[QUOTE]OST are often used by Seasonic, but they don't have the best reputation.[/QUOTE]
I'd amend that comment by adding "not uncommon for lower end Seasonic power supplies" to that because I defy you to find OSX caps in anything other than a lower-end Seasonic build....and in nothing that is placed as coming from a "premium power supply maker."
Anything other than a lower end build from Seasonic will have nothing less than RubyCon, Hitachi, and/or Chemi-Con caps in them....and to imply OSX caps are being used indiscriminately throughout all Seasonic's builds, which that single comment implies, is wholly incorrect.
Again, outside of basic bottom-feeder builds from Seasonic, such as an Antec Earthwatt, I defy you to find a single OSX cap anywhere in their builds.....Corsair certainly has none in any of theirs, Seasonic uses none in anything of 500W or greater capacity, and to find OSX caps in a PCP&C power supply, no matter the wattage, is to see the influence of OCZ finally rearing its ugly head in cost cutting. To save a couple of dollars in what it costs Seasonic to use OSX over Chemi-Con or Hitachi or the like is troubling, esp. given PCP&C's marketing position as a "premium" product.
all of the brands he mentioned were made by Seasonic.
The capacitor selection was also most likely the choice of PC P&C and Antec, not Seasonic.
Just like it was likely Antec's choice to use Fuhjyyu capacitors on the CWT built units.
What C'Dale Rider is objecting to is the author's assertion that it is somehow Seasonic's choice to use questionable capacitors, when it is more likely the choice of the company they make the unit for.
Corsairs are made by Seasonic and they use Hitachi and Nippon Chemicon capacitors..as do Seasonic's own branded PSUs.
Depends on the budget for the product. But don't think the buyer is going to search for components and probably even purchases them himself. The manufacturer chooses them according to budget and MAYBE presents a selection to choose from.
Is there any particular reason that all of the PSU reviews at Anandtech limit efficiency ratings to outputs of about 50W and higher?
Like it or not, I rarely shut my computer off. A typical week is probably 10 hours running and 158 hours in standby. For all I know, my power consumption during standby may exceed consumption during actual usage, depending on how low the efficiency during standby is. However, the point is that I don't know since there's no data available.
Just making it up, assume that power supply output during standby is 2W but efficiency is only 10%. That would be about 14 kilowatt-hours consumed during a month.
Given the trend to "green" electronics, it wouldn't hurt to include this spec in your reviews.
No, it is always true that standby uses less power. Even if efficiency goes down it is always inherant that it be using less total power FOR that efficiency to go down (it is the cause of efficiency going down).
Therefore, you are basically worrying about nothing important. If power consumption itself is important then hibernate, don't go only into standby.
Hi, I had standby efficiency ratings in the beginning but nobody seemed to be interested in it. Standby efficiency is never really good but i can surely include then in future again. :)
Also, I think it would be cool if you guys did an article (or blog, I'm not picky) about computers vs the electric bill. For example, how to calculate how much of an effect a given computer will have on electricity usage. Like, add up the power usage of graphics cards, CPUs, etc., in use and at idle, and factor in the time it's on/standing by/off.
I'm still in shock that for the month out of the year that my computer was out of commission (i broke it good). My electric bill was only about 1/2 - 2/3rds of what it normally is. That's a huge difference even though I live in a small condo. And you could bring it back around to the performance side of things by showing what a difference overclocking can make on power usage.
I know it's a depressing subject, but efficiency and envirenment-consciousness are important to a lot of people right now.
The Anandtech component reviews which include power numbers are IIRC done at the wall, so you can look at those for a variety of power draw scenarios. How long you use the system and how heavily you use it, how large and power-efficient your monitor is, etc. will all influence your power usage.
They have stated before they would like to come out with a guide for power draw for some common components, but that is probably more of a back-burner item.
They really should drop the 'Silencer' name from their PSUs that are noisy. I've owned two Silencer models in the past and they were among the noisiest PSUs I've ever had. As a consumer it's nice to have some truth in advertising, especially from a reputable company.
