[EOL] PSU Tier List rev. 14.8

[SexualHealing]

2 hours ago, LukeSavenije said:

About the video @SexualHealing sent earlier

 

a huge shoutout to Poiu from Tweakpc for checking this and talking through it

 

I don`t understand the gear used but even before they measured temps the slim guy actually setting it looked at the oscilloscope and realized the ripple was likely caused by temps because of the wave pattern, then he measured temps and it was super high. They tuned the psu off on purpose to give it another chance.

[LukeSavenije]

1 minute ago, SexualHealing said:

I don`t understand the gear used but even before they measured temps the slim guy actually setting it looked at the oscilloscope and realized the ripple was likely caused by temps due to the of the wave pattern, then he measured temps and it was super high. 

and we came to the conclusion it's a measurement error, either by actual wrong measure or by added noise

 

what i wanted to show was this

 

sunmoon:

 

Chroma

 

video

 

those spikes there aren't normal

[TVwazhere]

5 hours ago, SexualHealing said:

I particlarly do not trust you when it comes to Corsair, you are the marketing director of Corsair. Your job is to promote Corsair at any cost. Aris owns a company with you.

Maybe when he's acting as a representative for Corsair, but right now he's just a guy on a forum. If he was here representing Corsair, he'd have an IA badge and probably would have a user name reflecting that. We have other Corsair reps here already who do just that.

[LukeSavenije]

4 minutes ago, TVwazhere said:

Maybe when he's acting as a representative for Corsair, but right now he's just a guy on a forum. If he was here representing Corsair, he'd have an IA badge and probably would have a user name reflecting that. We have other Corsair reps here already who do just that.

he even said it himself a couple times

 

[SexualHealing]

5 minutes ago, TVwazhere said:

Maybe when he's acting as a representative for Corsair, but right now he's just a guy on a forum. If he was here representing Corsair, he'd have an IA badge and probably would have a user name reflecting that. We have other Corsair reps here already who do just that.

 

Nothing particular about him, it is just me. 

 

Regarding a possible error in the tests, i can`t have any word on it but luckly he will ge to know and give his version about it.

 

 

[Juular]

1 hour ago, SexualHealing said:

even before they measured temps the slim guy actually setting it looked at the oscilloscope and realized the ripple was likely caused by temps because of the wave pattern

That what he says ?

[SexualHealing]

1 hour ago, Juular said:

That what he says ?

 

Before the second temp check while pointing to the wave abnormality the "slim guy" says "it is probably the primary suffering with heat".

 

I would get surprised if he did anything wrong, also by my surprise he ended recommending the PSU in that same video but only up to 550W.

 

[Guest]

9 hours ago, SexualHealing said:

 

How is it bullshit?

 

 

 

Because he's overloading the PSU.

 

He's expecting optimal performance while the PSU is out of spec.

 

Who in the real world is going to use a 550W PSU for a 600W load?  You wouldn't.  So you would never see these results.

 

I'm not saying his RESULTS are bullshit.  I'm saying that his expectations for these outlier tests are unrealistic and should not be a focal point.

 

It's like a car review driving a car stuck in 1st gear, keeping it at redline for half an hour and then saying the car is shit because the engine blows up.

[LukeSavenije]

1 minute ago, jonnyGURU said:

I'm not saying his RESULTS are bullshit.

can you look at what Poiu said about it? i feel like the results might actually be a bit odd

[Guest]

9 hours ago, SexualHealing said:

 

jonnyguru and the reviewer with a Corsair t-shiirt. In the video jonny claims to be the marketing technical manager of Corsair.

 

Pretty sure everyone here already knows that, so I don't get your point.

 

To show that I'm not being biased, I pointed out that even a high end Seasonic Prime PSU does the exact same thing when you overload it.  

[SexualHealing]

3 hours ago, jonnyGURU said:

Pretty sure everyone here already knows that, so I don't get your point.

 

To show that I'm not being biased, I pointed out that even a high end Seasonic Prime PSU does the exact same thing when you overload it.  

 

Even if there is an error causing high ripple and temperature it shows the PSU has no working protections anyway 

[SexualHealing]

Teclab is live on Youtube right now, he said the CX550 result was atypical and that he studied it. The only conclusion he could reach is that it was an issue with the PSU ressonance, there is no operational mistake from his part. 

