Jump to content

Talk:Power supply unit (computer): Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Browse history interactively
← Previous edit
Content deleted Content added
VisualWikitext
→top: Added automated archiving configuration
m Removed deprecated parameters in {{Talk header}} that are now handled automatically (Task 30)
 
(43 intermediate revisions by 21 users not shown)
Line 1: Line 1:
{{Talk header}}
{{Talk header}}
{{Auto archiving notice }}
{{WikiProject Computing|class=C|importance=mid|hardware=yes|hardware-importance=high}}
{{WikiProject banner shell|class=C|
{{Auto archiving notice |age=90 |units=days |bot=Lowercase sigmabot III}}
{{WikiProject Computing|importance=mid|hardware=yes|hardware-importance=high}}
}}
{{User:MiszaBot/config
{{User:MiszaBot/config
| archiveheader = {{Talk archive navigation}}
| archiveheader = {{Talk archive navigation}}
Line 11: Line 13:
| archive = Talk:Power supply unit (computer)/Archive %(counter)d
| archive = Talk:Power supply unit (computer)/Archive %(counter)d
}}
}}

== No mention of... ==
*Most PSU I think have like 4x 12v-lines. *Does not mention EPS12V.
*Does not mention Power_Good signal (sometimes called Power_OK or PWR_OK).
*8-pin Molex connectors for modern CPUs? [[User:Story Weaver|Story Weaver]] 00:28, 17 December 2006 (UTC)

Most PSU's don't have 4 independent 12v lines, in fact most high end power supplies have 3x 12v lines, if that. I'm not sure what you're getting at with an "8pin molex connector for CPUs". --[[User:71.113.162.164|71.113.162.164]] 01:27, 5 February 2007 (UTC)

* The latest Video cards mated with the latest Intel processors do benefit from 4 rails (12VDC lines) At this time there are numberous 3 & 4 Rail power supplies available now.
* The EPS12V form factor that needs to be added, has the 8pin molex connector.
* The BTX form factor needs to be added, even if it will not be widely implimented.
* I would like to see the Mechanical specifications (Size of, Form Factors) for all of the various types. Some are cross compatable and usable for upgrading an old chassis. IE: The BTX or EPS12V may fit into your favorite Micro-ATX case.
* According to this site they have a product that covers 3 form factors ATX, BTX & EPS12v: http://www.highpowersupply.com/hpc620A12s.html<br />
Thanks for the cheese.

== Ideas for improvement ==
* General clean up.
* Get some wiring diagrams of the different connectors.
* Clarify distinction between different connectors, and list standards such as EPS+12v etc.
Any other ideas would be appreciated, as well as consideration of this list. [[User:Damien Shiest|Damien Shiest]] 20:26, 28 November 2006 (UTC)

The "Things to consider" section needs to be changed from a list to something else, right now it seems very un-encyclopedic. --[[User:71.113.162.164|71.113.162.164]] 01:27, 5 February 2007 (UTC)
:I'd like to add to your list a discussion of pre-ATX power supply designs, such as AT power supplies. I can think of two differences right now: the motherboard power connector (was a pair of connectors, usually wired in parallel, unlike today's 20/24 pin block) and the hardwired power switch. &mdash; [[User:EagleOne|EagleOne]]\<sup>[[User_talk:EagleOne|Talk]]</sup> 03:42, 5 May 2007 (UTC)
::Yes, the addition of that information would be good. I am not sure how to structure the article, have an article for AT and one for ATX, or have just computer power supply, and have chapters for AT and chapter for ATX, or just have it all together. Please feel free to improve the article. -- [[User:Frap|Frap]] 10:02, 5 May 2007 (UTC)
:::Aside of organizing the article, I added a section for the differences between AT and ATX power supplies. Feel free to expand or clarify. [[User:Thief12|Thief12]] 03:15, 12 July 2007 (UTC)

What need is there for any mention of AT power supplies other than ancient history? AT power supplies haven't been relevant for over a decade now. I think any discussion of them belongs in an "Evolution of standards" section, and make it clear to people that AT has not been in active use for some time, which the article currently fails to do. And I also see that the [[ATX]] article already includes pinouts, discussion of the difference between ATX and AT, and a ton of detail on different versions of ATX power supply standards. Either the articles should be merged or all the power supply info in [[ATX]] should be moved into this article given "Computer Power Supply" generally means an ATX PSU these days. [[Special:Contributions/24.179.151.124|24.179.151.124]] ([[User talk:24.179.151.124|talk]]) 08:45, 3 November 2010 (UTC)

== PFC ==
Curious about the power factor corrections... if a power supply does not indicate whether it has PFC or not, is it just "passive PFC" or is it "no PFC?" Secondly, isn't PFC required on all power supplies sold in Europe? ---Ransom (--[[User:71.4.51.150|71.4.51.150]] 23:10, 6 December 2006 (UTC))
:Yes, all PSU sold in Europe must have PFC. So I assume if it does not mention anything about PFC, it is passive PFC. Active PFC is better and less common. -- [[User:Frap|Frap]] 21:53, 11 December 2006 (UTC)

::No, we cannot assume passive PFC for those unindicated PSUs. Some are just regular PSUs without any form of PFC. [[User:Kiwi8|Kiwi8]] 12:56, 18 January 2007 (UTC)

:::Since 2001, the [[IEC]] standard 61000-3-2 came into effect in Europe any new electronic equipment consuming more than 75Watts has to meet certain standards for harmonic content. This basically mandated some form of PFC. All of Britain, Japan and China have adopted similar standards though I know of no standard for the US.

