The Power Supply Thread

[jonnyGURU]

Sorry I missed the edits. I didn't come back into the thread because there were no new posts.

Originally posted by: GalvanizedYankee

EDIT: The question is. Why do most all of the higher end PSUs have APFC.?

Since the American market is so large and APFC is so non-essential why offer it here
in the first place.?

I think at that point, it's like, "Why not?" I'm quite sure that a manufacturer that has a higher-end product is going to want it to be available around the world. That can't happen without APFC. You're not supposed to even sell a product that doesn't have APFC in the EU. So if you have, let's say, a $100 PSU and only $3 of that makes up the cost of the APFC, is it worth it to have two SKU's, one for the US and one for the EU?

The Dan's DATA link has been posted on this forum many times and no one disputes
it ever.
I look forward to a direct answer or links other than Xbit Labs.

I don't necessarily wish to endorse creedence with all of what Dan has to say. I found that article simply looking for an easy to understand explaination of what PFC is.

EDIT#2

Only because it is so easy to search, I went to Newegg and looked at PSUs w/o PFC,
price range from $65 to $165 and found few that I would seriously consider buying.

I searched again, $65 to $165 w/Active PFC and found many worthy of my consideration
and monie$.

My question that you chose to avoid was not about a particular PSU but about the quality
of PSU construction. It seems that those w/APFC are more highly rated.

I didn't mean or choose to avoid the question. I simply didn't realize that a question was presented to me.

APFC PSU's are more expensive and more expensive PSU's have APFC. That's a fact. My ONLY point, and there isn't anything that you should be read into this at all, is not ALL great choices/value have APFC. You said it yourself; you searched between $65 and $165 and found a few that you would consider. At least that's a few. You said yourself that your PC doesn't have a PSU w/o APFC, and neither does mine.. or my wife's. Will the next one? Maybe. It depends on all of the other features and performance and aesthetics of the PSU.

Many times, someone posts the classic "what PSU would you suggest" thread and I'll ask for system specs, price range, etc and then throw up about a dozen suggestions. 3/4 of them get knocked off the list by the person because of APFC. Although I can respect the CHOICE to go w/ APFC, point it out as a deciding factor and know why you want APFC before you discount everything that DOESN'T have APFC. That's all.

I once used to say, "if the power supply has active PFC then you can be pretty confident that the power supply is of decent quality" but then I ran into a brand that had active PFC that was labeled as a "680W" but kept blowing fuses at 550W. It had max, not sustained, ratings on the label and only promised 10% tolerance! Needless to say, that would not be a recommended power supply.

Last post in this thread.

 

[jonnyGURU]

Originally posted by: computer
Can anyone tell me if such a thing exists?

A PS with adjustable rails, AND a "modular" cable design like these?
http://www.ocztechnology.com/images/products/accessories/b/ModStreamfull.jpg
Even OCZ doesn't. That one at the link doesn't have adjustable voltage, and theirs that DO have adjustable voltage doesn't have the separate cables! Tagan makes a really nice 530 and 580 watt unit with the separate cables, but no adjustable voltage.

Thanks.

IMHO (opinion, and only opinion) adjustable rails are totally a marketing gimmick.

OPINION:

The rails should be correct from the factory and rails should be stable enough to where loads do not cause them to fluctuate enough to require adjustment.

Yeah, yeah, PCP&C Turbo-Cool's have holes that allow for access to pots, but they don't give you documentation explaining adjusting them and I really think those are for the factory to make adjustments after final assembly, not for the end user.

"Well I want to overclock." That's a myth. The regulators that adjust the voltage for RAM, etc. the kinds of reasons why someone might want to adjust their voltages is tolerant of about 10% voltage fluctuation. If you have that kind of fluctuation, you need to chuck the power supply.

 

[computer]

Originally posted by: JEDIYoda
you really should be making a new thread....

This is NOT about particular brands per say...lets keep this thread clean!!

What's your problem?? Who the hell mentioned a brand????? I think you need to read my post again and stop playing "thread police".
Can anyone tell me if such a thing exists?
A PS with adjustable rails, AND a "modular" cable design like these?
The question is, as you should be able to see, does such a thing as described exist. The friggin' URL was ONLY FOR REFERENCE PURPOSES as were THE OTHER COMMENTS.

