PSU tier list 2.0

[turkey3_scratch]

I've still never seen a PSU review where the reviewer points out bad silicon.

 

[prince_13]

how bout my seasonic m12ii 620w non evo ? what tier is this ?

 

[RCFProd]

how bout my seasonic m12ii 620w non evo ? what tier is this ?

Did you look into the tier list? Or on the frontpage?

 

[toyftw]

We don't have a 'squintskii' registered here. But yes, there are a handful. And that attitude is hardly going to get us more.

+1

 

[toyftw]

Like you said, though, there is so much stuff that needs taken into consideration, like how many units were sold, and what volume of people whose units break come to online forums for help. I kind of want to start telling people on this forum to send me their dead PSUs so I can look inside and see what happened. Also, with dead units, how much the user used the computer and at what level of stress plays a major role.

What data may be better is how many warranties were issued out / units sold, which would be info from the company.

And no company is going to give out that kind of data. Actually that brings to mind an idea if you wanted a boring repetitive task, lol. With thousands of posts regarding PSU failures, you could mine them for PSU failure data. It may not be a very scientific analysis, but you could compare the number of failures of a brand to the total number of posts and start to draw a picture of which brands have more reported problems.

Hi - That data(RMA,etc) is available somewhere. I remeber reading somewhere FSP stating they had the
best RMA ratio in the PSU industry. Not sure how old(& truthful) the data was, but FSP must have had access to
failure rates, etc to make that statement.

 

[turkey3_scratch]

Let's not forget those who don't RMA, for instance. Also, if you have a shorter warranty, you may have less RMAs than if you offer a long one.

 

[Darkbreeze]

Yes, JG works for Corsair. Has for several years.

So you're saying that if you replace the caps in a random cheapo Chinese unit, suddenly it won't have any long-term degradation issues? Silicon can fail under overload/excessive heat too. Bridge rectifiers and the switching transistors being the common ones.

Correlation does not imply causation. You can't say that the PSUs with Chinese caps failing over time means it's actually the chinese caps that are failing, and not the fact that the mfr cheaped out on other stuff like heatsinking, correctly sized components etc. At least not unless you open lots up and find bulged caps.

Or even placement. We've seen poor arrangements before where relatively good quality components failed because they were hidden behind something, out of the path of any airflow, or were too bunched together. Lots of ways they can mess up what might otherwise have been a reliable configuration by a poor design, even when quality components are used.

 

[Darkbreeze]

I don't think a list with failures is going to work. Having people report what brand and model they have, and that it failed, isn't going to tell you if they were using it with hardware that should have had a PSU twice the capacity as what they were using, if the unit was 8 years old and simply had outlived it's useful lifespan, whether other factors such as allowing buildup in very dusty environments without periodic cleaning are involved or even simply the fact that they pooched something themselves by not fully seating a connector or trying to put one where it doesn't belong.

There are probably fifty other reasons why a unit could fail through no fault of it's own, like other faulty hardware that pulls overvoltage due to a short or some other kind of issue, all of which would skew that data beyond usefulness.

 

[turkey3_scratch]

Well then, I give up.

 

[BadActor]

I don't think a list with failures is going to work. Having people report what brand and model they have, and that it failed, isn't going to tell you if they were using it with hardware that should have had a PSU twice the capacity as what they were using, if the unit was 8 years old and simply had outlived it's useful lifespan, whether other factors such as allowing buildup in very dusty environments without periodic cleaning are involved or even simply the fact that they pooched something themselves by not fully seating a connector or trying to put one where it doesn't belong.

There are probably fifty other reasons why a unit could fail through no fault of it's own, like other faulty hardware that pulls overvoltage due to a short or some other kind of issue, all of which would skew that data beyond usefulness.

Maybe so, but all PSU's will be subjected to the same factors. I think over a large enough sample, some clear winners and losers will emerge. The budget PSU's are naturally going to have the worst showing and the most likely to be abused. This may still be useful for demonstrating why it's not a good idea to use them in anything other than low end builds.

 

[Darkbreeze]

All PSUs will NOT be subjected to the same factors. Some people will leave ECO switches on, when gaming under full loads consistently, will use 500w units with cards that really ought to have 650w units, use 750w units with dual card configurations that absolutely require 850-1000w units, may continue to use a faulty driver, even after it has been determined to have problems that result in miscalculated power draw loads. They'll turn PSUs upside down when bottom mounted and intended to have the intake vents facing the bottom where cool air can be used.

Some will never clean their units, despite living in regions where dust or sand are prevalent in the air. Others will use them in top mounted configurations, where they live in heat all the time and use internal heated case air for cooling while under terrific loads.

