PSU Fundamentals & Recommendations

Want to talk about one of the articles in SPCR? Here's the forum for you.
dhanson865
Posts: 2198
Joined: Thu Feb 10, 2005 11:20 am
Location: TN, USA

Post by dhanson865 » Fri Sep 01, 2006 11:08 am

Code: Select all

Model       Output (W)  65      90      150     200     250     300     400     500     600

S12-330     Efficiency  75.5%   77.9%   80.3%   82.0%   80.9%   79.9%   **
S12-430                 78.3%   80.5%   81.6%   81.8%   80.5%   79.6%   *
Graph those two and see what the "right size" is then. You end up with a graph telling you to get the S12-330w if you are hitting 200 to 300w and thelling you to get the S12-430w if you are staying below 200w, exactly the opposite of the conclusion suggested in the section in question.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Fri Sep 01, 2006 12:09 pm

dhanson865 wrote:

Code: Select all

Model       Output (W)  65      90      150     200     250     300     400     500     600

S12-330     Efficiency  75.5%   77.9%   80.3%   82.0%   80.9%   79.9%   **
S12-430                 78.3%   80.5%   81.6%   81.8%   80.5%   79.6%   *
Graph those two and see what the "right size" is then. You end up with a graph telling you to get the S12-330w if you are hitting 200 to 300w and thelling you to get the S12-430w if you are staying below 200w, exactly the opposite of the conclusion suggested in the section in question.
Right size for what? Isn't it obvious that the actual difference in WATTs is so miniscule it's not really worthy of concern? Different samples or test equipment might give you opposite results.

If you want to choose a PSU only on the basis of highest efficiency for the exact load your computer is using, you're not going to be buying a retail PSU in the first place. You'll be working for a company that's concerned about a bank of hundreds or thousands of servers. Then you will have access to the testing and measurement gear that lets you determine precise what your computers' power demands are in the first place.

For any individual SPCR consumer, efficiency is only one of several considerations, including: noise, cooling, price, availability, and power headroom -- which ultimately also includes "comfort level".

dhanson865
Posts: 2198
Joined: Thu Feb 10, 2005 11:20 am
Location: TN, USA

Post by dhanson865 » Fri Sep 01, 2006 7:29 pm

MikeC wrote: Right size for what? Isn't it obvious that the actual difference in WATTs is so miniscule it's not really worthy of concern? Different samples or test equipment might give you opposite results.
I think you are taking this way too personally. I like the article, I love the website. I'm just trying to give some constructive feedback.

The point is that the article is an introductory article for newcomers and longtimers alike. Seeing that graph back in the day gave me the impression I was going to have to be way more careful choosing the rightsized power supply.

The only way I knew that the S12 series is so close to a non issue in relation to that concept is to read 3 or 4 specific reviews over several years time and compile the table myself as no single article on SPCR makes the full series comparison. Maybe you think I came to the wrong conclusion?

I think it would be a plus to add an additional chart or a small disclaimer that some high quality power supplies almost totally negate the concept. Just as often as you see someone asking about a 600w power supply that isn't needed, I see someone asking about a 330w or less power supply that might be better served by getting the S12-380w or S12-430w instead.

If you think that a change to the article isn't worth doing that is your choice. If you think I'm wrong please correct me.

If you think you've already made that point somewhere else prior to this discussion then forgive me, I must have overlooked it.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Fri Sep 01, 2006 8:07 pm

dhanson865 --

I didn't think I was taking it personally, I just didn't see your point. Now, I think I do. You're right that very high efficiency PSUs with flat efficiency curves tend to make "PSU sizing" something of a non-issue. This seemed so self-evident to me, but I guess it's not to the readers. My apologies. I will make further clarifications at the next opportunity.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sat Sep 02, 2006 2:06 am

dhanson865@I think you are taking this way too personally.
Tone is something that is very open to misinterpretation in a textual medium like the internet. Smilies and other emoticons are supposed to alleviate this textual deficit, but really they are poor substitutes for the subtle nuances of human verbal communication; the same sentence can be an insult or a joke, depending on the person's expression (ie non-verbal cues). For my part, I don't think Mike was being defensive or overreacting, but refinements to the article have been agreed, so all's well that ends well.