Go back through these 6 pages and try to tell me that NO ONE mentioned any brands! You can't! I just glanced at only two pages and see THREE brands and models listed on them!

 

[computer]

Originally posted by: jonnyGURU

Originally posted by: computer
Can anyone tell me if such a thing exists?

A PS with adjustable rails, AND a "modular" cable design like these?
http://www.ocztechnology.com/images/products/accessories/b/ModStreamfull.jpg
Even OCZ doesn't. That one at the link doesn't have adjustable voltage, and theirs that DO have adjustable voltage doesn't have the separate cables! Tagan makes a really nice 530 and 580 watt unit with the separate cables, but no adjustable voltage.

Thanks.

IMHO (opinion, and only opinion) adjustable rails are totally a marketing gimmick.

OPINION:

The rails should be correct from the factory and rails should be stable enough to where loads do not cause them to fluctuate enough to require adjustment.

Yeah, yeah, PCP&C Turbo-Cool's have holes that allow for access to pots, but they don't give you documentation explaining adjusting them and I really think those are for the factory to make adjustments after final assembly, not for the end user.

"Well I want to overclock." That's a myth. The regulators that adjust the voltage for RAM, etc. the kinds of reasons why someone might want to adjust their voltages is tolerant of about 10% voltage fluctuation. If you have that kind of fluctuation, you need to chuck the power supply.

I know what you're saying, and I agree one should not have to adjust them, but I've gotten hi-end PS's that were maybe 11.7v on the 12v, still within spec, but a bit too low for my taste. You can't return those because the dealer or manufacturer refuses to take them back because they are still within the +/- 5% spec, I know, I've tried it. Every PS I've used I've had to open it up, find the pots, and adjust at least one of the voltages. Then there is natural semiconductor degradation that will happen over the months & years which will change the voltages, requiring some adjustments. It's just really convenient to have these adjustable pots.

Do all of the Turbo Cool's have those holes?

 

[jonnyGURU]

Originally posted by: computer

Originally posted by: jonnyGURU

Originally posted by: computer
Can anyone tell me if such a thing exists?

A PS with adjustable rails, AND a "modular" cable design like these?
http://www.ocztechnology.com/images/products/accessories/b/ModStreamfull.jpg
Even OCZ doesn't. That one at the link doesn't have adjustable voltage, and theirs that DO have adjustable voltage doesn't have the separate cables! Tagan makes a really nice 530 and 580 watt unit with the separate cables, but no adjustable voltage.

Thanks.

IMHO (opinion, and only opinion) adjustable rails are totally a marketing gimmick.

OPINION:

The rails should be correct from the factory and rails should be stable enough to where loads do not cause them to fluctuate enough to require adjustment.

Yeah, yeah, PCP&C Turbo-Cool's have holes that allow for access to pots, but they don't give you documentation explaining adjusting them and I really think those are for the factory to make adjustments after final assembly, not for the end user.

"Well I want to overclock." That's a myth. The regulators that adjust the voltage for RAM, etc. the kinds of reasons why someone might want to adjust their voltages is tolerant of about 10% voltage fluctuation. If you have that kind of fluctuation, you need to chuck the power supply.

I know what you're saying, and I agree one should not have to adjust them, but I've gotten hi-end PS's that were maybe 11.7v on the 12v, still within spec, but a bit too low for my taste. You can't return those because the dealer or manufacturer refuses to take them back because they are still within the +/- 5% spec, I know, I've tried it. Every PS I've used I've had to open it up, find the pots, and adjust at least one of the voltages. Then there is natural semiconductor degradation that will happen over the months & years which will change the voltages, requiring some adjustments. It's just really convenient to have these adjustable pots.

Do all of the Turbo Cool's have those holes?

Point taken. I guess I'm just spoiled by a big pile of power supplies. I never have one in a build for more than two years and if the voltages are not between 11.9 and 12.1 under load, I just don't use it.

As far as I've seen, PCP&C's have the holes. But I've yet to see a PCP&C need adjustment (even given your scenarios.) An option would be to pick something up by Performance-PC's that's been modified as a modular.

