Help with psu please

[pavle]

LLC resonant is the best option indeed.
I must correct myself: FSP ACRF topology is IMO too complicated and upon looking at a review, where ripple suppression can sometimes be an issue (at transient loads),
I'd advise buying a normal two transistor forward design or even better, LLC resonant and if it's Corsair, even better; I have quite a few of their PSUs and they're very good.

 

[ExcuseMeWtf]

be-quiet! Pure Power will offer lowest noise output of all of them and is most efficient.

CX is viable alternative if you don't care about that in particular.


Out of cheaper ones Corsair CV650 is full range* and supposed to be DC-DC based according to JonnyGURU, who works at Corsair, so I'd go with this. But not CV550 since it's older group regulated design.

*Even if you live in 230 VAC area, being full range can be useful, if you experience frequent brownouts. PSUs with limited input range (200-240 VAC or so) tend to have weaker primary components and with brownouts those could get overloaded and PSU's lifespan would get shortened

 

[dirtyferret]

The CX have variable bom - some are made by CWT, some by GW (according to this, at the very least ...) - it's possible the GW ones all dud out and drag the entire series down with them.

GW's take on the CX650, fairly impressive considering the budget constraint. Silver efficiency, 105c caps with Rubycon, elite, and a few teapo. Too bad it won't be available in the states.

 

[Bill_Bright]

The CX have variable bom

Variable what? "Bom" is not a standard "electronics" term. In my shop we use it on the consulting side of the business when we submit a bid for a new contract. There it stands for a "bill of materials".

Please clarify. Did you mean OEM? If so, then that makes sense.

Too bad it won't be available in the states.

I don't know how many other companies do this (two very different models, made by two different OEM suppliers, clearly using different designs but with the same model number) but I find this very frustrating.

GW's take on the CX650, fairly impressive considering the budget constraint

"Fairly impressive"? I don't agree. The ATX Form Factor standard for ATX power supplies establishes the "minimum" requirements that all ATX PSUs, even budget models, must meet in order to be ATX compliant. That GW CX650 failed!

"Hold-up time" is the amount of time a PSU maintains full continuous output voltages during a power interruption. The ATX Form Factor standard requires ATX compliant PSUs to "hold up" output power for at least 17ms during such power interruptions. Note as seen here on page 20, section 3.2.9, that is a "requirement", not a "recommendation" as that Tom's article and JG state.

That CS 650W supply failed miserably, holding output for only 8.3ms - less than half the required time! That means any loss of power (or a simple dip or sag below 90VAC) of even 10ms (which is MUCH faster than the human eye can detect) will result the computer suddenly crashing, potentially resulting in lost data and/or a corrupt drive.

Power dips (opposite of spikes) and sags (opposite of surges) and brownouts (long duration sags) are much more common than most people think - even with stable power grids. And such power anomalies don't have to originate off the grid. A refrigerator or AC unit cycling on, for example, can cause a voltage drop on that circuit.

Consumers need to expect - or rather demand that any PSU claiming to be ATX Form Factor compliant must meet the minimum specs laid out in that standard. PERIOD!

That means that entry level PSUs, that is, those built with "budget constraints" must meet those minimum specs too. It is up to the upper tier models to exceed those specs - which many do, with ease!

That Corsair failed. It is not ATX compliant. I do not recommend anyone buy it. If you do, make sure you put it on a "good" UPS with AVR. And keep regular backups of your data.

 

[Selaya]

Yeah, I meant bill of materials with it, I guess OEM may have been more correct (different OEM kinda implies different bill of materials, does it not xd).

 

[Bill_Bright]

(different OEM kinda implies different bill of materials, does it not xd).

No. Not really.

Many companies "outsource" different OEMs for their products or the parts they use to assemble their products. They do this when one factory or assembly plant does not have the capacity to meet demand. Or they may do this to ensure their supply chain is not totally halted should an Earthquake, for example, destroy the OEM facility.

Car manufacturers often do this for their most popular models. So do hard drive makers. But typically, the same model # uses the exact same design and parts (BOM). The only difference is the factory they were "assembled" in.