Ardemus
Posts: 4
Joined: Sat Aug 05, 2006 3:00 pm
Location: San Luis Obispo, CA USA

Typos - Chipping in for our image ;)

Post by Ardemus » Sun Sep 03, 2006 3:31 pm

This is a great write up, I've read it in bits and pieces (and I'm not finished) but I noticed a few simple typos. *I* don't care, but it's good for the image to not have them :P

Page 3 - Second to last paragraph.
"What does all this mean? The :arrow: safetyl benefit of dual 12V lines is questioned by the engineers I spoke with."
Page 3 - Last paragraph.
If no specific guidlines are given by the manufacturer regarding which components should be connected to which connectors and/or cables (especially with :arrow: detable output cable models), then it is probably safe to assume that the PSU does not have really independent, separate 12V lines.
Page 4 - Last paragraph - Questionable logic?
The models in the Seasonic S12 series, for example, differ so little in efficiency at the same 65~300W power levels that there's no real power consumption :arrow: benefit in choosing a high power model even if it is going to be used at lower power levels.
Perhaps you meant, "Power consumption cost in choosing a high power..." or "Power consumption benefit in choosing a low power..."

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Sun Sep 03, 2006 6:02 pm

All fixed. Thanks.

HueyCobra
Posts: 217
Joined: Tue Feb 14, 2006 12:16 am
Location: Australia

Post by HueyCobra » Sat Sep 16, 2006 6:25 pm

Just noticed that the Antec Phantom 350 and Antec Phantom 500 links in the Recommended PSU Lists point to Antec Online Casino Reports (www.antec-inc.com) rather than the Antec website correctly given for the Antec NeoHE 430 and Antec SmartPower 2-450.

Edit: Something went awry - now the entire Power Supply Fundamentals & Recommendations article appears in the Reference / Recommended section (with correct links) and the Recommended PSU Lists remain incorrect :o

HueyCobra
Posts: 217
Joined: Tue Feb 14, 2006 12:16 am
Location: Australia

Post by HueyCobra » Sat Sep 23, 2006 4:13 pm

The antec-inc links remain in Recommended PSU Lists.

Also, POSTSCRIPT 2 ADDED in the Antec Neo HE 430 power supply review points to itself rather than Postscript 2.

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Sun Sep 24, 2006 10:53 am

All fixed now.

Jose Hidalgo
Posts: 81
Joined: Mon Dec 11, 2006 3:25 pm
Location: Lyon, France

Post by Jose Hidalgo » Mon Dec 11, 2006 3:38 pm

Hi everybody,

May I just ask if there is a reason why hasn't the mCubed EF28 been reviewed yet ?

This PSU might be IMHO one of the very best fanless power supplies out there, plain and simple :D :
- With 200W of permanent power (280W peak power) it can drive most HTPCs and even more,
- With the AC/DC power brick being external, there is almost no heat generated inside the case (let's remember that almost all fanless PSU generate a lot of heat inside the cases, which often requires fans in the end),
- With only 12x6 cm for the internal board, it will fit inside even the tiniest of cases.

What do you think ? I would certainly like to see this PSU reviewed very soon and ranked against all other quality fanless PSUs. :wink:

Thanks in advance,

Jose

wim
Posts: 777
Joined: Wed Apr 28, 2004 5:16 am
Location: canberra, australia

Post by wim » Mon Dec 11, 2006 8:07 pm

hi, it's not a proper review, but there is some mention of it in the HFX mini article
http://www.silentpcreview.com/article678-page1.html
The first component that I listened closely to was mCubed's new EF28 power brick which they introduced in our interview with them back in August. It's basically a fanless external AC-DC convertor that provides up to 210W at 12V for an internal DC-DC adapter. I had to move to within 20cm of the unit in order to be able to pick up a faint electronic buzzing.

Jose Hidalgo
Posts: 81
Joined: Mon Dec 11, 2006 3:25 pm
Location: Lyon, France

Post by Jose Hidalgo » Tue Dec 12, 2006 2:43 am

Yes, thank you wim. I remember that article. And I think the EF28 really deserves to be reviewed, and maybe even recommended by SPCR. I personally don't see another fanless PSU that even comes close to it, becase of its original "internal+external" design that limits heat inside the case. Fanless without the heat : best of both worlds ! :) What do SPCR people think ?

