25 Amp rails. If any of thoise individual rails get near the limit, the PSU will act as I described above. As long as each device connected to it draws some proportion less current than 25 Amps, leaving plenty of headroom of course, it should be fine. My 480's, being overclocked as they are would push that beyond the limit.

Before I had that BFG 1250 watt PSU, I ran an 850 watt from a vendor I don't remember. That PSU had noise on the lines that affected my stability with my overclocks. That was before Fermi GPU's were introduced, and I had a GTX 295. It turned out that due to the noise the PSU was putting out on it's outputs, every time I used a Killa-Watt meter, my system Blue Screened. Without the Killawatt meter is was fine. This was only a problem when overclocking the CPU.

At any rate, I do reccomend a single rail power supply with enough headroom to not be maxed out. It does make a difference.

Steve

I hear Ya Steve, But I don't have a lot of choice and the 1500w is doing pretty well, It will be replaced with either a 1200w single rail psu or 1350w Enermax Maxrevo, But that may not be until October 2012. But then I'm building the ultimate crunching PC(12 gpus), the prototype here(6 gpus) is doing very good now. Starting in Feb and going until Aug though I'll be trying to save up nearly $1,400.00 for moving and utility deposits so that I might be able to buy a house.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1173081 · Report as offensive
Profile ML1
Volunteer moderator
Volunteer tester

Send message
Joined: 25 Nov 01
Posts: 20147
Credit: 7,508,002
RAC: 20
United Kingdom
Message 1173228 - Posted: 23 Nov 2011, 14:10:02 UTC - in response to Message 1173023.  
Last modified: 23 Nov 2011, 14:11:43 UTC

... A standard 60w 230v light bulb may have an inrush current of 4 amp, but it isn't sitting there consuming 920w.

Quite so. Sometimes, that inrush current can cause a cold bulb filament to immediately fail.

However, do not confuse the effects of a cold and heat dependant resistive filament for how an inductive switch mode power supply operates. Also do not confuse the old style "inrush" (high) currents for charging (big) smoothing capacitors at start up.


For PSUs, a brief maximum current load can be expected at start-up, but that is limited by the switch mode circuitry and by the magnetic circuit in the inductor or transformer.

Look at those power numbers again for the example given. At maximum load, an efficiency of 50% is not unexpected.

As hinted in my original comment, I wouldn't expect the PSU to be that bad. Also note that those input currents will be for minimum acceptable input voltage and note that "Watts = Volts * Amps" so you'll see 'better' power figures than my (shouldn't be seen) dumb worse case assumption.

However, it does give a good example for people to work out the specs ;-)


Happy cool crunchin',
Martin
See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)
ID: 1173228 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1173279 - Posted: 23 Nov 2011, 17:49:23 UTC - in response to Message 1173228.  

... A standard 60w 230v light bulb may have an inrush current of 4 amp, but it isn't sitting there consuming 920w.

Quite so. Sometimes, that inrush current can cause a cold bulb filament to immediately fail.

However, do not confuse the effects of a cold and heat dependant resistive filament for how an inductive switch mode power supply operates. Also do not confuse the old style "inrush" (high) currents for charging (big) smoothing capacitors at start up.

For PSUs, a brief maximum current load can be expected at start-up, but that is limited by the switch mode circuitry and by the magnetic circuit in the inductor or transformer.

Look at those power numbers again for the example given. At maximum load, an efficiency of 50% is not unexpected.

As hinted in my original comment, I wouldn't expect the PSU to be that bad. Also note that those input currents will be for minimum acceptable input voltage and note that "Watts = Volts * Amps" so you'll see 'better' power figures than my (shouldn't be seen) dumb worse case assumption.

However, it does give a good example for people to work out the specs ;-)

Happy cool crunchin',
Martin

I think you might be nuts to think that PSU would have a 50% rating, but that is OK. :D Different views for different people and what not.