:::Given 2001 isn't that long ago some older PSUs without any PFC may still be sold (illegally now I imagine) in those territories that have these rules on harmonics. ps: as a geek I have opened PSUs and replaced the fans with LED fans for [[case mods]]. Stuff the warranty - see the light. [[User:Ttiotsw|Ttiotsw]] 13:47, 18 January 2007 (UTC)

::::Yes, but as u said, it is in certain countries. In Singapore here we still have regular PSUs without any PFC on sale. Thus the statement about unindicated PSUs implying passive PFC cannot apply across the board. [[User:Kiwi8|Kiwi8]] 14:14, 18 January 2007 (UTC)

== MTBF ==
Looking at the MTBF, I think 100K hours is a bit unrealistic. Antec PSUs, which are higher end, are rated at 80K MTBF. Anyone else have any thoughts? - [[User:MSTCrow|MSTCrow]] 08:29, 7 December 2006 (UTC)

I agree, Antec generally has higher quality PSUs, and rates them very realistically, I think the section on MTBF should be removed, unless it is further explained. --[[User:71.113.162.164|71.113.162.164]] 01:27, 5 February 2007 (UTC)

:I heard Antec have different manufacturer of the supplies for different lines/products. And that some are good, some are bad. -- [[User:Frap|Frap]] 11:31, 5 February 2007 (UTC)

:I think 100kh MTBF is not in any way a meaningful number. This is about 20 years. And this is supposed to be the MEAN time between failure, with all the DOA units factored in? Have you ever seen the insides of a PSU after 5 years? The .. dust? Did I mention the dust? And sometimes the layers of smoke residue? I think whether they tell you 100k hours or 200k hours, or 258.2 k hours -- who is to check that number, when 80% of PSUs get junked after 3 years anyway when the industry figures out they need just another "standard" and new, new, all new. <small>—The preceding [[Wikipedia:Signatures|unsigned]] comment was added by [[User:145.253.2.234|145.253.2.234]] ([[User talk:145.253.2.234|talk]] • [[Special:Contributions/145.253.2.234|contribs]]) 13:29, June 14, 2007 (UTC{{{3|}}})</small>

:The Lifespan section needs a re-write as it fundamentally misunderstands MTBF. A good example of the difference between MTBF and lifetime in the context of PSUs can be seen here: http://www.xppower.com/pdfs/MTBF.pdf <span style="font-size: smaller;" class="autosigned">— Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[Special:Contributions/118.71.9.127|118.71.9.127]] ([[User talk:118.71.9.127|talk]]) 06:38, 25 June 2014 (UTC)</span><!-- Template:Unsigned IP --> <!--Autosigned by SineBot-->

:: That's a very good point, thank you! Went ahead and {{Diff|Power supply unit (computer)|614862260|614843508|corrected and expanded}} the section, please check it out. &mdash;&nbsp;[[User:Dsimic|Dsimic]]&nbsp;([[User talk:Dsimic#nobold|talk]]&nbsp;|&nbsp;[[Special:Contributions/Dsimic|contribs]]) 09:21, 29 June 2014 (UTC)

== Laptop supplies ==
The article currently does not mention laptop power supplies, which are quite different. It is particularly important to cover them, because they are generally much more energy efficient than desktop computers. Also, there should be mention of specialized low-power supplies, such as used in carputers.-[[User:69.87.202.60|69.87.202.60]] 12:25, 17 May 2007 (UTC)
:I just added a section for laptop power supplies. Feel free to expand or clarify. [[User:Thief12|Thief12]] 03:15, 12 July 2007 (UTC)

:: I agree that laptop power supplies are notable enough to mention somewhere in Wikipedia.
:: I am restoring the "Laptops" section in this "Power supply unit (computer)" article for now.
:: Or would some other Wikipedia article be a more appropriate place to discuss laptop computer power supply units? --[[Special:Contributions/68.0.124.33|68.0.124.33]] ([[User talk:68.0.124.33|talk]]) 04:59, 12 May 2009 (UTC)

== neutrality of the Dell reference ==

at the end of the first paragraph, there are references to dell producing non-standard ATX pinouts, and sugguesting it was an attempt to punish consumers using non-dell parts. first, the reference is speculation, as there is no evidence that it wasn't a cost-saving design decision that made a stupid decision of reusing connectors. secondly, compaq did this with many componets right up until the HP buyout, but it's not mentioned here. dell is neither the do-no-wrong company, nor the evil empire. --[[User:66.69.175.41|66.69.175.41]] 13:20, 12 June 2007 (UTC)
::Agreed, I think we should remove it. The practice isn't limited to Dell[[User:Achilles2.0|Achilles2.0]] 05:00, 9 July 2007 (UTC)
::I've put back a non-accusatory mention of this fact. It is important and should have prominence because knowing it will prevent people from damaging their computers. It's an easy mistake to make. [[User:Martinwguy|Martinwguy]] ([[User talk:Martinwguy|talk]]) 15:56, 7 September 2009 (UTC)

== power measurement ==

Is there any way of measuring the power being supplied by a system's PSU at a given time? [[User:Drutt|Drutt]] 18:42, 27 June 2007 (UTC)

:I'm not sure about that, but you can use a device to measure the power being ''used'' by a PSU. The Kill-a-Watt electricity usage monitor is such a device. Some PSUs have also been tested for their energy efficiency at different loads. If you have that information you could calculate how much power your PSU was supplying based on the number of watts being used and the % efficiency at the % capacity of the load.
[[User:Mspandana|Mspandana]] 06:00, 4 August 2007 (UTC)

::The output is just DC. Use an ammeter on the various output lines (12v, 5v, etc.) There are also special devices for testing power supplies that allow one to provide reference loads and more conveniently measure the outputs also. See for instance the power supply reviews at silentpcreview.com [[User:Zodon|Zodon]] ([[User talk:Zodon|talk]]) 09:42, 18 February 2008 (UTC)
::Good power supplies have an typical efficiency of 80%, but cheap ones may have very low efficiencies like 60%. --[[User:MrBurns|MrBurns]] ([[User talk:MrBurns|talk]]) 09:31, 17 January 2009 (UTC)


== Label ==
== Label ==
Line 127: Line 53:
|}
|}


== External links modified ==
== Removal of paragraph about modular cabling creating high resistance. ==