 

[computer]

Originally posted by: jonnyGURU

Point taken. I guess I'm just spoiled by a big pile of power supplies. I never have one in a build for more than two years and if the voltages are not between 11.9 and 12.1 under load, I just don't use it.

As far as I've seen, PCP&C's have the holes. But I've yet to see a PCP&C need adjustment (even given your scenarios.) An option would be to pick something up by Performance-PC's that's been modified as a modular.

I've used a Fortron Source FSP400-60PFN now for about 3 years! When I first got it, it was 11.7v and I had to adjust it, the 3.3 and 5v were fine (since then, the 3.3v has dropped a bit). I swore I'd never upgrade again unless "something totally new that revolutionized the computer market" came along, since my PC now is overkill for what I do (business related email, internet and my websites). The only reason I need a new PS now is I just added an AIW 9800 Pro (to finally make use of my 8x AGP slot, and my AIW 7500 was "stuttering" under massive multitasking), and I think my PS is getting a bit "choked" by the extra power the 9800 needs.

Hmmm, Doesn't look like any of the Turbo Cool PS's have modular cables. :disgust:

 

[jonnyGURU]

That FSP400-60PFN only has 15A on the 12V rail. I think somewhere between the 2.4GHz running at 3.4 and the Radeon 9800 Pro you were taxing the 12V rail a bit more than FSP would like you to. That would explain the drop in voltage.

I think you just need to give a really good power supply a chance... not that the FSP isn't really good... but like I said... you know?

 

[jonnyGURU]

DP!

Wow. I coughed as I clicked the mouse button and it made me double click!!

 

[computer]

Originally posted by: jonnyGURU
DP!

Wow. I coughed as I clicked the mouse button and it made me double click!!

:laugh: My "Tourette's" makes me do that too. LOL.

 

[computer]

Originally posted by: jonnyGURU
That FSP400-60PFN only has 15A on the 12V rail. I think somewhere between the 2.4GHz running at 3.4 and the Radeon 9800 Pro you were taxing the 12V rail a bit more than FSP would like you to. That would explain the drop in voltage.

I think you just need to give a really good power supply a chance... not that the FSP isn't really good... but like I said... you know?

No I think mine has on it 24A on the 12V according to the label. The reason I got it was because it was the top-rated <500 watt PS at AT reviews and two other sites that tested them. It's one of the rare cases where a PS is actually rated LOWER than it's actual specs. For one, it has 430w. If (they) made a modular cable unit w/adjustable voltage and with more power, I'd get it.

I actually have TWO PS's in my case. The Forton, and an additional small AT PS! The AT powers my fans. I use a large server tower case and at the AT supply is on top the ATX, right in front of a 92mm rear exhaust case fan. (There's also 2 other rear fans). I modded the case a bit to make it look professional (as I always do ) Cut a rear hold for the power cable, and cut a hole in the front for a rocker power switch for it.

Anyway, when I decide on a new PS, I can stop having to separately turn on/off that rocker switch for the AT supply since I'll get one powerful enough to run everything.

 

[jonnyGURU]

Really? Hmm... I've seen FSP400's with 24A on the 12V rail, but not the PFN. The PFN is what comes with a lot of the AOpen cases that I used in builds. I never worried about the 15A on the 12V rail because these were builds for SBS servers or Voicemail servers and would often not even have more than on-board video.

 

[computer]

Originally posted by: jonnyGURU
Really? Hmm... I've seen FSP400's with 24A on the 12V rail, but not the PFN. The PFN is what comes with a lot of the AOpen cases that I used in builds. I never worried about the 15A on the 12V rail because these were builds for SBS servers or Voicemail servers and would often not even have more than on-board video.

Maybe it does have 15A, I found one of the reviews and it says 15A here.
http://www.tomshardware.com/2003/06/09/high_power_for_power_users/page19.html
http://www.tomshardware.com/2003/06/09/high_power_for_power_users/page17.html
I was wrong about it having 430w, it's actually 453w. Since it has higher wattage it probably has more than a rated 15A.