In some cases, for example, cars, different factories are located in different countries because the market is different. For example, left hand or right hand driving, or different emissions standards.

In other cases, even if manufactured in different countries, they are the exact same. This is especially true with computing electronics because globally, they all comply with the same universal standards. Motherboards and CPUs, for example all use the exact same DC voltages, support the exact same operating systems, programs, peripherals, etc. Most power supplies, including these Corsairs, have universal power inputs. So if you live in a country that uses a 100VAC, 120VAC, or 240VAC, 50 or 60Hz grid, the exact same PSU can be used. All that is different is the physical shape of the plug that goes into the wall outlet.

In this case here, it appears the two Corsairs, even though they have the same model number, are different inside, more than just the name of the OEM supplier.

 

[dirtyferret]

That CS 650W supply failed miserably, holding output for only 8.3ms - less than half the required time! That means any loss of power (or a simple dip or sag below 90VAC) of even 10ms (which is MUCH faster than the human eye can detect) will result the computer suddenly crashing, potentially resulting in lost data and/or a corrupt drive.

At full load so what are the chances the power goes out at full load and wouldn't your recommended UPS solve the issue? Regardless, yes it's obviously a cost cutting move.

I don't know how many other companies do this (two very different models, made by two different OEM suppliers, clearly using different designs but with the same model number) but I find this very frustrating.

You would be surprised, we do it also. Netgear with routers is another example, chaning factories and internal parts while keeping the same basic model number.

 

[Bill_Bright]

No, I would not be surprised. I didn't say no other companies do it. I just said I don't know how many do.

At full load so what are the chances the power goes out at full load

Full load? I never said anything about full load. Regardless, the load, full or otherwise, has nothing to do with the hold up "time" where the PSU is required to provide all output voltages (again, regardless of load - as long as the load does not exceed published capacity) during such power interruptions that don't exceed 17ms.

wouldn't your recommended UPS solve the issue?

Ummm, that's why I recommended it. But compensating for inadequate power supplies that fail to meet industry standards is NOT the reason for using a "good" UPS with AVR.

And cost cutting moves are no excuse for failing to comply with industry standards either. Ask Volkswagen. If any company claims their products complies with a standard, it should comply with that standard. Period.

If that supply only achieved 78% efficiency at 50% load, should they still be able to claim 80PLUS compliance? No! So why should ATX compliance be any different?

 

[dirtyferret]

Full load? I never said anything about full load.

You need to read your own links Bill, from your own documentation that you linked above.

Electrical20 336521 - 0023.2.8+5V DC / +3.3V DC Power Sequencing - Required
The +12V1 DC / +12V2 DC and +5 VDC output levels must be equal to or greater than the +3.3 VDC output at all times during power-up and normal operation. The time between any output of +12V1 DC / +12V2 DC and +5 VDC reaching its minimum in-regulation level and +3.3 VDC reaching its minimum in-regulation level must be ≤20 ms as shown in the below figure. Figure 3-2: Power Supply Timing3.2.9Voltage Hold-U p Time - Required
The power supply should maintain output regulations per Table 3-2 despite a loss of input power at the low-end nominal range-115 VAC / 47 Hz or 230 VAC / 47 Hz – at maximum continuous output load as applicable for a minimum of 17ms (T5+T6).

 

[Bill_Bright]

And you need to read and understand what that means. Apparently, you don't.

It does NOT mean "only" at maximum continuous output load. It simply means at any load "up to and including" the maximum load, the PSU must maintain the output voltages.

That means if the computer is only pulling 50% of the PSU's rated capacity, the PSU must still maintain the output voltages (ie, +3.3VDC, +5VDC and +12VDC) any time the input voltage is interrupted for at least 17ms. Same with 10%, load, 30% load, 64.2% load all the way up to 100% load.

So again, I never said anything about full load. But regardless, it does not matter. What matters is,

Regardless, the load, full or otherwise, has nothing to do with the hold up "time" where the PSU is required to provide all output voltages (again, regardless of load - as long as the load does not exceed published capacity) during such power interruptions that don't exceed 17ms.