I've removed the statements about modular cabling creating high resistance because these claims are biased. This claim is made by a company that themselves does not produce modular power supplies. I just found out, however, that there can be a sginificant amount of resistance across modular connectors, but it depends on the age of the PSU, the manufacturer, and the number of times the plug was insterted/removed. The article is found here: http://www.motherboards.org/articles/guides/1488_11.html I wasn't sure whether this was reliable, however, so I left my edit alone. I know anyone can make a website, but it looks like a long-running and well-established one, and I wanted to be sure. Thanks. [[User:Totakeke423|Totakeke423]] ([[User talk:Totakeke423|talk]]) 05:35, 30 March 2008 (UTC)
::I'm kind of wondering how an unused connector port can provide ANY resistance.----[[User:Asher196|Asher196]] ([[User talk:Asher196|talk]]) 05:01, 30 March 2008 (UTC)
::::I think it's only for the ones that are connected. I think the resistance comes from loose connections.[[User:Totakeke423|Totakeke423]] ([[User talk:Totakeke423|talk]]) 05:36, 30 March 2008 (UTC)
::::::I rewrote the section I removed about modular cabling and added a source website. [[User:Totakeke423|Totakeke423]] ([[User talk:Totakeke423|talk]]) 07:37, 30 March 2008 (UTC)


==Few points==
"Quality construction consisting of industrial grade electrical components and/or a higher speed fan can help to contribute to a higher MTBF rating by keeping critical components cool, thus preventing the unit from overheating. Overheating is a major cause of PSU failure."

OTOH higher speed fans dont last as long, and fan failure is also a cause of many PSU failures. The best PSUs use larger than usual lower speed fans.


"While it is hoped that eliminating the excess cables will improve the flow of cooling air inside the computer case,"

only to a trivial extent, not really a significant issue


"While it's relatively easy and inexpensive to lubricate/replace a fan, opening a power supply can be dangerous and usually voids the warranty, so it's best left to a professional."

a dubious conclusion, and clearly wrong for a lot of people.
[[User:Tabby|Tabby]] ([[User talk:Tabby|talk]]) 12:59, 10 May 2008 (UTC)

it took me 8 years to figure out how to build a computer now that i think about it. The ps i got had 2 fans blowing air out. Now that i look at it, one fan blows air in the other blows it out! <span style="font-size: smaller;" class="autosigned">—Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[Special:Contributions/69.255.42.105|69.255.42.105]] ([[User talk:69.255.42.105|talk]]) 09:45, 18 May 2011 (UTC)</span><!-- Template:UnsignedIP --> <!--Autosigned by SineBot-->

==Incorrect Information==

"Using a power supply that is larger than necessary can significantly increase operating costs by wasting energy."

Is completley false, a power supply does not draw more electricity than it needs to. A 600W psu will not use anymore power than a 300W psu when powering the same hardware.

"Most power supplies have an efficienty at about 50% of the rated power load."

Makes no sense. What does it mean? Their efficiency is highest at 50% load? They become more efficient after 50% load? They become less efficient after 50% load?

[[User:Comrinec|Comrinec]] ([[User talk:Comrinec|talk]]) 01:04, 31 May 2008 (UTC)
:Yes, it varies but it generally peaks at a certain percentage and below and above that, it may be have lower efficiency. -- [[User:Frap|Frap]] ([[User talk:Frap|talk]]) 02:19, 31 May 2008 (UTC)

The larger point is important to include, which most people would not know about or consider, which is that the average power supply actually produces dramatically less power than what the sticker on the outside says (i.e. what it is "rated" as). From what I have read, 50% of rated output is not an unrealistic figure, at what point in the power usage graph it might appear is secondary the the interests of the average person. In this same regard, I think it would be useful for the article to detail the importance of "Active PFC" and it's association with an "80+" rating as an objective way of differentiating between higher-quality power supplies, and the substandard junk that is so common, and causes so many problems with people's computers.
[[User:Jonny Quick|Jonny Quick]] ([[User talk:Jonny Quick|talk]]) 21:57, 30 July 2011 (UTC)Jonny Quick

== Rename article to Power Supply Unit (Computer) ==

I think that this article has got the wrong name. The technical name for computer power supplys is power supply unit (PSU). I propose we rename the article as above. Do people agree? <small>—Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[User:WikiJonathanpeter|WikiJonathanpeter]] ([[User talk:WikiJonathanpeter|talk]] • [[Special:Contributions/WikiJonathanpeter|contribs]]) 19:25, 31 May 2008 (UTC)</small><!-- Template:Unsigned --> <!--Autosigned by SineBot-->
::I agree. I don't like the current name of the article.[[User:Asher196|Asher196]] ([[User talk:Asher196|talk]]) 23:15, 31 May 2008 (UTC)
::: Thank you. Since noone has left a negative comment regarding this, I will go ahead with the change. Thanks. [[User:WikiJonathanpeter|WikiJonathanpeter]] ([[User talk:WikiJonathanpeter|talk]]) 22:05, 12 June 2008 (UTC)

== Contradiction ==

The section ''Power rating'' says <blockquote>Some believe that using a PSU with more power potential than you need, wastes energy. This is a myth.</blockquote>
The section ''Energy efficiency'' says <blockquote> It's important to match the capacity of a power supply to the power needs of the computer. The energy efficiency of power supplies drops significantly at low loads. Efficiency generally peaks at about 50-75% load.</blockquote>
This is a serious contradiction. -- [[User:JorgePeixoto|Jorge Peixoto]] ([[User talk:JorgePeixoto|talk]]) 00:36, 19 August 2008 (UTC)

== what? ==

what? [[Special:Contributions/66.203.20.96|66.203.20.96]] ([[User talk:66.203.20.96|talk]]) 13:57, 4 September 2008 (UTC)

== Troubleshooting ==

Am I right that the fan will operate on ''all'' PC PSU's when it is first connected to the power supply... and that if the fan does not turn then the PSU itself has failed? vs the fan turns, but the PC can't turn on - the fault is probably with the motherboard.

