Discussion PSU recommendations and power supply discussion thread - Tom's hardware

[Karadjgne]

Technology will Always leap ahead of safety. Safety doesn't even become a twinkling in someone's brain patterns until a need is discovered. It's not until the technology is viable that ppl like Aris and JG and OW start picking it apart and finding its shortcomings. 20 years ago nobody cared about transient response, holdup times, and half a dozen other things that seem to be becoming more important. 6 years ago, nobody really cared about inline caps, JG tested with and without and basically said they help but didn't know why they weren't on the inside of the psu. Now? Ppl griping about sleeves, lack of real need etc.

Hindsight is 20/20, easy to see needs for improvement or fixes or precautions, but if the tech doesn't yet exist, can't really predict any of those. So safety by default can only come after and only if there's a need for it.

 

[Karadjgne]

Riotoro has been around for several years, they made mostly cases and peripherals and are quite well known in South America markets. It's only recently they really branched out into psus, and teaming up with Seasonic is always a solid bonus. Whether they'll stick with that or try and flood the market by using cheaper platforms, I dunno. It'd be nice if they didn't, I like having something solid I can recommend, without having to play guessing games.

 

[Mezoxin]

Hindsight is 20/20, easy to see needs for improvement or fixes or precautions, but if the tech doesn't yet exist, can't really predict any of those. So safety by default can only come after and only if there's a need for it.

yes the tech has to be present in order to discover its shortcomings , but it doesn't have to be released to discover ALL of it , as there is a lot of predictive analytical tools available for that
Back in the dark ages the only way of predicting the future was by words of oracles and superstition , but ever since Pascal and Fermat founded the basics for which would be later known as probability theory , there has been a scientific method of predicting the future , and although they were made to solve a puzzle and win games of chance, later on their first business application was with life insurance policies in the uk 18th century and has been the founding stone of what is known as risk management , whether in finance or reliability engineering or safety and also for conducting valid researches and linking a cause to an effect with enough certainty
So not everything we see in safe design or protection features are reactive most of them are predictive as well

After i read some pretty good reviews and saw some good deals i reccomended a Riotoro Enigma 850w gold for somewhere around 80 or 85 bucks.

I think i would genuinely concider a Riotoro PSU for a new build assuming it is a well reviewed model.

yes that older unit was reviewed by aris , i would feel confident as well about the newer version if it has a review of similar caliber

 

[NightHawkRMX]

yes that older unit was reviewed by aris , i would feel confident as well about the newer version if it has a review of similar caliber

The newer Enigma G2 Is basically a SeaSonic focus plus gold and carries over the same 10 year warranty. This unit is actually superior to the Antec Earthwatts gold which are based on the Focus gold (not plus) and have less modularity and have a shorter warranty length.

Here are reviews varying in detail.

Riotoro Enigma 850W G2 Power Supply Review - KitGuru

Today we take a look at the new Riotoro Enigma 850W G2 power supply which is 80 Plus Gold Certified,

www.kitguru.net

Riotoro Enigma G2 650W Gold | Power Supply Review

The RIOTORO Engima G2 650W Gold is a very good power supply when it comes to performance and sound, thanks to the new Seasonic platform it uses

tech-legend.com

I feel the Enigma G2 should find there way onto this recommended list since its nearly identical SeaSonic twin is recommended.

 

[NightHawkRMX]

Here is a review for the OG Enigma gold 850w. It is clearly a fair bit different from the newer G2 as it is based on a completely different Great Wall platform.

It also appears the Enigma 850w is a fair but inferior to the newer Enigma G2 models, however it seems pretty alright. I don't know whether it would make this list or not.

Apparently the Enigma Gold 850w is "virtually identical to the Corsair CS850M..."

Riotoro Enigma 850W Power Supply Review - PC Perspective

Riotoro Enigma 850W Power Supply Review Introduction

pcper.com

https://www.tomshardware.com/reviews/riotoro-enigma-850w-psu,4910.html

 

[Mezoxin]

@NightHawkRMX
the older version atleast had a thorough review , so you would know what you are exactly getting
the G2 is better than the older version that's true as its based on a superior platform, but at this price point i would wait and see a proper thorough review for that unit before recommending over something that is similarly priced and thoroughly tested like the corsair TXM , as cheaper rebrands may have some components stripped off and the missing mention pf protective features in the product page makes a little bit suspicious

 

[Karadjgne]

Like all RIOTORO PSUs, the Enigma G2 features full voltage, current, power, and short-circuit protections for improved safety.

Riotoro spec sheet Enigma G2.

I’m glad to see overtemp protection being mentioned, though

Riotoro Builder 500W Power Supply – JonnyGURU.com

www.jonnyguru.com

And any protections are mentioned where in that professional review? On the box itself, but not even totalled by OW.

The older Onyx psus web pages have at least a mention of 'full voltage, current, power and short circuit' protections.

