JonnyGURU Reviews Corsair HX620 620W PSU

[perkam]

Check it out: http://www.xtremesystems.org/forums/showthread.php?t=109412

You can give feedback to him in that thread :) Not a bad PSU at all at $169.

Perkam

[nn_step]

this worries me
http://www.jonnyguru.com/PSU/HX620W/Oscope/Test5-12V3.jpg
and it gets a "9" :confused:

[[XC] Lead Head]

this worries me
http://www.jonnyguru.com/PSU/HX620W/Oscope/Test5-12V3.jpg
and it gets a "9" :confused:

...That looks like a sine wave...if its DC it should be completely flat...

[arisythila]

a sine wave is a bit different, but it defentally has the characteristics of a sine wave. DC should be completely flat... (true DC) but we run into problems with AC to DC sometimes.. Basically, when you convert the two, you sometimes have this problem when you dont filter it enough.

I wouldnt recommend this PSU for anybody.

~Mike

[WeStSiDePLaYa]

a sine wave is a bit different, but it defentally has the characteristics of a sine wave. DC should be completely flat... (true DC) but we run into problems with AC to DC sometimes.. Basically, when you convert the two, you sometimes have this problem when you dont filter it enough.

I wouldnt recommend this PSU for anybody.

~Mike


well you can look inside and see that there is not many large inductors, which are used to filter AC ripple.

not to mention the AC line filtering looks pretty weak.

[jonnyGURU]

this worries me
http://www.jonnyguru.com/PSU/HX620W/Oscope/Test5-12V3.jpg
and it gets a "9" :confused:

I wouldn't poke fun.

You haven't seen what the GameXstream and FSP Epsilon's look like under that same scope. ;)

[Nedjo]

well show us!!

[Oklahoma Wolf]

I've seen worse scope results, and the FSP/OCZ is where I've seen them ;)

[jonnyGURU]

well show us!!

I need to find out where my OCZ 700W RMA is. I sent that thing back two weeks ago. :(

[[XC] Lead Head]

I need to find out where my OCZ 700W RMA is. I sent that thing back two weeks ago. :(


This it?

http://photos1.blogger.com/blogger/4411/1898/1600/Test5-12V2.jpg

[arisythila]

well you can look inside and see that there is not many large inductors, which are used to filter AC ripple.

not to mention the AC line filtering looks pretty weak.

yeah, looks pretty :banana::banana::banana::banana:ty...

~Mike

[arisythila]

I need to find out where my OCZ 700W RMA is. I sent that thing back two weeks ago. :(

OCZ has made good PSU's before, I hope the GamerX isnt this bad... (I have a GamerX PSU)

~Mike

[Otaking71]

if you look at the that last chart....and notice the scale...it's only a .1v flux...that's nothing at a full load. (or is it .5?)

[WeStSiDePLaYa]

if you look at the that last chart....and notice the scale...it's only a .1v flux...that's nothing at a full load. (or is it .5?)


no, its not talking about voltage drop, as those are in different chart.

what we are talking about is line ripple.

[Praxis1452]

which it seems is what he meant by flux westside..

I think it's .5 voltage ripple. Not too bad at full load.

[Oklahoma Wolf]

Ripple is well inside rated specs.

not to mention the AC line filtering looks pretty weak

This one uses a forward converter (which is why I think it's a Seasonic)... circuit design will be different from most PSU's. Wouldn't shock me at all if the new PC P&C Silencer 750W is built the same way.

[WeStSiDePLaYa]

Ripple is well inside rated specs.


but even with the new atx standards, there is allowed 5%(+/-) variance on the rails. which most overclockers would find unnacceptable, especially for the prices of psu's today.

so saying that its inside rated spec isnt saying much.

EDIT: also i wouldnt call that well within spec, more like just within spec. and like ive already said, people here dont just want better than the minimum requirements, they expect it.

[Oklahoma Wolf]

people here dont just want better than the minimum requirements, they expect it.

At full power? That rules out an awful lot of PSU's.

[jonnyGURU]

100mV ripple is a widely accepted tolerance for most power supplies. Even cheapies.

The Corsair only exhibited 100mV (not even) ripple under test 5.

I didn't want to throw OCZ's GameXstream under the bus because OCZ is a great company and there have been no reports of wide-spread problems with the GameXstream, but the 700W I had exhibited nearly 100mV ripple on ALL tests and exceeded 100mV ripple during the last test.

And to prove that this wasn't an isolated incident, I tested an FSP 600W and it exhibited the exact same ripple.

Now... not to show biased for Corsair and against OCZ, I want you to take a look at some of the other PSU's I've reviewed and scoped. There's a reason I keep the Oscope to the same scale regardless of the PSU I'm testing.