And those are just the most common probable factors that could skew the results. I do agree that you're likely to at least find a few units that tend to outlast all others regardless of what conditions they are exposed to, but for the most part we already know what those units are anyhow based on extreme hotbox lab testing.

 

[toyftw]

Yes, JG works for Corsair. Has for several years.

So you're saying that if you replace the caps in a random cheapo Chinese unit, suddenly it won't have any long-term degradation issues? Silicon can fail under overload/excessive heat too. Bridge rectifiers and the switching transistors being the common ones.

Correlation does not imply causation. You can't say that the PSUs with Chinese caps failing over time means it's actually the chinese caps that are failing, and not the fact that the mfr cheaped out on other stuff like heatsinking, correctly sized components etc. At least not unless you open lots up and find bulged caps.

Or even placement. We've seen poor arrangements before where relatively good quality components failed because they were hidden behind something, out of the path of any airflow, or were too bunched together. Lots of ways they can mess up what might otherwise have been a reliable configuration by a poor design, even when quality components are used.

Of course you can alleviate this by just removing an MOV (after it's been tested, of course!!)

 

[turkey3_scratch]

For a group regulated unit, if you're pulling let's say 50A on the 12V rail, and 0.5A on the 3.3V and 5V rails, with this crossload, will increasing amperage on the 3.3V and 5V rails actually increase the 12V voltage even if the 12V rail load remains at 50A?

What I'm trying to get at here, is does crossloading of this type induce such a low voltage on the 12V rail only because you can pull more amps from it when there is less load on the minor rails, or do the minor rails themselves actually affect this voltage directly? I think the latter but want to make sure.

 

[Someone Somewhere]

I've still never seen a PSU review where the reviewer points out bad silicon.

Sure, every time they say "this bridge rectifier isn't high enough capacity".

 

[turkey3_scratch]

That's not saying the silicon is bad, it's saying that the silicon can't handle the current the PSU creates. I'm talking about silicon that has[/i[ enough capacity.

 

[Someone Somewhere]

Reviews are always new units, so you won't see degraded silicon in them, plus you can't see it with the naked eye. That's long term; have you ever seen a failed cap in a review of a new PSU? Of course not.

Picking the wrong silicon is akin to picking a 16V cap instead of a 25V one. Will it work initially? Yup. Will it be cheaper? Yup. Will it cause issues down the track? Probably.

 

[turkey3_scratch]

Yeah but how are you going to know if one silicon will degrade faster than another? If they are both rated at the same capacity. Even if silicon degrades, the fan and caps will probably fail long before the silicon does.

 

[Karadjgne]

No way to tell. It's no different than the cpu or ram lottery. The silicon used in any given batch is neither labeled or specific. The only reliance you have is with certain companies using a higher grade of silicon, which is offset by a higher price per final unit, just as Japanese caps use a higher grade of aluminum, but even then, there's no actual guarantee that any given batch ends up on the highest purity of that grade scale, or the lowest.

 

[turkey3_scratch]

So when silicon degrades... How does this happen? Silicon is an element, and there is also n-doped and p-doped for transistors. What exactly happens that causes it to "degrade"? I mean, it's not like the atoms are going to go flying into the void of space.

 

[MasterMace]

So when silicon degrades... How does this happen? ...

What exactly happens that causes it to "degrade"? I mean, it's not like the atoms are going to go flying ...

Usually with heat.


And with heat, that's precisely what happens

 

[turkey3_scratch]

So when silicon degrades... How does this happen? ...

What exactly happens that causes it to "degrade"? I mean, it's not like the atoms are going to go flying ...

Usually with heat.


And with heat, that's precisely what happens

That doesn't explain anything though. I was just sitting out in 90 degree whether and I feel more lively than ever.

 

[Someone Somewhere]

I'm mainly talking about silicon that is rated differently - it's more than just current; withstand voltage applies too. Heatsinking also matters.

It can cause individual silicon atoms to migrate away from where they're supposed to be, or punch a tunnel through the insulator. Heat makes it easier for stuff too move around - it basically gets softer (and remember electrical resistance generally increases with heat). More voltage is more force pushing on each electron/atom.

 

[turkey3_scratch]

It can cause individual silicon atoms to migrate away from where they're supposed to be, or punch a tunnel through the insulator.

Now that's what I'm looking for, thanks 😉 So basically deformation?

 

[Someone Somewhere]

Kinda. Not visible to the naked eye, because it's all happening inside the die. Migration is the technical term; I believe that instead of getting a nice sharp line between different regions, they start to blur.

 

[turkey3_scratch]

So what would make one silicon higher grade than another? Would it be other substances mixed in with the silicon, or something like the smoothness of the silicon and transistors?