 

[Mezoxin]

@karadigne
was able to find the spec sheet but on a different part of the website that strangely links to a file uploaded on amazon
https://www.riotoro.com/support/enigma-g2-750w-atx-power-supply-1
https://s3.amazonaws.com/riotoro-live/downloads/8e1e08f3-6497-42bb-a826-fe715a4c96cf.pdf

so according to their spec sheet which is very rough when it comes to mention of protections they may have stripped off OTP
yes OW did not update his methodology for quite a long time , Aris continually does though , last thing i think he added was in 2018 was EMI testing

 

[Vic 40]

Think Aris would have mentioned it if certain protections wouldn't be there,
Riotoro Debuts Enigma G2 Power Supplies | Tom's Hardware

 

[Mezoxin]

Think Aris would have mentioned it if certain protections wouldn't be there,
Riotoro Debuts Enigma G2 Power Supplies | Tom's Hardware

that is just a n early news report and not a review , he is just saying what has been told to him by representative at CES 2018

 

[Vic 40]

You'd think he does at least some research ...

 

[Mezoxin]

You'd think he does at least some research ...

of course he does , but he would never evaluate a unit based on paper or word of mouth , that's why he says he expects it to perform similar to seasonic , the news article was only meant to be news not a recommendation or an evaluation, nor is he supposed to critique a product based on speculation

 

[Vic 40]

Not talking about what performance to expect,just discussing protections and those not being mentioned and him doing some research to what protections might have been included. Think he would at least look for that kind of info and mention discrepancies.

Do not know how he goes about it,just me thinking out loud.😉

 

[NightHawkRMX]

I was trying to get a glimpse at the protections IC from the pictures included in the KitGuru review, but I really don't know what to look for nor are the pictures detailed enough to make out writing on chips when the image is magnified.

I would HIGHLY doubt Riotoro would spring for a Seasonic focus platform, use mostly the same components inside as its seasonic counter part, then offer crappy protections.

The PSU at least has OCP from the testing on kitguru. Infact the 850w g2 delivered 900w while maintaining 90% efficency (ill chalk this up to LLC resonant working wonders) and cut off at 912w.

The next PSU i get WILL have llc resonant for the stable efficiency no matter the load.

 

[Mezoxin]

The PSU at least has OCP from the testing on kitguru. Infact the 850w g2 delivered 900w while maintaining 90% efficency (ill chalk this up to LLC resonant working wonders) and cut off at 912w.

well that was OPP not OCP , as seasonic focus plus as well as most single 12v psu's dont need and normally dont have OCP on the 12v rail , they only have it on the minor rails
https://www.tomshardware.com/reviews/seasonic-focus-plus-gold-850-psu,5247-6.html

either the reviewer took a wrong measurement when he tested when it would cut off or that is the first evidence that there is a difference in protection implementation between the g2 and focus plus
because the focus plus OPP set point was at 1160w / 137% of nominal wattage , while the g2 cut off at 912w which is 107%
but my guess is that the kitguru reviewer took a wrong measurements because OPP is never set that low

 

[NightHawkRMX]

well that was OPP not OCP , as seasonic focus plus as well as most single 12v psu's dont need and normally dont have OCP on the 12v rail , they only have it on the minor rails
https://www.tomshardware.com/reviews/seasonic-focus-plus-gold-850-psu,5247-6.html

either the reviewer took a wrong measurement when he tested when it would cut off or that is the first evidence that there is a difference in protection implementation between the g2 and focus plus
because the focus plus OPP set point was at 1160w / 137% of nominal wattage , while the g2 cut off at 912w which is 107%
but my guess is that the kitguru reviewer took a wrong measurements because OPP is never set that low

That does seem pretty low at 107%, but there are some capacitor differences between a Seasonic and the G2, so this may have an affect.

 

[Mezoxin]

That does seem pretty low at 107%, but there are some capacitor differences between a Seasonic and the G2, so this may have an affect.

almost all PSU's now have their OPP at 125% or more , some reaching 140%
912w that would cause problems with something like a vega 64 when combined with a powerful cpu , check Rouge leader experience with seasonic prime 650w ( had OPP at around 900w) and the vega 64 , it was a couple of posts back in this thread

 

[Karadjgne]

Huh, backwards maybe? I don't think it's psus causing problems with Amd gpus, as much that in typical and historic fashion amd gpus and power spikes are causing problems for psus.

 

[Rogue Leader]

Huh, backwards maybe? I don't think it's psus causing problems with Amd gpus, as much that in typical and historic fashion amd gpus and power spikes are causing problems for psus.

The PSU wasn't causing my issue, I agree. a 650w unit SHOULD have been enough, but the Vega 64 had power spikes in certain games that would hit OPP and shut the system down. Replacing it with a 750 fixed it. Both Seasonic Prime Titanium.

My 5700XT outperforms it and doesn't have such issues.