Don't know if this is just my observation of a limited range of equipment or pretty much the way these things all work.

[[User:GarrieIrons|Ga]][[User_talk:GarrieIrons|rr]][[Special:Contributions/GarrieIrons|ie]] 00:12, 19 October 2008 (UTC)

==I want to delete these sections==

*troubleshooting
*capacitors
*noisy fan

It is not wikipedia's purpose to provide trobleshooting, and it is a dangerous thing as well. capacitor plague does not deserve it's own heading, and is not needed anyway. This isn't any more of an issue with power supplys than it is with TV's and mb's. Noisy fan - along the same lines. and to say it's not that hard to lubricate - are you kidding, you can kill yourself lubricating a psu fan.

Also, all and every single point listed under troubleshooting psu can be caused by other components as well.

P.S. feel free to delete them for me
[[User:Silverxxx|Silverxxx]] ([[User talk:Silverxxx|talk]]) 15:25, 24 March 2009 (UTC)
:I agree 100%- It also is a [[WP:HOWTO|how-to]]. I'll just delete it. [[User:Gboycolor|Gboycolor]] ([[User talk:Gboycolor|talk]]) 02:04, 4 November 2009 (UTC)
::Sigh... I wish there was a "NOTWIKIPEDIA" that simply listed all the useful (but not encyclopedic) information that someone took the time to add to wikipedia only to have it deleted by some later editor. I mean, seriously, how hard would it be for Wikipedia to have a checkbox labeled "Move to Non-encyclopedic" next to the save button that would move any deleted information to some other Wiki? I think such a thing might even encourage people to be more scrupulous in cleaning up HOWTOs and other such obviously useful information since they'd know that the information was being preserved, just not in Wikipedia. <span style="font-size: smaller;" class="autosigned">— Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[Special:Contributions/24.18.238.89|24.18.238.89]] ([[User talk:24.18.238.89|talk]]) 18:05, 4 November 2013 (UTC)</span><!-- Template:Unsigned IP --> <!--Autosigned by SineBot-->

==Power Rating section needs cleanup==
Lots of the information near the beginning of this section is unsourced and seems potentially incorrect. Specifically, the claim regarding 300-500W power supplies stating that "the use of which is limited to Internet-surfing and burning and playing DVDs." is highly dubious - a personal computer with a 500W power supply is capable of doing a much wider range of activities than this implies! And as is mentioned later in this section (with citation), the very large wattages listed near the beginning for "enthusiasts" are not actually required in most circumstances.

I think the "surfing and burning DVDs" phrase can simply be deleted now, and the claims of ranges and usages should probably be given a "citation needed" tag and looked into further.
[[User:Jnthornh|Jeremy Thornhill]] ([[User talk:Jnthornh|talk]]) 14:03, 10 April 2009 (UTC)

== Power Supply and Automatic Voltage Regulators? ==

Ever since the start of computing, there is always the conception on requiring every PC unit to have its own AVR (automatic voltage regulator). Here in our country, I rarely see anyone who connects the computer directly to the power outlet. Can you please post an answer related to the computer power supply unit's dependency to the automatic voltage regulator or even a surge protector cause it is really needed. [[User:Triadwarfare|Triadwarfare]] ([[User talk:Triadwarfare|talk]]) 14:15, 11 October 2009 (UTC)
:Maybe you live in the bush or something with crap electricity. A surge protector is a nice thing to have on expensive equipment, but is not necessary. Here in our country, I rarely see anyone with an surge protector. -- 16:47, 11 October 2009 (UTC)

::What about surges from lightning? Unless all your power cables are underground? There are other sources of spikes, surges, brown-outs etc that probably make some form of protection/ power conditioning advisable, especially in 'critical' applications. U.P.S.s, Uninterruptible Power Supplies provide a lot of protection, apparently, and are becoming cheaper. ALDI, for example, no commercial endorsement, sells them.

::Don't forget Triadwarfare, this is Wikipedia, not a blog or bulletin board. Or did you mean you wanted this content in the article? --[[Special:Contributions/220.101.28.25|220.101.28.25]] ([[User talk:220.101.28.25|talk]]) 03:53, 25 October 2009 (UTC)

::Well, I want it there in the article since there is a misconception here. AVR/Surge protector here is very popular that every computer unit must come with an AVR/Surge protector. I just want to include and verify if Computer Power supplies have their own surge protector (which negates the need for an AVR/surge protector). Anyway in our country, electricity is not underground... they are hung by electric posts. [[User:Triadwarfare|Triadwarfare]] ([[User talk:Triadwarfare|talk]]) 06:15, 15 November 2009 (UTC)

::The design and control circuitry of the modern PSU does provide some (limited) surge protection, but if you have regular lightning strikes on overhead mains distribution, then additional surge protection would be a very good idea. I've had phone systems, video recorders and TVs damaged by lightning strikes, but never a computer - though perhaps I've just been lucky in not having one turned on at the time of the surge. [[User:Dbfirs|''<font face="verdana"><font color="blue">D</font><font color="#00ccff">b</font><font color="#44ffcc">f</font><font color="66ff66">i</font><font color="44ee44">r</font><font color="44aa44">s</font></font>'']] 07:50, 14 May 2010 (UTC)

== Update Needed of Power Supply Pictures ==

The photos of the open power supply, and the one of the output cables are ''seriously'' out of date. The latter picture is 5-1/2 years old that is ANCIENT.