Sorry your Silencer is not quiet, mine is virtually impossible to hear over the hard drive inside the case (and this is a hard drive that I pulled from a Tivo2, so it is a fairly quiet drive itself.)
My Silencer 470 is so quiet that I have to double and triple check that the fan is even moving.
I notice every time these PC Power and Cooling reviews come out there are noise bashing comments and no mention of the load or temperature that you are exposing them to. Please qualify your comments with some observed facts.
It is always possible my system is drawing less, but I doubt it.* Specs change, but currently I run a P5NE-SLI with a 7300GT (formerly 8800GTS 320MB, no difference, even with some overclocking and under 5-hour gaming sessions with multi-sampling and shader generation by the game as part of level creation.)
2GB DDR2, 80GB WD, Celeron 430 @ 3.17Ghz.
I have at times run an overclocked 7900GS (650mhz core, voltmod) on a 2.6Ghz Athlon 3200+ with a DFI Infinity SLI-M2 and a 6600GT for extra screens, same DDR2 for memory, also an HD tuner card and another HDD have been used at the same time.
The case I am running now has top-to-bottom mesh for the front, only USB and front audio at the very bottom, and a DVD at the top, 9 bays worth of mesh run from top to bottom. Scythe 120mm intake at the bottom. Scythe Infinity with 120mm in the middle for the CPU, and 92mm AC exhuast fan. I try to keep the heat away from the PCPower as much as possible.
I look forward to trying a Dual/Quad overclocked and an nvidia 200 series card overclocked with a voltmod to see if it is too loud, currently I cannot afford that, but prices drop, so I will in the future. I suppose I could put my 7900GS OC back in and load it and the 7300 with Ati-tool (the 7300 is long-overdue for a mod to 7600GT clocks, just need to get the cooling and volt-mod out of the way.), chuck in another hard drive, and start recording off of the HD tuner. But why bother? It is quiet, and my system is drawing plenty of power for it to ramp up, my room gets 90°F+, so I don't think noise is a problem.
I am not saying that they are all good, for all I know there is different software, or different settings, or somebody threw a loud fan in from the wrong bin on the assembly line. More likely the specs change over time, and all systems are different. I can only report on the unit I have with the hardware I am using.
Why is there no more reporting about the quality of the DC output, Ripple and Noise, in your power supply reviews??? I think the last time ripple was reported was in Nov 2007 in the 400-450w PSU Roundup.
Is there a problem in measurement??? If not shouldn't there at least be a comment about current output quality???
That was actually the first one I didnt include a comment about it. But no worries, if something would have been wrong be sure i would have added it to the list.
I find Ripple/Noise also very important but since nobody said a word I thought I could leave the graphics out and just add a comment. I always needed quite some time for the ripple graphs. I will add at least the comments again in the future and new models I get will have the ripple graphs again... alright? It's just there are not many people actually understanding what this topic is about and we need to still keep it accessible. No worries, I am working 7 years with power supplies and had several product managing positions in several companies. I know it's an important topic :)
Since I am just a layman comments are the most important IMO unless a graph is needed to make a point about the power's quality. Reporting the results can help decide if one PSU is worth a premium in price or is a steal compared to other products. Of course only if the better power supply reviewers measure the same way so we can rely on the results and compare reviews from different sites. There are only a few of you on the net and am sure you are at least aware of each other. You all have methodolgy articles post your procedures there.
Your opinion of the results would be most beneficial qualifying the results. I know the lower the number the better the results but context of the results is needed. As an example a maximum of 40mV of noise on the 12 volt rail at full power is a great result but is 95mV at full power on another PSU still OK because at 50% power the ripple and noise is 40mV???
Please let importance dictate what you report not unware opinion or no comments. :)
I have only realized this year about the importance of clean power while reading reviews to make a decision of which PSU to purchase. So I am still trying to understand how important it is and which power supplies provide it.