 

 

 

 

[Guest]

58 minutes ago, SexualHealing said:

 

Even if there is an error causing high ripple and temperature it shows the PSU has no working protections anyway 

No.  It doesn't.

 

There is no protection for "too high ripple" unless you have a digital PSU.  Which you don't.

 

It's not uncommon for a PSU under loads beyond it's rated specification for either maximum output or temperature to have an increase in ripple.

 

Therefore, the only solution for an analog PSU to avoid this issue is to reduce the OCP, OPP or OTP... which is not practical.

 

THAT is why I say the test is bullshit.

 

[SexualHealing]

34 minutes ago, jonnyGURU said:

No.  It doesn't.

 

There is no protection for "too high ripple" unless you have a digital PSU.  Which you don't.

 

It's not uncommon for a PSU under loads beyond it's rated specification for either maximum output or temperature to have an increase in ripple.

 

Therefore, the only solution for an analog PSU to avoid this issue is to reduce the OCP, OPP or OTP... which is not practical.

 

THAT is why I say the test is bullshit.

 

 

So the ripple supression and temperature are the issue, not the lack of protection..He is testing it again now. 

But people asked to test if it does explode.

 

Edit: The PSU showed the same bad beahaviour, but if i got it right that ripple interference and high temperature went away after changing the voltage to 230 volts. 

[Guest]

38 minutes ago, SexualHealing said:

 

So the ripple supression and temperature are the issue, not the lack of protection..He is testing it again now. 

But people asked to test if it does explode.

 

Edit: The PSU showed the same bad beahaviour, but if i got it right that ripple interference and high temperature went away after changing the voltage to 230 volts. 

Because at 230V input, the primary side is working at half the current and therefore, lower temperatures and farther from the OPP threshold.

 

[Firewrath9]

Also, @SexualHealing, I still don't really see the issue with the CX550. Lets suppose it sucks when it is a 600-700W. Why does that matter? ofc it's going to suck and have bad ripple. There's a reason why its 550W. As for OCP, if you're running 600W-700W sustained on a 550W PSU, then don't blame the PSU. It's whoever decided to stick a 2080 Ti and a 9900KS and overclock the hell outta both then run furmark + AIDA64 FPU's fault. Short spikes up to 600W won't effect the performance (if the OCP was lower than spikes would trigger it [at least thats how I think it would work], so there's a reason why OCP is high), and shorts would instantly trip OCP, so no problem there.

Saying the CX550 is bad for performing badly at 600W is like saying a 3700X is bad because it degrades at 1.6v. A 3700x is not designed for 1.6v, just like how the CX550 is not designed for 600W. The CX550 is designed for 550W. People who buy the CX550 won't go over 550W. 

[SexualHealing]

8 hours ago, Firewrath9 said:

Also, @SexualHealing, I still don't really see the issue with the CX550. Lets suppose it sucks when it is a 600-700W. Why does that matter? ofc it's going to suck and have bad ripple. There's a reason why its 550W. As for OCP, if you're running 600W-700W sustained on a 550W PSU, then don't blame the PSU. It's whoever decided to stick a 2080 Ti and a 9900KS and overclock the hell outta both then run furmark + AIDA64 FPU's fault. Short spikes up to 600W won't effect the performance (if the OCP was lower than spikes would trigger it [at least thats how I think it would work], so there's a reason why OCP is high), and shorts would instantly trip OCP, so no problem there.

Saying the CX550 is bad for performing badly at 600W is like saying a 3700X is bad because it degrades at 1.6v. A 3700x is not designed for 1.6v, just like how the CX550 is not designed for 600W. The CX550 is designed for 550W. People who buy the CX550 won't go over 550W. 

 

Firewrath9.

It matters for your computer safety, you may have not dimensioned psu power well. Overclock may even double power comsumption, Also as you states a GPU particularly AMD may also have unspecified power draw spikes. It is not a problem to me either but i know what i am doing when building a computer, most customers and even vendors do not know. LukeSavenije totally agrees it matters, Seasonic Focus Plus v2 was downgraded in his Tier list exactly because at 120% load it would have high ripple according to a single Chinese web page. I even had a discussion with him about the Seasonic and got upset but whatever he believes that PSU has high ripple at 120% and won`t put it back in a high Tier. 