 

[John Naylor]

One of the things to consider when selecting a product is "who is reviewing it" If I don't see a review of a monitor on tftcentral, it falls into the category of "not interested" as it's unparalleled in the depth of their reviews. When you see a PSU being reviewed it's either because the manufacturer set them a free one or they obtained one (by request or purchase) because it's of significant interest to their subscriber base. And one thing about PC enthusiasts are guilty of is that they tend to be loyal to their earliest data sources and they tend not to "let go" of outdated info. You still see peeps talking about Rad Fans as if we still had the 30 fpi rads from the 90s. Brand loyalty and source loyalty does not account for the fact that "things change" and that in this field, they change faster than others. There are no bests .... only bests for the particular application(s) and best at this point in time.

Our old trusted sites are not "the same". The Purch Group started as a "product placement entity" and has tried to mask that origin since as it began purchasing independent PC review web sites. Hard to take advice from a company that made itself buy taking money to get product placements in the media. As Joe Pesci said in My Cousin Vinny ... "I got no use for dis guy".

But one thing you will notice about the CX series on the Real Hardtech site

15 reviews 2009 - 2012
4 reviews 2013- 2014
5 reviews 2015 - 2016

In 2015 , HardOCP gave the CX750 a FAIL rating / TPU gave it an 8.3 in 2012
Jonnyguru gave the CX 750 an February 2015, and 8 later that year in may and an 8.7 in 2016 ....

Not exactly consistent.

Also... I like to look at user reviews. Granted there are going to be bad reviews from those who blame their lack of knowledge and experience on the product . But when I look at newegg at the CX750M and see only 44% 5 egg ratings and 22% 1 egg ratings, that's not a good sign. And when ya can grab the Seasonic Focus Gold within $10 or so o the CXM, Im gonna get the Seasonic. Hoever these days with the price of PSUs and ither components being what they are, waiting the game we're playing most at this time.

 

[Deleted member 193596]

Be Quiet Pure Power 11 600W - Gold
maybe look on a price comparison site for a 750W one (sometimes they are just 10-15 bucks apart)

 

[Bill_Bright]

There are no bests .... only bests for the particular application(s) and best at this point in time.

^^^THIS^^^ I wish everyone understood this. And I wish everyone understood this applies to software applications too. For example, there is no one browser or security app or OS that is best at everything.

There is nothing wrong with brand loyalty, BTW - as long as it does not blind you to other equally or even better alternatives.

Brand loyalty and source loyalty does not account for the fact that "things change" and that in this field, they change faster than others. There are no bests .... only bests for the particular application(s) and best at this point in time.

Ummm, I think you forgot that truism and fact-of- life when you made the very biased and contradictory claim of,

If I don't see a review of a monitor on tftcentral, it falls into the category of "not interested" as it's unparalleled in the depth of their reviews.

TFTCentral may be a decent review site, but to summarily dismiss all others in this "brand loyalty" manner is very contradictive. "There are no bests", remember? Maybe this day, but not that day. That is, no one product (or site) that is best at all points in time. Even if that site focuses their reviews on one category only.

If you only read one professional review, its like getting your news from only one news source; you only get one side of the story. And most likely, that review came from reviewing just one sample of the product - a sample that might have been cherry picked to get a good review, or conversely, a sample that just happened to be a lemon. Testing multiple samples increases accuracy as even the best makers can produce a lemon on occasion. And will do so until Man can create perfection 100% of the time.

As for user reviews like those on Newegg or Amazon, look at user reviews but do NOT put too much stock in them. With user reviews, it is important to note happy users don't complain and don't generally write good reviews either. And most reviewers don't come back after they have used the product for a while to write reviews. Also, users typically don't have comparable products to compare with for proper side-by-side (A-B) comparisons, nor do they have properly equipped testing facilities, or the necessary technical training for a proper evaluation.