It gets worse. Believe me. I've got about three PSU's on the bench know that I'm typing up copy for. It gets a lot worse. ;)

If you're basing my numbers on numbers seen elsewhere... don't. It won't jive.

I've seen reports of as little as 40mV on certain websites discussing certain power supplies that I've scoped and found twice the ripple. This ripple is measured at the PSU's PWM chip. My ripple is measured at the output. As much credit I have to give ANY website that measures ANY ripple at all, this comparison is a kin to measuring voltages inside the PSU on the PCB versus at the tip of the connector where it plugs into the motherboard. Readings WILL be different. Readings at the PWM do not take into consideration anything that can potentially cause ripple from where the wires are soldered down to the PCB forward to the component that's getting plugged in.

Personally, I would rather know the voltage and ripple of the components that's getting the power at the point it's getting it's power than from a PWM.

:toast:

[bullet2urbrain]

Wow, lots of critics here, that looks halfway decent to me, especially at full load.

but i guess we have a bunch of Xtreme PSU testers.


Thanks for the results JonnyGURU

[Anemone]

Well I'm in the mix having a GX 700 but not yet having it in use (and not sure I will either). If it weren't for the noise of the pcp&c I'd probably use one. The minute you try other alternatives to get rid of the noise, you end up with more issues it appears.

Ripple is rarely tested, so yes, kudos to those testing it at all. But to see these results on the GX is quite disheartening to say the least.

/sigh

So I'm back to the noise and perfect power or trying to find near perfect and quieter... or hoping that round 2 of the 700 sees some light at the end of the tunnel that isn't the 5:05pm train.

[pazza316]

I have a GX 700 by OCZ and my PSU runs flawlessly with my Conroe system and Crossfire graphics Cards (X1900 XTX) all heavily overclocked.

On the price versus performance scales the GX700 gets my vote every time. Yes I agree their are better PSU's but for the price the GX700 goes for, the unit is truly great and in my opinion their is no better!

[RyderOCZ]

Its been said before and I am sure it will be again.

You can't have everything for nothing, if you absolutely feel you need a PSU with 2mv of Ripple, spend the money, that is the only way to get it.

OCZ, Corsair...anybody can market a PSU that uses regulation as tight as you can want it.....but are any of you going to spend $650.00 when 95% of you will do just fine to spend $150.00 and never have a worry in the world when it comes to your rig?

EDIT: Sorry about that Jonny....very thorough review as always :) :thumbsup:

[Plywood99]

Its been said before and I am sure it will be again.

You can't have everything for nothing, if you absolutely feel you need a PSU with 2mv of Ripple, spend the money, that is the only way to get it.

OCZ, Corsair...anybody can market a PSU that uses regulation as tight as you can want it.....but are any of you going to spend $650.00 when 95% of you will do just fine to spend $150.00 and never have a worry in the world when it comes to your rig?

EDIT: Sorry about that Jonny....very thorough review as always :) :thumbsup:

Qft. Mountain out of a mole hill is the motto of some folks here...

[Sanborn]

thats less than 50mV fluctuation, stop your crying.

[jonnyGURU]

thats less than 50mV fluctuation, stop your crying.

Actually, the scale is .05v (50mv) per 2ms, so that's 100mV at the most, but still... like Ryder said... it's not even worth mentioning and to assume that another power supply would offer considerably less for the same amount of dough is naive.

[Praxis1452]

Actually, the scale is .05v (50mv) per 2ms, so that's 100mV at the most, but still... like Ryder said... it's not even worth mentioning and to assume that another power supply would offer considerably less for the same amount of dough is naive.
I'm sry but I found that really funny. Maybe I'm just tired... :D

[SoddemFX]

With almost all switching supplies the manufacturer has to choose a balance between steady state ripple and transient response, 100mv steady state ripple on a 12v rail isnt anything to worry about.

Sure a munufacturer can reduce output filter inductance and compensate with proportionly higher output capacitence but inductance is cheap whereas capacitence isnt. As Ryder said, its all about cost and a balance with performance is reached.

I think it looks like a good PSU :)

Tom

[nn_step]

Its been said before and I am sure it will be again.

You can't have everything for nothing, if you absolutely feel you need a PSU with 2mv of Ripple, spend the money, that is the only way to get it.

OCZ, Corsair...anybody can market a PSU that uses regulation as tight as you can want it.....but are any of you going to spend $650.00 when 95% of you will do just fine to spend $150.00 and never have a worry in the world when it comes to your rig?

EDIT: Sorry about that Jonny....very thorough review as always :) :thumbsup:
Hence I use a $600 PCP&C 850 SSI :D