If anyone has seen any PS reviews/roundups/tests, that ain't YEARS old, please LMK. Anand, nor Tom's I believe hasn't done any in 3 years. :roll:

 

[Howard]

Originally posted by: jonnyGURU
The Seasonic SS-600HT would actually NOT be a better choice, because that PSU only has two 12V rails and the PCI-e's are both on the same rail and there's no way to move on of them off onto 12V2.

So in a nutshell, if you were to have a problem with an ST65ZF, prior to trying any creative connector manipulation, you'd actually have the exact same problem with the Seasonic only there's nothing you can do about it.

Thank you.

You're right. I should have ignored his question about which was the better PSU.

That said, I'm wondering about some of your comments. The S12-600 PSU didn't have any problems with the test setup... which leads me to believe that it's not that both PCI-E connectors are on the same rail, but that the rails on the S12-600 don't have the 240VA limiter protection. If there isn't a limiter, then there's only one effective rail, right? So then it doesn't matter how much power is being drawn from each "rail".

I don't understand how using a 4-pin to PCI-E adapter would have solved the 5V minimum load issue for the Silverstone PSU, though.

EDIT: Besides, I didn't mention anything about the SS-600HT.

 

[jonnyGURU]

Originally posted by: Howard
The S12-600 PSU didn't have any problems with the test setup... .

Which test setup are we talking about? (Link) There were so many "PSU's don't work with high end SLI" articles last week, I lost track.

If there is no limiter, the PSU could still have two separate rails. The two really aren't synonymous. But typically the limiter is just set higher than what the label shows because anything > 20A would not be ATX12V2.2 compliant. Like Enermax is set to 22A, Ultra X2 is set to 25A, etc. I'm pretty sure that the Tagan 2-Force PSU's don't have limiters. Either that or the limiter is disabled when you throw the switch that runs the two rails in parallel, but I doubt they went through so much trouble.

Some power supplies don't care if they're 12V2.2 compliant. Check out the 30A 12V2 on this bad boy: http://www.nesteq.de/Specification_asm.pdf

So if both PCI-e's were on the same rail and the limiter was set to 25A or non-existant would the PSU pass? Sure. What's the Seasonic's limiter set to if one exists at all? Don't know. Guess I'll have to find an ST-600 and find out!

I also need a link that explains this "issue" with the "minimum 5V load requirement" and the Silverstone. I'm well aware of the minimum load requirements and the assumptions often made about them being "so high," but I've tested two different quad-rail Silverstones because people have exclaimed that the minimum 5V load requirement was too high and have yet to see any ill effects from loading one up to a minimum amperage while juicing the 12V rails all to max. Worst case scenaio is the 12V drops a tad, but no worse than most other power supplies on the market when crossloaded with a high 12V load and low 5V load.

In fact, the whole reason I started doing low 5V crossload tests in my reviews was because someone was "afraid" to use the ST65ZF because of it's high minimum 5V load "requirements."

And I plan to start "SLI certifying" PSU's as part of my tests. I'm making little PCI-e to molex converters I can plug into the load tester so I can load up the Molexes, 2x2, and ATX connector with typical loads, and then whack each PCI-e with 75W loads just to see if it actually passes or not.

 

[jonnyGURU]

To summarize that last post:

1. I do admit that I may be wrong about the Seasonic being able to handle two high end SLI cards and everything else on a system.

2. If it can handle it, then the limiters are set higher than the label spec, or there's no limiter.

3. If a PSU has no limiters on it's rails, that doesn't mean it still doesn't have two rails. But if the rails are run in parallel to make one rail, the limiter would have to be eliminated.

4. I'm aware of Silverstone minimum 5V load requirements, but have yet been shown how they are unrealisticly high or cause issues in themselves.

By all means, I'm open to enlightenment. I haven't learned what I know so far by being stubborn.

 

[Howard]

See table: http://www.extremetech.com/article2/0,1697,1932949,00.asp

And the minimum load issue: http://www.extremetech.com/article2/0,1697,1932948,00.asp

In other words, if you balance the loads on the different rails, then you won't have this kind of problem. So we took PCI Express power adapter cable to give this idea a test. These cables consist of two Molex four-pin connectors on one end, and a six-pin PCIe connector on the other end.