It has NO SATA or PCI-E connectors at all. It has ATX+12, and thats about ''it''. ''Very'' out of date. The photo of the SATA connector on the talk page should be placed in the article.
Further graphics should be available in the commons or from Wiki books(?). --[[Special:Contributions/220.101.28.25|220.101.28.25]] ([[User talk:220.101.28.25|talk]]) 04:10, 25 October 2009 (UTC)

== Claim regarding resistance of connectors vs wire is unsupported ==

Article currently says: "However, this [resistance of connectors varying by quality, age, insertions] is somewhat inconsequential as the amount of this resistance in a good connector is small compared to the resistance generated by the length of the wire itself.[9]"

Reference 9 is to an article where someone tested some power supplies, and did not specifically examine the resistance of connectors and their crimp joints, versus the resistance of the wires. The author of that article makes it clear in the forum comments:
"Agreed [that measured voltage drops could be attributed to various causes]. Could be the wire. Could be the crimp. That's why I said I couldn't explain it. Since the wires were brand new with no breaks in them and the gage of wire the same on all five PSU's, then I think problem would be how the wire is attached to the connector pins."

In short, this article is not a suitable reference for the point made here, which I believe to be suspect. The resistance due to wire length for various gauges of stranded wire can be looked up in many references. It is approx 6 ohms/1000 feet for 18 guage, which amounts to .012 ohm for two feet, resulting in 0.12 V drop per 10A (and bear in mind that in a PC PSU, the high currents are distributed across multiple wires, 4 or 5 for 3.3 and 5V, and 8 for ground in the 24-pin ATX12V, according to the current wikipedia article.).

I have not made an exhaustive study of connectors, but I have certainly measured higher voltage drops attributable to deteriorated connectors -- it is not uncommon to see the nylon connector bodies turned brown due to heat, which is an indicator that additional resistance has developed there. [[User:Gwideman|Gwideman]] ([[User talk:Gwideman|talk]]) 08:33, 14 October 2010 (UTC)

:I see that you're unfamiliar with editing. You seem to know something about the subject, and it'd be hard to make this article any worse. --[[User:Juventas|Juventas]] ([[User talk:Juventas|talk]]) 02:41, 15 October 2010 (UTC)

== Supposed power consumption, absurd? ==

"[...] in recent years the power demand of "video cards" in the ability to watch high definition (HD) media has led to even the average ATX computer to consume between 400 and 500 watts."

Completely false. Show me an "average" ATX computer that actually ''consumes'' 400 - 500 watts. That power consumption more accurately describes a high-end multi-GPU configuration under load. Hardly something needed for viewing HD media. [[User:AerisArchon|AerisArchon]] ([[User talk:AerisArchon|talk]]) 14:03, 19 July 2011 (UTC)

== It's a very good article ==

It would be wrong to loose sight of the great merits of this article which are lucidity and relevance to the likely concerns of whoever has searched it up - in my case the correct PSU for a graphics card.

[[User:Sporus|Sporus]] ([[User talk:Sporus|talk]]) 04:56, 20 September 2011 (UTC)

==Suggest merge from [[Power supply rail]]==
[[Power supply rail]] is a dictionary definition, followed by a lot of stuff that belongs only in an article about PC-compatible power supplies. Whatever is unique from that article should be moved here. --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 04:36, 27 October 2011 (UTC)

*'''support'''. Power supply rail is a general idea about power delivery (does it come "third rail live"?); the PSR article is about PC supplies. The merge is appropriate. I'd let PSR have a general description with a see also pointing here. [[User:Glrx|Glrx]] ([[User talk:Glrx|talk]]) 15:19, 29 October 2011 (UTC)

==Merge (really rename) to [[Power supply unit]]==
Someone has suggested renaming this article to "power supply unit". I oppose this renaming because this article is about a very specific application of power supplies to desktop personal computers - in principle , "power supply unit" should be more general, not a list of easy-to-research IBM PC trivia. --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 14:33, 7 December 2011 (UTC)
:There may of been some confusion here. If PSUs were primarily for computers, it would make sense to rename. Since they're not, and there is an article for [[power supply]], I've changed the redirect for [[power supply unit]] to that (instead of here). --[[User:Juventas|Juventas]] ([[User talk:Juventas|talk]]) 02:03, 8 December 2011 (UTC)
*'''Oppose'''. This article discusses PSU used in PCs, so (computer) disambig suffix is apt. There can be a more general PSU. There may be a better article name (perhaps something reflecting AT and ATX designation), but the article should stick around with a restricted title. [[User:Glrx|Glrx]] ([[User talk:Glrx|talk]]) 17:58, 9 December 2011 (UTC)
*'''Oppose''' - This is a better title (as the article is currently titled), because it's more specific. <small><font face="arial"><strong>[[User:Northamerica1000|Northamerica1000]]</strong><sup>[[User_talk:Northamerica1000|(talk)]]</sup></font></small> 15:15, 10 January 2012 (UTC)

== Suggest merge with [[modular power supply unit]] ==

Modular PSUs are an enthusiast-only product, and exist in a tiny fraction of 1% of PCs. If anything I think their current representation in '''power supply unit (computer)''' should be reduced. The most relevant place to merge them would be [[enthusiast computing]]. --[[User:Juventas|Juventas]] ([[User talk:Juventas|talk]]) 03:07, 31 December 2011 (UTC)

*'''Note''' - It has previously been suggested that [[Modular power supply unit]] be merged into this article, which, if I'm not mistaken, the user above opposes, instead stating that information about modular power supply units in this article should be merged into the [[Modular power supply unit]] article. (See the next section below for more information.) <small><font face="arial"><strong>[[User:Northamerica1000|Northamerica1000]]</strong><sup>[[User_talk:Northamerica1000|(talk)]]</sup></font></small> 15:04, 10 January 2012 (UTC)

*'''Support merging to [[Modular power supply unit]], but NOT to [[enthusiast computing]]''' - See my comments below (in the next section). Keeping some basic introductory information about modular power supply units in this article, with an included <nowiki>{{main|Modular power supply unit}}</nowiki> note at the section's lead seems to be a good fit. This type of reasonable merge won't affect no-one seeking information about this specific topic. <small><font face="arial"><strong>[[User:Northamerica1000|Northamerica1000]]</strong><sup>[[User_talk:Northamerica1000|(talk)]]</sup></font></small> 15:10, 10 January 2012 (UTC)