I´m also positive to the importance of reporting ripple figures, the standard is max.50mV ripple but lately there is a lot of psus sold happily stating 150mV ripple in their specs even though it´s out of the standard.
If such companies voluntarily reports 150mV one can only guess how high the real figures are. Or how many more companies out there are a little bit smarter and state 50mV ripple even though the real figures are mabe ~200mV.
High ripple can damage many parts in the computer, lowers performance and rises temperature. So, yes, it's pretty important and should be mentioned in all psu reviews. Not mentioning it is no guarantee that the performance is really great.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
22 Comments
Back to Article
irev210 - Wednesday, June 11, 2008 - link
Is it just me or is this just an overpriced antec earthwatts 500W PSU?This is EXACTLY the same internals as earthwatts at a 50% price premium.
How depressing :(
C'DaleRider - Tuesday, June 10, 2008 - link
[QUOTE]OST are often used by Seasonic, but they don't have the best reputation.[/QUOTE]I'd amend that comment by adding "not uncommon for lower end Seasonic power supplies" to that because I defy you to find OSX caps in anything other than a lower-end Seasonic build....and in nothing that is placed as coming from a "premium power supply maker."
Anything other than a lower end build from Seasonic will have nothing less than RubyCon, Hitachi, and/or Chemi-Con caps in them....and to imply OSX caps are being used indiscriminately throughout all Seasonic's builds, which that single comment implies, is wholly incorrect.
Again, outside of basic bottom-feeder builds from Seasonic, such as an Antec Earthwatt, I defy you to find a single OSX cap anywhere in their builds.....Corsair certainly has none in any of theirs, Seasonic uses none in anything of 500W or greater capacity, and to find OSX caps in a PCP&C power supply, no matter the wattage, is to see the influence of OCZ finally rearing its ugly head in cost cutting. To save a couple of dollars in what it costs Seasonic to use OSX over Chemi-Con or Hitachi or the like is troubling, esp. given PCP&C's marketing position as a "premium" product.
yyrkoon - Tuesday, June 10, 2008 - link
I find it a bit silly that you ask to 'not soil' your name brand of choice, but then you go and soil about 5 other brand names yourself.Well done ! (not).
HOOfan 1 - Tuesday, June 10, 2008 - link
all of the brands he mentioned were made by Seasonic.The capacitor selection was also most likely the choice of PC P&C and Antec, not Seasonic.
Just like it was likely Antec's choice to use Fuhjyyu capacitors on the CWT built units.
What C'Dale Rider is objecting to is the author's assertion that it is somehow Seasonic's choice to use questionable capacitors, when it is more likely the choice of the company they make the unit for.
Corsairs are made by Seasonic and they use Hitachi and Nippon Chemicon capacitors..as do Seasonic's own branded PSUs.
Christoph Katzer - Wednesday, June 11, 2008 - link
Depends on the budget for the product. But don't think the buyer is going to search for components and probably even purchases them himself. The manufacturer chooses them according to budget and MAYBE presents a selection to choose from.The Irish Patient - Monday, June 9, 2008 - link
Christopher,Is there any particular reason that all of the PSU reviews at Anandtech limit efficiency ratings to outputs of about 50W and higher?
Like it or not, I rarely shut my computer off. A typical week is probably 10 hours running and 158 hours in standby. For all I know, my power consumption during standby may exceed consumption during actual usage, depending on how low the efficiency during standby is. However, the point is that I don't know since there's no data available.
Just making it up, assume that power supply output during standby is 2W but efficiency is only 10%. That would be about 14 kilowatt-hours consumed during a month.
Given the trend to "green" electronics, it wouldn't hurt to include this spec in your reviews.
mindless1 - Monday, June 9, 2008 - link
No, it is always true that standby uses less power. Even if efficiency goes down it is always inherant that it be using less total power FOR that efficiency to go down (it is the cause of efficiency going down).Therefore, you are basically worrying about nothing important. If power consumption itself is important then hibernate, don't go only into standby.