Now, I can understand that this would bypass the limiter on a certain 12V rail, but I have no idea how this is related to the preceding paragraph (which is a quote) about the minimum load on the 5V rail.

If there is no limiter, the PSU could still have two separate rails

I thought all rails drew power from the same inverter/transformer? In that case, what differentiates the different rails?

 

[jonnyGURU]

Originally posted by: Howard

If there is no limiter, the PSU could still have two separate rails

I thought all rails drew power from the same inverter/transformer? In that case, what differentiates the different rails?

Even though the power in a power supply all goes through the same rectifiers, transformers, etc (which is why your PSU's total output is never the sum of all of it's rails.) final regulation is still done by an invidual zener diode and power transistor. These are what we call your actual "rails." Not the DC power feeding these circuits. So if there is more than one regulator regulating 12V DC output, it's still more than one regulator, more than one "rail." And if there is no limiter on this circuit, or circuits, that doesn't mean that the regulation doesn't still take place. If you have two 12V rails and they do not have limiters set on them, that doesn't mean that what's hooked up to the regulator 12V1 is not isolated from 12V2 unless the output of these regulators are put into parallel. Of course, the rails are going to be interactive with each other, but that's the case of any dual-12V-rail power supply simply because the source of the voltage being regulated is the same. One will always rob Peter to pay Paul. This is why you can't add multiple 12V rails to find the total combined 12Voutput of a multi-12V-rail power supply. If you only have 444W available to BOTH of the 12V rails, it really doesn't matter if there are limiters restricting only 20A to one rail and 20A to another or no limiter on either, as long as the total output doesn't exceed 444W. But the actual regulation of each 12V rail, down to 12V, and the inherent filtering and isolation of that rail due to the components used still remains.

I'll read the article and get back to you on the rest. I have to give my girl a bath right now.

 

[Howard]

Originally posted by: jonnyGURU

Originally posted by: Howard

If there is no limiter, the PSU could still have two separate rails

I thought all rails drew power from the same inverter/transformer? In that case, what differentiates the different rails?

Even though the power in a power supply all goes through the same rectifiers, transformers, etc (which is why your PSU's total output is never the sum of all of it's rails.) final regulation is still done by an invidual zener diode and power transistor. These are what we call your actual "rails." Not the DC power feeding these circuits. So if there is more than one regulator regulating 12V DC output, it's still more than one regulator, more than one "rail." And if there is no limiter on this circuit, or circuits, that doesn't mean that the regulation doesn't still take place. If you have two 12V rails and they do not have limiters set on them, that doesn't mean that what's hooked up to the regulator 12V1 is not isolated from 12V2 unless the output of these regulators are put into parallel. Of course, the rails are going to be interactive with each other, but that's the case of any dual-12V-rail power supply simply because the source of the voltage being regulated is the same. One will always rob Peter to pay Paul. This is why you can't add multiple 12V rails to find the total combined 12Voutput of a multi-12V-rail power supply. If you only have 444W available to BOTH of the 12V rails, it really doesn't matter if there are limiters restricting only 20A to one rail and 20A to another or no limiter on either, as long as the total output doesn't exceed 444W. But the actual regulation of each 12V rail, down to 12V, and the inherent filtering and isolation of that rail due to the components used still remains.

Ah, gotcha.

 

[jonnyGURU]

Originally posted by: Howard
See table: http://www.extremetech.com/article2/0,1697,1932949,00.asp

I don't know how the Seasonic passed. Maybe because the limiter is actually set higher than the label spec. That's good. Intel ATX12V2.2 is crap right now anyway. One year ago, they wouldn't have taken into consideration high power PCI-e cards. Heck! If they did there'd be SOME mention of them in the specification! 20A per rail is simply unrealistic.

Originally posted by: Howard

And the minimum load issue: http://www.extremetech.com/article2/0,1697,1932948,00.asp

In other words, if you balance the loads on the different rails, then you won't have this kind of problem. So we took PCI Express power adapter cable to give this idea a test. These cables consist of two Molex four-pin connectors on one end, and a six-pin PCIe connector on the other end.

Now, I can understand that this would bypass the limiter on a certain 12V rail, but I have no idea how this is related to the preceding paragraph (which is a quote) about the minimum load on the 5V rail.