 

Ripple kills computer components, during spikes to more sensitive parts like your HDD or SSD ripple may burn those parts out of the blue and you will lose all your data. It is very sad. And do you think manufacturers care? They don`t, Corsair made a reputation 10 years ago or more when they were actually putting out quality in reasonably priced products like the CX but the CX had cost cuts to increase profits year and after year. Today something like the CX M is no different of a $10 generic psu. The 60mv of the CX at normal loads is completely fine but when you compare to 10-20-30mv offered by Super Flower and Seasonic in smilarly priced psus it gets bad! The CX high ripple shown in 5v and 3v rails at only 110% load may kill any hdd and ssd.

 

Among well informed enthusiasts Corsair is known as the brand the offers less for more cash. The overpriced Corsair PSUs made by Seasonic, Flextronics are fine for the most part but CX, VS, CX M and probably the new CV are lemons with such issues and durability problems, but they have warranty! No, they do not. Their warranty is complete bullshit, they will have a burden of bureaucracy and if you pass over the bureaucracy they will try hard to find something to deny warranty, if you miss a cable warranty is denied. Someone close to me had his Corsair AXi warranty denied simply because Corsair would say it has no issues, it had a proven issue. In the end Corsair offered him a cheaper HX psu, he accepted because he was already tired trying to get his AXi replaced. And what if your computer was damaged over time due to a low quality psu? They won`t buy you a new computer as you may guess.

 

The CX high temperature is also an issue, the psu won`t last and may even get on fire. When they tested if the CX would explode in the last video, they were kind enough to change voltage to 230v. At 115v it would explode. 

 

If you do not patrol companies they will have all sorts of questionable practices. Ronaldo from the videos i posted have found very interesting cases, like a psu from Xigmatek with 400mv ripple at normal loads but the interesting thing is that it had a fake component made of concrete to make the psu internals look better and heavier. He also caught companies doing different batches of the same psu model, some batches that get reviewed would be of good quality but other batches found in retail stores would be of lesser quality. 

 

 

 

 

 

[komi1997]

FSP looks to have a new budget DC-DC PSU, but their naming is awful as always, Hydro Pro, come on, there are now 6-7 Hydro series by FSP by now...

[LukeSavenije]

8 minutes ago, komi1997 said:

naming is awful as always, Hydro Pro

worst part... this is also the second hydro pro

https://www.fsplifestyle.com/PROP197003683/

[komi1997]

Just now, LukeSavenije said:

worst part... this is also the second hydro pro

https://www.fsplifestyle.com/PROP197003683/

seems there is also Dagger Pro, 5 new Daggers incoming now I suppose

[LukeSavenije]

6 minutes ago, komi1997 said:

seems there is also Dagger Pro, 5 new Daggers incoming now I suppose

dagger pro was already a thing, actually

 

welcome to FSP

[Vejnemojnen]

33 minutes ago, komi1997 said:

FSP looks to have a new budget DC-DC PSU, but their naming is awful as always, Hydro Pro, come on, there are now 6-7 Hydro series by FSP by now...

at least it has black cables, DC-DC converter and Bronze Efficiency. Too bad that Bitfenix units have gotten more expensive. Back in august, you could have bought a 450w formula gold for ~19.000ft. Which is a much better-nicer unit IMHO.

[SexualHealing]

17 hours ago, jonnyGURU said:

 

Who in the real world is going to use a 550W PSU for a 600W load?  You wouldn't.  So you would never see these results.

 

I would not but you know that is not what happens in practice. You also know many PSUs turn itself off before putting the PSU or computer at risk, that is the whole point of protections. If people can only use it safely up to 550W the PSU should turn itself off at 550W,.

[Guest]

2 hours ago, SexualHealing said:

I would not but you know that is not what happens in practice. You also know many PSUs turn itself off before putting the PSU or computer at risk, that is the whole point of protections. If people can only use it safely up to 550W the PSU should turn itself off at 550W,.

The point of protections is to protect the PC and the user from unexpected failures, shorts, etc.

 

Not for people deliberately pulling more power than they're supposed to from a PSU.

 

 

[Guest]

6 hours ago, SexualHealing said:

 

 

 

The CX high temperature is also an issue, the psu won`t last and may even get on fire.

 

 

Also not true.