This link may prove useful for those trying to calculate their input current to their PSU.
http://www.enermax.com/home.php?fn=eng/product_a1_1_1&lv0=1&lv1=52&no=182
Under the section "89PLUS ready!". They are showing how their PSU uses less power compared to competitors with the platinum rating, but also demonstrates how to calculate the power draw based on the efficiency of the PSU.

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1173279 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1173281 - Posted: 23 Nov 2011, 17:57:32 UTC - in response to Message 1173279.  

That says my system requires a 625W. I have a 800W so I'm safe for now


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1173281 · Report as offensive
Profile Lint trap

Send message
Joined: 30 May 03
Posts: 871
Credit: 28,092,319
RAC: 0
United States
Message 1173328 - Posted: 23 Nov 2011, 21:27:08 UTC

That is an unusually detailed specification label for a PSU and quite interesting...


The label is from a ~4yo PSU made by Ultra.

Lt

ID: 1173328 · Report as offensive
Profile ML1
Volunteer moderator
Volunteer tester

Send message
Joined: 25 Nov 01
Posts: 20147
Credit: 7,508,002
RAC: 20
United Kingdom
Message 1173344 - Posted: 23 Nov 2011, 22:16:59 UTC - in response to Message 1173279.  
Last modified: 23 Nov 2011, 22:20:33 UTC

I think you might be nuts to think that PSU would have a 50% rating, but that is OK. :D Different views for different people and what not.

This link may prove useful for those trying to calculate their input current to their PSU.
http://www.enermax.com/home.php?fn=eng/product_a1_1_1&lv0=1&lv1=52&no=182
Under the section "89PLUS ready!". They are showing how their PSU uses less power compared to competitors with the platinum rating, but also demonstrates how to calculate the power draw based on the efficiency of the PSU.

Well... If you will go for a top-of-the-range PSU with "World’s leading modular cable PSU series with 89-94% efficiency @ 20-100% load."

For that one, it is good to see some real specs and charts for load and ambient temperature. Not sure on what the rest of their Marketing-speak is trying to exclaim. So yep, that's a nice one that might actually pay for itself in avoiding waste heat if you will be running it for long enough. Note the load efficiency graph lower down the page and compare with other PSUs operating with poorer design margins and poorer cooling.

Indeed, at the other end of the spectrum a certain review tested a range of PSUs for their claimed ratings. Some blew up before reaching the claimed max ratings...

Such is the real world...

Happy cool crunchin',
Martin
See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)
ID: 1173344 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1173392 - Posted: 24 Nov 2011, 1:23:45 UTC - in response to Message 1173344.  

I think you might be nuts to think that PSU would have a 50% rating, but that is OK. :D Different views for different people and what not.

This link may prove useful for those trying to calculate their input current to their PSU.
http://www.enermax.com/home.php?fn=eng/product_a1_1_1&lv0=1&lv1=52&no=182
Under the section "89PLUS ready!". They are showing how their PSU uses less power compared to competitors with the platinum rating, but also demonstrates how to calculate the power draw based on the efficiency of the PSU.

Well... If you will go for a top-of-the-range PSU with "World’s leading modular cable PSU series with 89-94% efficiency @ 20-100% load."

For that one, it is good to see some real specs and charts for load and ambient temperature. Not sure on what the rest of their Marketing-speak is trying to exclaim. So yep, that's a nice one that might actually pay for itself in avoiding waste heat if you will be running it for long enough. Note the load efficiency graph lower down the page and compare with other PSUs operating with poorer design margins and poorer cooling.

Indeed, at the other end of the spectrum a certain review tested a range of PSUs for their claimed ratings. Some blew up before reaching the claimed max ratings...

Such is the real world...

Happy cool crunchin',
Martin

I have used Enermax or Antec supplies for many years. Enermax does provide a good bit of information about all of their products, which I find helpful. I used that one as an example as I was looking at it recently for my next system.

Toms hardware has a nice section for power supply reviews. That could be good for answering questions or providing useful info when people go to buy a new supply.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1173392 · Report as offensive
1 · 2 · Next

Message boards : Number crunching : How to calculate PSU load right?


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.