Anandtech to start doing PSU reviews

[Blain]

Originally posted by: Acanthus
They are not above my head, give me one case where ripple matters at all, ever.

You dont power up your PC and go "wow this PSU has a much cleaner power signal!".

You dont see the difference, it doesnt affect performance, and if it works, why does it matter at all?

Data corruption. Voltage ripple is a component of PS "stability".

 

[ForumMaster]

Originally posted by: Acanthus

Originally posted by: rise
once again, the communication process has broken down...

Sorry, not used to all of the new categories yet.

That does kind of reinforce my point... There are very few factors people really have questions about.

How many rails it has is generally completely meaningless, you just add them together, and having multiple rails has never shown more stability than a single rail.

So you look at the amps, you look at the brand, and you estimate how much power you need...

The only real qustion out there, is which ones dont blow up after 6-12 months.

well i'm not sure if it really was the ripple (i think that when the voltage isn't stable?)
but my sister's PC had this crappy PSU and was really unstable. it would crash often. we replaced the PSU with an average one of the same capacity (i know different PSU's provide different real capacities) and all the problems disappeared. we did this cause the mobo reported unstable voltages.

 

[Acanthus]

Originally posted by: Blain

Originally posted by: Acanthus
They are not above my head, give me one case where ripple matters at all, ever.

You dont power up your PC and go "wow this PSU has a much cleaner power signal!".

You dont see the difference, it doesnt affect performance, and if it works, why does it matter at all?

Data corruption. Voltage ripple is a component of PS "stability".

I have *never* seen data corruption because of ripple.

 

[Acanthus]

Originally posted by: ForumMaster

Originally posted by: Acanthus

Originally posted by: rise
once again, the communication process has broken down...

Sorry, not used to all of the new categories yet.

That does kind of reinforce my point... There are very few factors people really have questions about.

How many rails it has is generally completely meaningless, you just add them together, and having multiple rails has never shown more stability than a single rail.

So you look at the amps, you look at the brand, and you estimate how much power you need...

The only real qustion out there, is which ones dont blow up after 6-12 months.

well i'm not sure if it really was the ripple (i think that when the voltage isn't stable?)
but my sister's PC had this crappy PSU and was really unstable. it would crash often. we replaced the PSU with an average one of the same capacity (i know different PSU's provide different real capacities) and all the problems disappeared. we did this cause the mobo reported unstable voltages.

Yes, there is a very real difference between a Dell 300w, and generic 300w, and an antec 300w.

You can tell just looking at the amperage ratings on each rail.

 

[JEDIYoda]

Originally posted by: Acanthus

Originally posted by: ForumMaster

Originally posted by: Acanthus

Originally posted by: rise
once again, the communication process has broken down...

Sorry, not used to all of the new categories yet.

That does kind of reinforce my point... There are very few factors people really have questions about.

How many rails it has is generally completely meaningless, you just add them together, and having multiple rails has never shown more stability than a single rail.

So you look at the amps, you look at the brand, and you estimate how much power you need...

The only real qustion out there, is which ones dont blow up after 6-12 months.

well i'm not sure if it really was the ripple (i think that when the voltage isn't stable?)
but my sister's PC had this crappy PSU and was really unstable. it would crash often. we replaced the PSU with an average one of the same capacity (i know different PSU's provide different real capacities) and all the problems disappeared. we did this cause the mobo reported unstable voltages.

Yes, there is a very real difference between a Dell 300w, and generic 300w, and an antec 300w.

You can tell just looking at the amperage ratings on each rail.

You just don`t get it do you???
Even with different amperage rating you coul still have 3 totally different PSU`s....from the standpoint of quality and components used.....
If we flip the coin over you could also have totally the same units from th stand points of quality and compnents used....

You cannot dummy down which PSU to buy strictly by what you say are important verses what johnnyGURU says......

Again I say -- it apears you have no clue other than to say if you plug it in and it works then all is good....hmmmm

 

[Acanthus]

Originally posted by: JEDIYoda

You just don`t get it do you???
Even with different amperage rating you coul still have 3 totally different PSU`s....from the standpoint of quality and components used.....
If we flip the coin over you could also have totally the same units from th stand points of quality and compnents used....

You cannot dummy down which PSU to buy strictly by what you say are important verses what johnnyGURU says......

Again I say -- it apears you have no clue other than to say if you plug it in and it works then all is good....hmmmm

Because that is exactly how it works with power supplies.

Its akin to reviewing fans almost... Beyond "working under X load" and "MTBF" there are no other meaningful statistics... You havent elaborated on why im wrong yet...

See ATs 1st PSU review up now... what meaningful info do you REALLY get from it?

"it works under x load"
"the safety features work"
"its $160"

Oh and thanks for the know-it-all comments about "not knowing what im talking about" because i didnt elaborate on the differences between the generic example i gave... Of course they are internally different...

Barring a dishonest OEM, all of these PSU reviews are going to look the same.

 

[Acanthus]

Originally posted by: JEDIYoda


After reading this whole thread seems to me like people need to just ignore the comments made by Acanthus!!
It`s obvious the technical aspects of PSU testing are shall we say above his head...

So youre saying things like efficiency are vital? And Ripple?

Aside from being written on the damn box if you really do care, does the $2 on your power bill really matter?

I do agree with the other poster that seeing whats under the hood does matter, like the capacitor makers... etc. But this is only because it can help you estimate the MTBF with a wild stab in the dark type of guess... It still doesnt solve the issue of the most prevalent statistic that people want to know isnt going to be done in any way by these reviews.

 

[GalvanizedYankee]

High ripple must be delt with by capacitors down stream, on the main board, on any slot cards and in the opticals. When the PSU is sending out dirty current other comp hardware life may suffer. The hotter(quieter) the rig is run the greater the suffering. In this context ripple does matter. Rigs will OC better with clean stable current.

We sure love fighting over PSU issues.

 

[Acanthus]

Originally posted by: GalvanizedYankee
High ripple must be delt with by capacitors down stream, on the main board, on any slot cards and in the opticals. When the PSU is sending out dirty current other comp hardware life may suffer. The hotter(quieter) the rig is run the greater the suffering. In this context ripple does matter. Rigs will OC better with clean stable current.

We sure love fighting over PSU issues.

I will concede that in a combination of crap components (ECS mobo + generic PSU + very high load) then it could come into play.

But in our bickering in this thread ( ) going with decent components pretty much averts this problem.

You dont need a $160 power supply to push a high end machine, likewise, you dont want to push a high end machine with a $30 550w or those freebies that come with low end cases.

Common sense tells you what to do in these situations, i dont think we really need numbers for that.

 

[maluckey]

Acanthus,

Common sense tells me not to listen to what is on the box, and instead listen and look for what the unit in question is built like and capable of. A manufacturer will issue claims based off of unreasonable testing and temperatures given the chance. Testing and publication...hanging out their dirty laundry so to speak is a good way to call their bluff.

you go ahead and buy your PSU based on manufacturers claims, word of mouth and pricing. You might get lucky. I won't take that chance.