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Tue Dec 12, 2006 4:33 am

I personally don't see another fanless PSU that even comes close to it, becase of its original "internal+external" design that limits heat inside the case.
There are plenty of AC/DC bricks on the market, they only differ in efficiency and capacity. So it would be a pretty short review.

Jose Hidalgo
Posts: 81
Joined: Mon Dec 11, 2006 3:25 pm
Location: Lyon, France

Post by Jose Hidalgo » Tue Dec 12, 2006 4:45 am

Plenty you say ? How many fanless bricks do you know in the 200+ Watt permanent power range ? I only know one (the EF28), but I would be very happy if you gave me some other names. I've spent a lot of time searching for them, and I've only found the EF28 in that power range.

BTW let's remember that the EF28 is not only a power brick but a power brick PLUS an internal DC-DC card. The power brick alone is quite useless.

And I don't see why power bricks should be tested separately. They are PSUs and they should be tested against the other PSUs, don't you think ? All I'm asking is for the EF28 to be tested and added to the big SPCR table of reviewed and recommended PSUs, that's all. ;-)

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Tue Dec 12, 2006 4:57 am

And I don't see why power bricks should be tested separately. They are PSUs and they should be tested against the other PSUs, don't you think ?
Yeah, sure, I don't have a problem with that, but once you start getting into the 100-200W level you are competing with Seasonic SS-300SFD, which is 85% efficient at the 200W level; the brick + DC/DC board is not going to be much more efficient than that, so the only reason to get one is if you have a very small case which will not take a SFX supply. Also it is pretty expensive (£100+). But we will see anyway if SPCR reviews it.

Jose Hidalgo
Posts: 81
Joined: Mon Dec 11, 2006 3:25 pm
Location: Lyon, France

Post by Jose Hidalgo » Tue Dec 12, 2006 5:04 am

jaganath wrote:so the only reason to get one is if you have a very small case which will not take a SFX supply
I think the real reason to get one is (1) to be completely fanless/silent (the Seasonic is not : http://www.silentpcreview.com/article286-page4.html ), and most of all, (2) to get a fanless PSU that will NOT generate any heat inside the case (almost). All internal fanless PSUs do generate a lot of heat, and that's very bad inside a 100% fanless case. With the EF28, not only you don't generate heat, but you save a lot of space, not only for small case reasons, but to allow a better airflow in case of natural convection for instance.

(IMHO of course :oops: )

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Tue Dec 12, 2006 5:13 am

Oh, I absolutely agree, for a totally fanless PC it's a fantastic idea (ie with HFX Classic or Mini); that would make a very cool (!) HTPC. Maybe I should start saving up....:wink:

Jose Hidalgo
Posts: 81
Joined: Mon Dec 11, 2006 3:25 pm
Location: Lyon, France

Post by Jose Hidalgo » Tue Dec 12, 2006 6:11 am

Yes, and that's exactly what I'm trying to build up right now. I'm designing my own anodized aluminium case, that will look a bit like HFX Classic (lateral heatsinks, extensive use of heatpipes) with some improvements. And the HTPC inside will be 100% fanless (all right, it will have 2 fans, but only for security reasons and they shall be turned off under normal operation). :wink:

Charles
Posts: 87
Joined: Mon Nov 14, 2005 8:03 am
Location: Budapest, Hungary

Post by Charles » Sat Dec 30, 2006 4:37 pm

Hello!

I have a Tagan Easycon 480 PSU. Where would you place this on the list mentioned in the article? :)

besonen
Posts: 22
Joined: Tue Dec 12, 2006 10:46 am
Location: Oregon, U.S.A.