Christoph Katzer - Monday, June 9, 2008 - link
Hi, I had standby efficiency ratings in the beginning but nobody seemed to be interested in it. Standby efficiency is never really good but i can surely include then in future again. :)icingdeath88 - Monday, June 9, 2008 - link
SecondedAlso, I think it would be cool if you guys did an article (or blog, I'm not picky) about computers vs the electric bill. For example, how to calculate how much of an effect a given computer will have on electricity usage. Like, add up the power usage of graphics cards, CPUs, etc., in use and at idle, and factor in the time it's on/standing by/off.
I'm still in shock that for the month out of the year that my computer was out of commission (i broke it good). My electric bill was only about 1/2 - 2/3rds of what it normally is. That's a huge difference even though I live in a small condo. And you could bring it back around to the performance side of things by showing what a difference overclocking can make on power usage.
I know it's a depressing subject, but efficiency and envirenment-consciousness are important to a lot of people right now.
mindless1 - Monday, June 9, 2008 - link
That would hardly require a blog. Just a chart or even rough guess then look at your power bill for the KWH cost and the math is simple.Or look at the power bill to get cost and program that into a Kill-A-Watt meter.
strikeback03 - Monday, June 9, 2008 - link
The Anandtech component reviews which include power numbers are IIRC done at the wall, so you can look at those for a variety of power draw scenarios. How long you use the system and how heavily you use it, how large and power-efficient your monitor is, etc. will all influence your power usage.They have stated before they would like to come out with a guide for power draw for some common components, but that is probably more of a back-burner item.
grunjee - Monday, June 9, 2008 - link
They really should drop the 'Silencer' name from their PSUs that are noisy. I've owned two Silencer models in the past and they were among the noisiest PSUs I've ever had. As a consumer it's nice to have some truth in advertising, especially from a reputable company.nubie - Monday, June 9, 2008 - link
Sorry your Silencer is not quiet, mine is virtually impossible to hear over the hard drive inside the case (and this is a hard drive that I pulled from a Tivo2, so it is a fairly quiet drive itself.)My Silencer 470 is so quiet that I have to double and triple check that the fan is even moving.
I notice every time these PC Power and Cooling reviews come out there are noise bashing comments and no mention of the load or temperature that you are exposing them to. Please qualify your comments with some observed facts.
The fans should ramp up if exposed to high temperatures and load, this should not be counted against them as the case design should take into account proper cooling for the PSU. See this guide from SilentPCReview.com to build a duct to cool the PSU (and prevent Fan ramp up) : http://www.silentpcreview.com/article293-page2.htm...">http://www.silentpcreview.com/article293-page2.htm...
If there is an actual problem with the PSU, you should take it up with the PC Power and Cooling.
Christoph Katzer - Monday, June 9, 2008 - link
Sorry but I am sure your PC has less than 100 watts power consumption. because then it is really quiet. Post your specs, we will see.nubie - Saturday, June 14, 2008 - link
It is always possible my system is drawing less, but I doubt it.* Specs change, but currently I run a P5NE-SLI with a 7300GT (formerly 8800GTS 320MB, no difference, even with some overclocking and under 5-hour gaming sessions with multi-sampling and shader generation by the game as part of level creation.)2GB DDR2, 80GB WD, Celeron 430 @ 3.17Ghz.
I have at times run an overclocked 7900GS (650mhz core, voltmod) on a 2.6Ghz Athlon 3200+ with a DFI Infinity SLI-M2 and a 6600GT for extra screens, same DDR2 for memory, also an HD tuner card and another HDD have been used at the same time.
The case I am running now has top-to-bottom mesh for the front, only USB and front audio at the very bottom, and a DVD at the top, 9 bays worth of mesh run from top to bottom. Scythe 120mm intake at the bottom. Scythe Infinity with 120mm in the middle for the CPU, and 92mm AC exhuast fan. I try to keep the heat away from the PCPower as much as possible.