Plus often times you will see products down-rated because it was a different color than shown on Newegg, the Post Office delivered to the house next door, UPS delivered it a day late, or the box looked like it fell off the FedEx truck. Those reviews skew the results unfairly. So if I am considering a product, I read the user reviews but I don't put a lot of stock in them UNLESS there are several complaining about the EXACT SAME genuine fault/defect with that actual product.

Edit comment: fixed a couple typos.

 

[manos]

So, i raised my budget a bit and now i m between
1. Seasonic CORE-GM-650
2. Corsair TX-M Series TX650M
Both cost the same. Which one you suggest?

 

[Mr.Scott]

So, i raised my budget a bit and now i m between
1. Seasonic CORE-GM-650
2. Corsair TX-M Series TX650M
Both cost the same. Which one you suggest?

They're virtually identical, including warranty. Choose your favorite brand.
Me?, I'm a Seasonic guy.

 

[dirtyferret]

So, i raised my budget a bit and now i m between
1. Seasonic CORE-GM-650
2. Corsair TX-M Series TX650M
Both cost the same. Which one you suggest?

It's the gold TX650M and not the bronze? If so, get the cheaper of the two as you won't notice a difference.

 

[Bill_Bright]

I hate fan noise. The Seasonic has a hybrid/silent mode where there fan stays off until a specific thermal threshold is crossed. It is my experience that this feature works great and for most tasks like "working" the forums, processing emails, editing Word docs or Excel spreadsheets, surfing the Internet, those fans never need to spin up. This assumes, of course, the PSU is properly sized and case cooling is setup properly.

So for me, I would get the Seasonic.

 

[ExcuseMeWtf]

I hate fan noise. The Seasonic has a hybrid/silent mode where there fan stays off until a specific thermal threshold is crossed. It is my experience that this feature works great and for most tasks like "working" the forums, processing emails, editing Word docs or Excel spreadsheets, surfing the Internet, those fans never need to spin up. This assumes, of course, the PSU is properly sized and case cooling is setup properly.

So for me, I would get the Seasonic.

Seasonic Core does not have fanless mode though.

 

[Bill_Bright]

Did you follow your own link? Scroll down and it clearly says,

S2FC - SMART AND SILENT FAN CONTROL

 

[Assimilator]

Did you follow your own link? Scroll down and it clearly says,

Which is not fan-off mode, as the graph on that same page shows. The fan will always spin, but not at maximum RPM.

 

[dirtyferret]

Did you follow your own link? Scroll down and it clearly says,

according to seasonic, S2FC is not fanless. Hence the 7 year warranty vs the ten year on the plus series

 

[manos]

So if we put aside the fanless mode, which one do you prefer, Seasonic CORE-GM-650 or Corsair TX-M Series TX650M?

 

[dirtyferret]

So if we put aside the fanless mode, which one do you prefer, Seasonic CORE-GM-650 or Corsair TX-M Series TX650M?

What's the price, I know you said it's close but how close?

 

[Bill_Bright]

Gee whiz! It seems folks are so eager to show where someone else is wrong, they don't even look at their own sources!

Which is not fan-off mode, as the graph on that same page shows. The fan will always spin, but not at maximum RPM.

according to seasonic, S2FC is not fanless.

Many PSUs have a feature where the fan does not start to spin until a certain thermal threshold is crossed inside the PSU.

DF - If you look at your own graph, it clearly says "Fanless" Mode! And look at the graph and at the magenta line, it does indeed starts at 0 RPM. That's off or "Fanless" to me.

Assimilator - If you look at that graph too, note also the "*" and how it says " *Fan mode activation points are estimated." If not activated to until those points, that means is NOT spinning until those points.

Now if you guys don't believe your own sources, go tell Seasonic they are wrong, not me! But please don't refer to graphs and links that don't even support your inaccurate claims.

So if we put aside the fanless mode, which one do you prefer, Seasonic CORE-GM-650 or Corsair TX-M Series TX650M?

Well, I have had inconsistent service from Corsair so I personally prefer Seasonic. But that's me.

 

[manos]

So if we put aside the fanless mode, which one do you prefer, Seasonic CORE-GM-650 or Corsair TX-M Series TX650M?

Both cost 90 euros