 

[Mezoxin]

If a PSU manufacturer releases a 650w PSU with 4 (6+2)pin Pcie connectors , this means that by PCI-SIG requirements that PSU can supply up to 600w of power through the 12v pcie cables + 75w for the pcie slot by the 24pin , why would PSU manufacturers allow for up to 112% of nominal power to be used for gpu's although that wouldn't leave enough power for the rest of the system , it would make more sense to top it off at 3 connectors max

Huh, backwards maybe? I don't think it's psus causing problems with Amd gpus, as much that in typical and historic fashion amd gpus and power spikes are causing problems for psus.

currently not only AMD though , problematic power spikes seems to be common now for top of the line GPU's some versions of 2080Ti's experienced similar behaviors with some 650w psu's

 

[Karadjgne]

In all reality a 6pin can provide upto 120w, an 8pin can provide upto 180w, and pcie slot somewhere around 90w. And for the daisy chained 6/6+2 it's somewhat less, limited more by guage than pins. Kinda throws your math off.

 

[Mezoxin]

In all reality a 6pin can provide upto 120w, an 8pin can provide upto 180w, and pcie slot somewhere around 90w. And for the daisy chained 6/6+2 it's somewhat less, limited more by guage than pins. Kinda throws your math off.

you are absolutely correct that in real use it could allow for more or less than what PCI-SIG requires (6pin=75w , 6+2pin=150w) according to variables you mentioned , i slightly disagree with the difference you mentioned between 6 and 8 pin ability in real world
but my point was is that by the availability of these extra connectors they allow for user error in overloading a PSU ,
which defies the purpose of why we have the cosmetic difference between 6pin pcie and 8pin pcie as the only reason why we have that cosmetic difference was to prevent user error in underpowering a GPU , as both connectors have the same number of 12v lines and are physically able to supply the same load.

 

[Darkbreeze]

It's not just about the cables though. The CIRCUIT that the cable is plugged into or part of also needs to be capable of supporting the desired or required load. For a 6 pin only ciruit, the internal circuit is probably NOT going to be designed to handle the same load that it would be if it were a 6+2 or 8 pin cable. Which is part of the reason why it is a bad idea to use a 6 pin to 8 pin adapter. It doesn't matter that both the 6 and 8 pin connectors have the same number of 12v wires or that they are both the same gauge, IF the internal circuit was not designed to withstand the same kind of load as one equipped with an 8 pin would have been, and it WON'T have been otherwise it WOULD have been given a 6+2 or 8 pin cable to acknowledge that fact.

 

[Mezoxin]

It's not just about the cables though. The CIRCUIT that the cable is plugged into or part of also needs to be capable of supporting the desired or required load. For a 6 pin only ciruit, the internal circuit is probably NOT going to be designed to handle the same load that it would be if it were a 6+2 or 8 pin cable. Which is part of the reason why it is a bad idea to use a 6 pin to 8 pin adapter. It doesn't matter that both the 6 and 8 pin connectors have the same number of 12v wires or that they are both the same gauge, IF the internal circuit was not designed to withstand the same kind of load as one equipped with an 8 pin would have been, and it WON'T have been otherwise it WOULD have been given a 6+2 or 8 pin cable to acknowledge that fact.

that could be very true for the circuitry inside the gpu thanks for bringing this to my realisation , but still from the psu side all 18AWG to pins need to adhere to intel ATX v 2.2 standard of supporting from 6-8A per pin of 12v rails

and thats why you see the pcie cable that has 2x (6+2) 8 pin is actually sourced from only 3 12v pins from the psu side in something like the corsair type 4 cable

so mostly by intel atx standard the PSU will support higher loads per 18 awg pinned wires than what is required by PCI-SIG for PCIe connectors , taking that into consideration , It supports that the difference between 6+2 and 6pin from the (PSU side at least) , was only for cosmetic reasons and to prevent user error of using a power supply that will underpower a gpu .

 

[Karadjgne]

6pin has 2x 12v hot, 3x ground and a sense. 8pin has 3x 12v hot, 3x ground and 2x sense. But the pins themselves are rated at 5A, so that's 60w per pin max, 120w on 6pin and 180w on 8pin. On psu side, pins can be rated for 9-11A, same as the pins on the motherboard side of EPS.

Since they all go back (especially true for daisy chained 6/6+2) to the same source, even on most multiple rail psus, whats available to each pcie will be what's available to all pcie in total. So technically you could try to pull all 40A+ through a single 6pin, which is where SCP, OCP, OPP should be kicking in to prevent such from happening.

It's with cards like the R9 295x2, a 2x 8pin gpu, that could easily pull in excess of 450w at default settings, where the standard 150w per 8pin gets tossed out.

Daisy chained pcie change the game. An 18" 18ga is rated for @ 8A per wire, or @ 284w per chain. Leaving 2x 8pin on a daisy somewhat chancy with cards that can pull that much through pcie which is supposedly rated at 2x 150w.

What a psu can supply at the point of exit, what a component can draw and what the actual connectors and wiring can handle are not very often, if almost never, the same thing. Having a pcie outlet capable of 24A on 6pin doesn't mean the cable can handle that amount of power, nor the connectors on the other end. Psu manufacturers over-enable modular plugs to prevent them from being the weak link, if anything is to burn up, it's the wire or the component end plug, not the psu side.