== Suggest Merge from [[Modular power supply unit]] to this article ==
A merge to tag exists on the [[Modular power supply unit]] article, to merge information from that article to this one.
*'''Oppose''' - Rationale:
:*This article is already lengthy, and the addition of more information will make it even lengthier, and difficult to navigate
:*The Modular power supply unit article is a reasonable content fork for the concept of computer power supply units
:*The Modular power supply unit article would benefit from improvements, such as the addition of more sources (I've already added a few to that article)
:*Note that User:Juventas in the section above this one has proposed the opposite, merging information about modular power supply units from this article to the Modular power supply unit article.
::{{mdash}} <small><font face="arial"><strong>[[User:Northamerica1000|Northamerica1000]]</strong><sup>[[User_talk:Northamerica1000|(talk)]]</sup></font></small> 15:08, 10 January 2012 (UTC)
*'''Support''' - Rationale:
:*Having two articles is quite in keeping with the Wikipedia tradition of obsessively cataloging recent comptuer trivai while ignoring the overview. More part numbers, please. Maybe we can add a list of Alibaba power supply vendors. That would be verifiable reference information, and is not Wikipedia the ultimate collection of randomly-associated facts? --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 15:07, 11 January 2012 (UTC)
:*A modular power supply merely has plugs and sockets to individually connect the device cables to the power supply, rather than having a big snarl of unused wire. Other than that it is an identical device to this one. Otherwise, I suppose we could two sets of articles for every device where the power cord is permanently attached, vs separable... [[User:DMahalko|DMahalko]] ([[User talk:DMahalko|talk]]) 08:46, 12 January 2012 (UTC)
:* MPS article doesn't say much (but its title could apply to wallwarts, open frames, and any colored-box unit). Even with a patent, I'm not sure MPS would pass muster as [[WP:N]]. [[User:Glrx|Glrx]] ([[User talk:Glrx|talk]]) 18:44, 13 January 2012 (UTC)

== 48v power for high-end video cards and systems? ==

Does anyone know if 48v power will eventually be used inside computers?

It's stupid for high wattage machines to need multiple 12v power rails when the internal devices that are requiring it don't even use that voltage for its intended purpose.

+12v was originally meant for driving motor loads, and video cards and so forth don't use it for that but instead use it to drive onboard voltage converters for more +5v / +3.3v etc on the card. Manufacturers don't expect this voltage from the power supply though because the amperage and wire gauge would be massive.

In the same manner, modern computer motherboards don't directly use the power supply voltage as provided but rely on separate voltage regulators on the motherboard to step down to about 1 volt for the processor.

It would make more sense to distribute 48v internally so that wiring can be smaller and more efficient. Cards that need more power can convert down from 48v to whatever they need just as is done now with 12v, but at 1/4th the amp rating.

The whole field of "standard voltages" is probably ripe for revamp. How many systems actually directly use the +5v from the PSU, vs lower voltages provided by onboard regulators?

[[User:DMahalko|DMahalko]] ([[User talk:DMahalko|talk]]) 08:59, 12 January 2012 (UTC)
:[[WP:BOLD|If you can cite it, write it.]] I haven't seen anyone writing on this topic, and Chinese manufacturers seem to have the art of the $25 500-watt power supply down to a fine art (even with all the UL and CSA stickers that they have to print on them). 48 volts is nice because you can make that voltage from a set of 24 lead-acid wet cells. --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 14:45, 12 January 2012 (UTC)

==Suggest merge==
[[Entry-Level Power Supply Specification]]is a variant type of PC-compatible power supply; this article could be merged here to give it context. --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 19:28, 28 January 2012 (UTC)
*'''Support''' - [[Entry-Level Power Supply Specification]] is a short article, so merging it to this one wouldn't make this article prohibitively longer. The merge would also provide additional context for this article. <small><font face="arial"><strong>[[User:Northamerica1000|Northamerica1000]]</strong><sup>[[User_talk:Northamerica1000|(talk)]]</sup></font></small> 02:21, 30 January 2012 (UTC)
*'''Support'''. That spec is dated 2004, and it seems that subsequent ATX supplies followed suit. [[User:Glrx|Glrx]] ([[User talk:Glrx|talk]]) 17:51, 30 January 2012 (UTC)

== Very narrow scope, take those PC blinders off ==

Seems like a typical article written by non-experts for non-experts. But that's why this entire article isn't just a minor subsection of [[switched-mode power supply]].

Hey, there were computer power supplies as discrete units before the IBM PC? And gasp, linear supplies at one point, too? Wow! Why not start with talking about those?

Explaining how it was all mashed together to begin with in early computing devices, but then separated out into a removable unit, might also be a good idea.

Oh, don't wanna get too technical though, or this article might start duplicating [[switched-mode power supply]] too much. [[User:DMahalko|DMahalko]] ([[User talk:DMahalko|talk]]) 16:13, 2 February 2012 (UTC)
:Experts get paid to write articles -you pays your money and you gets what you gets. There's lots of easy-to-verify train-spotting trivia about PC-compatible power supplies that we love to catalogue here; not one hint of *why* all these variations exist, nor of how important any of them were, but lots of replication of data sheets. I look forwward to a new direction for the article, but as the [[2N3904]] debacle taught me, this is a parts catalog, not an encyclopedia, when it comes to items electrical. --[[User:Wtshymanski|Wtshymanski]] ([[User talk:Wtshymanski|talk]]) 16:23, 2 February 2012 (UTC)

== Switching the PSU On and Off ==

A footnote to the diagram for the 24-pin ATX12V 2.x connector indicates that the PSU can turned on by grounding pin 16.
When 16 is isolated from ground the PSU switches off; correct? So the system board holds 16 to ground for the PSU to continue supplying power. The last step of system shut-down is to isolate pin 16 from ground. The power switch on the back of the PSU switches the AC input. The power switch on the front of the computer is a normally open momentary contact. Closing it activates the circuitry on the system board which grounds pin 16. Regards, ... [[User:PeterEasthope|PeterEasthope]] ([[User talk:PeterEasthope|talk]]) 00:36, 24 November 2012 (UTC)