I look forward to trying a Dual/Quad overclocked and an nvidia 200 series card overclocked with a voltmod to see if it is too loud, currently I cannot afford that, but prices drop, so I will in the future. I suppose I could put my 7900GS OC back in and load it and the 7300 with Ati-tool (the 7300 is long-overdue for a mod to 7600GT clocks, just need to get the cooling and volt-mod out of the way.), chuck in another hard drive, and start recording off of the HD tuner. But why bother? It is quiet, and my system is drawing plenty of power for it to ramp up, my room gets 90°F+, so I don't think noise is a problem.
* according to http://extreme.outervision.com/PSUEngine">http://extreme.outervision.com/PSUEngine my current system is 245W with 89% load on the CPU and 90% system load, which is the default, I didn't futz with that to generate a fake report (overclock generated at 1.3v 3.17ghz).
I am not saying that they are all good, for all I know there is different software, or different settings, or somebody threw a loud fan in from the wrong bin on the assembly line. More likely the specs change over time, and all systems are different. I can only report on the unit I have with the hardware I am using.
Merman - Monday, June 9, 2008 - link
Why is there no more reporting about the quality of the DC output, Ripple and Noise, in your power supply reviews??? I think the last time ripple was reported was in Nov 2007 in the 400-450w PSU Roundup.Is there a problem in measurement??? If not shouldn't there at least be a comment about current output quality???
Christoph Katzer - Monday, June 9, 2008 - link
That was actually the first one I didnt include a comment about it. But no worries, if something would have been wrong be sure i would have added it to the list.Merman - Monday, June 9, 2008 - link
Christoph is your comment condescending or an indication of the importance of Ripple and Noise???Christoph Katzer - Monday, June 9, 2008 - link
I find Ripple/Noise also very important but since nobody said a word I thought I could leave the graphics out and just add a comment. I always needed quite some time for the ripple graphs. I will add at least the comments again in the future and new models I get will have the ripple graphs again... alright? It's just there are not many people actually understanding what this topic is about and we need to still keep it accessible. No worries, I am working 7 years with power supplies and had several product managing positions in several companies. I know it's an important topic :)Merman - Tuesday, June 10, 2008 - link
Since I am just a layman comments are the most important IMO unless a graph is needed to make a point about the power's quality. Reporting the results can help decide if one PSU is worth a premium in price or is a steal compared to other products. Of course only if the better power supply reviewers measure the same way so we can rely on the results and compare reviews from different sites. There are only a few of you on the net and am sure you are at least aware of each other. You all have methodolgy articles post your procedures there.Your opinion of the results would be most beneficial qualifying the results. I know the lower the number the better the results but context of the results is needed. As an example a maximum of 40mV of noise on the 12 volt rail at full power is a great result but is 95mV at full power on another PSU still OK because at 50% power the ripple and noise is 40mV???
Please let importance dictate what you report not unware opinion or no comments. :)
I have only realized this year about the importance of clean power while reading reviews to make a decision of which PSU to purchase. So I am still trying to understand how important it is and which power supplies provide it.
Thank you for your consideration.
xeizo - Monday, June 9, 2008 - link
I´m also positive to the importance of reporting ripple figures, the standard is max.50mV ripple but lately there is a lot of psus sold happily stating 150mV ripple in their specs even though it´s out of the standard.If such companies voluntarily reports 150mV one can only guess how high the real figures are. Or how many more companies out there are a little bit smarter and state 50mV ripple even though the real figures are mabe ~200mV.
High ripple can damage many parts in the computer, lowers performance and rises temperature. So, yes, it's pretty important and should be mentioned in all psu reviews. Not mentioning it is no guarantee that the performance is really great.
HOOfan 1 - Monday, June 9, 2008 - link
50mV is the max specification on the 3.3V and 5V rails. The max spec on the 12V rails is 120mV.