PC Power & Cooling reviews

Post by besonen » Sat Jan 06, 2007 11:34 am

has SPCR reviewed any PC Power & Cooling (PCPC) power supplies in the last 12 months?

i noticed a comment in an SPCR article that stated that PCPC power supplies are actually mod'd Fortron power supplies. can someone point me to information that validates this statement?

in the past (years ago) i heard rumor that PCPC was selling mod'd Fortron power supplies but i never found a confirmation. at that time it actually seemed as though Seasonic and PCPC had some identical power supplies (so i concluded one of them was supplying the other).

merlin
Friend of SPCR
Posts: 717
Joined: Mon Jun 13, 2005 6:48 am
Location: San Francisco, CA

Re: PC Power & Cooling reviews

Post by merlin » Sun Jan 07, 2007 12:44 am

besonen wrote:has SPCR reviewed any PC Power & Cooling (PCPC) power supplies in the last 12 months?

i noticed a comment in an SPCR article that stated that PCPC power supplies are actually mod'd Fortron power supplies. can someone point me to information that validates this statement?

in the past (years ago) i heard rumor that PCPC was selling mod'd Fortron power supplies but i never found a confirmation. at that time it actually seemed as though Seasonic and PCPC had some identical power supplies (so i concluded one of them was supplying the other).
I don't believe they have reviewed much by them. But the history of pcpc low end psus is this: They used to use Fortron for their silencer series, now they use Seasonic for the silencers. I have no clue what their Turbo-Cool series is, however I can guarantee it is different from their lower end with very tight voltage regulation. With the seasonic-based builds, I'd expect similar performance to the rest of them out there...

besonen
Posts: 22
Joined: Tue Dec 12, 2006 10:46 am
Location: Oregon, U.S.A.

Re: PC Power & Cooling reviews

Post by besonen » Sun Jan 07, 2007 8:12 am

merlin wrote:now they [pcpc] use Seasonic for the silencers.
how was this determined? i assume PCPC doesn't state it any of their sales brochures?
Last edited by besonen on Sun Jan 21, 2007 12:43 pm, edited 1 time in total.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sun Jan 07, 2007 8:41 am

If they are Seasonics they must be based on older PCB designs, because the lower-end PCP&C Silencers have quite poor efficiency by today's standards (max 74%). The higher-powered models have excellent efficiency however (83%).

besonen
Posts: 22
Joined: Tue Dec 12, 2006 10:46 am
Location: Oregon, U.S.A.

Re: PC Power & Cooling reviews

Post by besonen » Sun Jan 21, 2007 1:09 pm

besonen wrote:
merlin wrote:now they [pcpc] use Seasonic for the silencers.
how was this determined? i assume PCPC doesn't state it any of their sales brochures?
i'd still love to be pointed to some confirmation of this information. especially since it seems to be held as common knowledge here and since i have yet been able to confirm it myself after some searching.

thanks,
david

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Wed Apr 04, 2007 4:48 pm

The PSU Fundamentals article has been updated again. This time, it's a new high power sample system in the REAL SYSTEM POWER REQUIREMENTS section.

Chocolinx
Posts: 311
Joined: Thu Jan 19, 2006 5:14 am
Location: Toronto
Contact:

Post by Chocolinx » Wed Apr 04, 2007 10:03 pm

That's insane! 350W Seasonic PSU can handle a lot! lol. Man I kind of wish I just got that instead of this S12-500 now lol.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Wed Apr 04, 2007 10:21 pm

Chocolinx wrote:That's insane! 350W Seasonic PSU can handle a lot! lol. Man I kind of wish I just got that instead of this S12-500 now lol.
Actually, the point was not that the 350W Seasonic can handle the X1950XTX-512 + oc'd high spec P4 D950 system; the point is that ANY good quality, honestly rated 350W PSU can handle it. The total system power draw was only 256W.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Fri Apr 06, 2007 1:56 pm

Is it an X1900XTX or an X1950XTX? The table states 1950XTX but the text under the table states that an X1900XTX was used. The power consumption between them do differ.

Is it possible that you sometime later will ad a similar setup, using CF or SLI just to see a really worst case scenario.

I've seen tests from techreport.com showing 8800GTX in SLI with an C2D X6800 not using more than 300W during load!

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Fri Apr 06, 2007 2:06 pm

X1950XTX.

SLI -- only if we get another sample... but as that Tech Report article indicates, 8800s in SLI or x1950xtx in xfire don't pull anywhere near the power officially recommended: 373W AC and 400W AC, respectively, from an 80% efficient PSU.

Post Reply