== Pictures of connectors, please ==

The end of the article has a nice chart showing the differences between the AT and ATX connector wiring, but doesn't have an actual picture of what the connectors look like or how they plug into a motherboard. <span style="font-size: smaller;" class="autosigned">— Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[Special:Contributions/24.18.238.89|24.18.238.89]] ([[User talk:24.18.238.89|talk]]) 17:53, 4 November 2013 (UTC)</span><!-- Template:Unsigned IP --> <!--Autosigned by SineBot-->

: There you go, just added a picture of a 24-pin ATX power conector. :) The attachment to the motherboard connector can be seen at the image below it, depicting a PSU tester connected to an ATX PSU. Unfortunately, I have no AT PSUs to take pictures of. -- [[User:Dsimic|Dsimic]] ([[User talk:Dsimic#nobold|talk]]) 00:05, 5 November 2013 (UTC)

== Connector Standarisation ==

The ATX motherboard connector, PCIe 6 and 8 pin connectors and CPU connectors all seem to be very similar, and most modular PSUs use essentially the same connectors on the PSU end. What I'm wondering is are these a standard type of connector that's available for other uses, or is specific to computer power supplies? I couldn't find it mentioned anywhere. -- [[User:Haravikk|Haravikk]] ([[User talk:Haravikk|talk]]) 14:42, 21 November 2013 (UTC)


Hello fellow Wikipedians,
: Those 6- and 8-pin connectors are standard types, and they can be used for any purpose &ndash; just like the "Molex" power connector for PATA HDDs and CD-ROMs. It's just they're very common on PSUs. ATX power connector is somehow specific to the PSUs, haven't seen it used for other purposes.
: Just as a note, talk pages are not supposed to be used for general-purpose questions, only improvements to the articles should be discussed there. -- [[User:Dsimic|Dsimic]] ([[User talk:Dsimic#nobold|talk]]) 14:53, 21 November 2013 (UTC)


I have just modified 8 external links on [[Power supply unit (computer)]]. Please take a moment to review [[special:diff/820065509|my edit]]. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit [[User:Cyberpower678/FaQs#InternetArchiveBot|this simple FaQ]] for additional information. I made the following changes:
== Bizarrely narrow ==
*Added archive https://web.archive.org/web/20080511164749/http://www.hardwaresecrets.com/article/534 to http://www.hardwaresecrets.com/article/534
*Added archive https://web.archive.org/web/20070927210041/http://www.hardwaresecrets.com/article/397/5 to http://www.hardwaresecrets.com/article/397/5
*Corrected formatting/usage for http://cache-www.intel.com/cd/00/00/52/37/523796_523796.pdf
*Added archive https://web.archive.org/web/20070227053257/http://www.efficientpowersupplies.org/pages/PC_Energy_Star_Recommendation.pdf to http://www.efficientpowersupplies.org/pages/PC_Energy_Star_Recommendation.pdf
*Added archive https://web.archive.org/web/20070514033001/http://www.hardwaresecrets.com/article/370 to http://www.hardwaresecrets.com/article/370
*Added archive https://web.archive.org/web/20080511164749/http://www.hardwaresecrets.com/article/534 to http://www.hardwaresecrets.com/article/534
*Added archive https://web.archive.org/web/20080512153640/http://www.hardwaresecrets.com/article/181 to http://www.hardwaresecrets.com/article/181
*Added archive https://web.archive.org/web/20110713121329/http://www.journeysystems.com/support/calculator/ to http://www.journeysystems.com/support/calculator/


When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.
This section has a title that led me to expect a discussion of power supplies for computers, in general. Instead, what I found is a discussion of power supplies for PC compatibles, a very narrow category. The section that purports to be a history is written as if the only power supply that ever mattered was the supply on the first PC. What about the huge motor generator sets that provided power to many 1950s era computers? What about the constant-voltage transformers and "computer grade" capacitors the size of hand grenades that were lined up in ranks in the power supplies of 1960s era computers? What about the advent of switching power supplies in 1970s-era minicomputers? Either there's an awful lot missing here or the page title should be changed to '''PC power supply'''. [[User:Douglas W. Jones|Douglas W. Jones]] ([[User talk:Douglas W. Jones|talk]]) 17:36, 21 November 2013 (UTC)


{{sourcecheck|checked=false|needhelp=}}
: You're more than welcome to expand this article so the complete history of PSUs is covered! -- [[User:Dsimic|Dsimic]] ([[User talk:Dsimic#nobold|talk]]) 21:08, 21 November 2013 (UTC)


Cheers.—[[User:InternetArchiveBot|'''<span style="color:darkgrey;font-family:monospace">InternetArchiveBot</span>''']] <span style="color:green;font-family:Rockwell">([[User talk:InternetArchiveBot|Report bug]])</span> 21:55, 12 January 2018 (UTC)
: If you've ever looked at mainframe-era PSU's you'll see that the principles have really not changed all that much. In fact it was probably the computer's need for low voltages at substantial currents that drove the widespread adoption of the Buck converter in place of inefficient linear regulators, in most electronic gear. The main difference was that earlier designs tended to be secondary switching because the power transistors then available could not withstand primary voltages. I've never seen capacitors the size of hand grenades, mostly they were substantially larger than that, although of not much greater capacity than smaller modern units. Such is progress. --[[User:Anteaus|Anteaus]] ([[User talk:Anteaus|talk]]) 16:28, 18 January 2016 (UTC)


== Does 3.3v serve any purpose? ==


== A Tear Down of PSU Transformers and Chokes ==
Point of interest, I believe the 3.3v rail was introduced with the first 'Slot A' ATX processors as the core voltage. Processors then started using bespoke core voltages, thus the 3.3v rail lost its original purpose. As the article says, it is easier to regulate the 5v or 12v rail to provide a selectable core voltage, which is what most if not all mobo manufacturers do. It's also on SATA power plugs, but again it's questionable as to whether it's used by any modern disk. It would be interesting to know if the 3.3v rail is doing anything at all in the modern PC, or if it's just a piece of historical baggage. Anyone got the lowdown on that? --[[User:Anteaus|Anteaus]] ([[User talk:Anteaus|talk]]) 17:02, 18 January 2016 (UTC)