It's more a factoid blurb than a caveat. Just pointing out the nature of a switch mode power supply and the typical "problems" you encounter witha dual-rail are going to be doubled with a quad-rail. What they're saying is for you to get maximum performance out of your 12V rail, you need to have a minimum load on the 5V rail. That's all. And since most machines that are "performance" in the sense of high end video cards, CPU, etc. resulting in a high draw on the 12V rail also have a number of drives to help "pad" that 5V rail some, then "crossloads" shouldn't be an issue.

I guess if you had a PC with two 7800GTX's and an X2 or Pent D and only one stick of RAM and one hard drive you might run into a problem. But who's going to do that?

 

[Howard]

Originally posted by: jonnyGURU
I guess if you had a PC with two 7800GTX's and an X2 or Pent D and only one stick of RAM and one hard drive you might run into a problem. But who's going to do that?

Somebody who buys a computer from Best Buy?

 

[Dorkenstein]

Is it okay to use a PSU with only one 12v rail?

 

[jonnyGURU]

Originally posted by: Dorkenstein
Is it okay to use a PSU with only one 12v rail?

Why wouldn't it be?

If it's a well made power supply, I would have no reservations.

 

[Luckyboy1]

This is such an excellent thread. It has been plainly laid out, but still, there seems to be some confusion. I can tell this by people asking questions like...

Is XYZ a good power supply brand?

You CANNOT simply add up the watts and turn off your brain and order a certain brand because it simply doesn't work that way in most cases. Most Power supply marketers sell several levels of quality. Take for example, my OCZ 520 Watt Pwerstream power supply is not only 30% more powerfull than the OCZ 520 Watt Modstream power supply, it is also shielded better and has more dedicated rails for video cards and hard drives. As a result, you have to read the complete specs on a power supply. If the vendor sends you on a long journey just to find the complete specs, you can immedialtely smell a rat in most cases.

Then you can easily take off 20-30% of the power supplies capabilities if it is housed like most computer cases, at the top of the case. Some of the Lin whatever cases actually are designed with the power supply mounted down low to help in this matter. Remember, your case will run 10-20 F hotter at the top of the case than the bottom. For every 10 F higher temp, you can knock off 10% in the power suppy's ability to deliver juice.

 

[JEDIYoda]

Originally posted by: Luckyboy1
This is such an excellent thread. It has been plainly laid out, but still, there seems to be some confusion. I can tell this by people asking questions like...

Is XYZ a good power supply brand?

You CANNOT simply add up the watts and turn off your brain and order a certain brand because it simply doesn't work that way in most cases. Most Power supply marketers sell several levels of quality. Take for example, my OCZ 520 Watt Pwerstream power supply is not only 30% more powerfull than the OCZ 520 Watt Modstream power supply, it is also shielded better and has more dedicated rails for video cards and hard drives. As a result, you have to read the complete specs on a power supply. If the vendor sends you on a long journey just to find the complete specs, you can immedialtely smell a rat in most cases.

Then you can easily take off 20-30% of the power supplies capabilities if it is housed like most computer cases, at the top of the case. Some of the Lin whatever cases actually are designed with the power supply mounted down low to help in this matter. Remember, your case will run 10-20 F hotter at the top of the case than the bottom. For every 10 F higher temp, you can knock off 10% in the power suppy's ability to deliver juice.

"Then you can easily take off 20-30% of the power supplies capabilities if it is housed like most computer cases, at the top of the case."

Please post supporting documentation...thx

Because that statement - just is not true. Alot has to do with the Power Supply.
Also more has to do with what type of ventilation you have...
Sorry but thats just not is a true or factual statement!!

 

[Luckyboy1]

Well, even something as simple minded as this PC Power & Cooling page tells you that ambient temperatures dramatically affect most power supplies...

http://www.pcpowerandcooling.com/technology/terminology/

not only will a 10 F drop in temp increase the unit's efficiency, it will increase its lifespan as well.

Yes, of course, case cooling affects power supply output, so you make my point for me.

I'm not going to dig for obscure sources to prove the point further on a subject that is common knowledge in EE circles.