<gallery>
: Hello! For example, PCI Express slots provide 3.3&nbsp;V power supply lines, see [http://cds.linear.com/docs/en/design-note/dn346f.pdf this document] for more details. Pretty much the same applies to conventional PCI slots. &mdash;&nbsp;[[User:Dsimic|Dsimic]]&nbsp;([[User talk:Dsimic#nobold|talk]]&nbsp;&#124;&nbsp;[[Special:Contributions/Dsimic|contribs]]) 22:14, 18 January 2016 (UTC)
Speicherdrossel PC-Netzteil IMG 20180621 1358.jpg
Main Transformer SPI 8TG00048 1FC 0310 of a PC PSU IMG11706.jpg
Main Transformer PW-0003-Y of a PC PSU IMG11710.jpg
</gallery>
Basics of the main transformer and chokes are missing in the article.
--<span style="color:#00A000;">Hans Haase ([[User talk:Hans Haase|有问题吗]])</span> 21:36, 24 November 2018 (UTC)
:Here[https://www.youtube.com/watch?v=tWG0SNYzFRI ] is a sample, somebody did it before, but smaller supplies tend to have less windings of wire. --<span style="color:#00A000;">Hans Haase ([[User talk:Hans Haase|有问题吗]])</span> 20:06, 24 December 2018 (UTC)


== Protections ==
== Suggest adding paragraph for Top Mount vs Bottom Mount design ==


Hello. This article completely lacks info about Power Supply Units protections, like HIPOT, OVP (Over Voltage Protection), OPP (Over Power Protection), SCP (Short Circuit Protection), etc . These are important info to look when buying PSUs, as the lack of those protections means the PSU doesn't have much quality and might damage the computer. They are, however, briefly mentioned at [[Power supply#Specification]].--[[User:MisterSanderson|MisterSanderson]] ([[User talk:MisterSanderson|talk]]) 17:01, 30 November 2018 (UTC)
I looked up this article hoping to find more information on when PSUs changed from being Top Mounted in the computer case to Bottom Mounted, along with some info on the efficiency of the Bottom Mount design. I think this would be a good addition to this article and help explain why PSUs are mounted in cases the way they are. [[User:Dlsimon|Dlsimon]] ([[User talk:Dlsimon|talk]]) 14:48, 5 May 2016 (UTC)
:The truth is, in large, not special produced series, there are no dedicated components inside what care for this specifications. HIPOT is a 5kV insulation test of the transformers, before brought to the supply production. If the PSU is specified, the whole PSU is testet, before leaving the factory. OVP requires to monitor each output rail on voltage, which is generating a error signal to stop the smps-controller. SCP is not measured, the difference to a charger is to allow for a limited time not reaching the output voltage on full power, to make hard drives spin-up of peak power consuming operations of the CPU not to shutdown the psu immediately. Smps-controllers like the TL494 or KA7500 are made for one output voltage, only. But can be used for PSU by designing a circuit around it. The KA3511 supports 3.3, 5, and 12V rails. But there are still other variants on the market. --<span style="color:#00A000;">Hans Haase ([[User talk:Hans Haase|有问题吗]])</span> 07:59, 3 December 2018 (UTC)

Latest revision as of 10:13, 8 October 2024

Label

[edit]
Example ATX power supply label
AC Input 100–240 Vac — 10–5A 50–60Hz
DC Output +3.3V +5V +12V1 +12V2 +12V3 +12V4 −12V +5Vsb
Max Output Current 36A
(118.8 W) 		30A
(150 W) 		18A
(216 W) 		18A
(216 W) 		18A
(216 W) 		18A
(216 W) 		0.5A
(6 W) 		3.0A
(15 W)
Max Combined Power 155W 680W 20W
Total Power 700W
[edit]

Hello fellow Wikipedians,

I have just modified 8 external links on Power supply unit (computer). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 21:55, 12 January 2018 (UTC)[reply]


A Tear Down of PSU Transformers and Chokes

[edit]

Basics of the main transformer and chokes are missing in the article. --Hans Haase (有问题吗) 21:36, 24 November 2018 (UTC)[reply]

Here[1] is a sample, somebody did it before, but smaller supplies tend to have less windings of wire. --Hans Haase (有问题吗) 20:06, 24 December 2018 (UTC)[reply]

Protections

[edit]

Hello. This article completely lacks info about Power Supply Units protections, like HIPOT, OVP (Over Voltage Protection), OPP (Over Power Protection), SCP (Short Circuit Protection), etc . These are important info to look when buying PSUs, as the lack of those protections means the PSU doesn't have much quality and might damage the computer. They are, however, briefly mentioned at Power supply#Specification.--MisterSanderson (talk) 17:01, 30 November 2018 (UTC)[reply]

The truth is, in large, not special produced series, there are no dedicated components inside what care for this specifications. HIPOT is a 5kV insulation test of the transformers, before brought to the supply production. If the PSU is specified, the whole PSU is testet, before leaving the factory. OVP requires to monitor each output rail on voltage, which is generating a error signal to stop the smps-controller. SCP is not measured, the difference to a charger is to allow for a limited time not reaching the output voltage on full power, to make hard drives spin-up of peak power consuming operations of the CPU not to shutdown the psu immediately. Smps-controllers like the TL494 or KA7500 are made for one output voltage, only. But can be used for PSU by designing a circuit around it. The KA3511 supports 3.3, 5, and 12V rails. But there are still other variants on the market. --Hans Haase (有问题吗) 07:59, 3 December 2018 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/enwiki/w/index.php?title=Talk:Power_supply_unit_(computer)&oldid=1250075140"