Discussion PSU tier list discussion thread

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

Juular

Respectable
BANNED
Mar 14, 2020
1,061
258
1,940
these are almost peripheral issues to the reasons for why a unit doesn't make a certain tier or some units being where they shouldn't be at all.
Sound all issues you think are there, as detailed and structured as possible, don't just say 'there are issues'. Preferably also make suggestions on how you think they should be fixed.
We were going back and forth for three pages just to get to the poing where Rogue suggests smth that actually proves to be useful without any apparent drawbacks. And i don't want another 5 pages worth of discussion for every single issue. You see the issue ? Describe it and suggest a some way to fix it.
 

rubix_1011

Contributing Writer
Moderator
This is one reason why using the color grey is an issue - Dark Mode + dim backlighting (which I do) for eye strain:

8qQc2zv.png
 
  • Like
Reactions: King_V

Juular

Respectable
BANNED
Mar 14, 2020
1,061
258
1,940
So, is it a requirement that I have to change to light mode to view the list?
Blame default color palette. Well i wanted to revisit colors at some point anyway, it's not hard, just find-replace for RGB codes but i guess since people liked variant D and i did too (we still have to cast a vote in the team tho and folk at LTT actually seems to like current version more, go figure) we might as well just get rid of gray, and other colors aren't as problematic.
 

Juular

Respectable
BANNED
Mar 14, 2020
1,061
258
1,940
I've edited the poll only to include two options (variants A and D previously) as the most voted to determine what people want more. Also for some reason we have more than three times more votes on that other forum so maybe pin it at least temporarily so people would see it ?
 

Rogue Leader

It's a trap!
Moderator
I've edited the poll only to include two options (variants A and D previously) as the most voted to determine what people want more. Also for some reason we have more than three times more votes on that other forum so maybe pin it at least temporarily so people would see it ?

I was actually just about to reply to this post this morning to see what the resolution was on this. Hopefully we get some good responses with this update.
 

LukeSavenije

Prominent
Apr 4, 2019
140
25
620
I'm curious what the "various issues and concerns" are with the EVGA GT?
lack of info
I'm curious as to what lands a PSU on a given tier if it is "lacking info".
quoting the methodology:
"In tier A units should have detailed reviews on the this or other units using this platform in similiar configuration to be in normal priority subtier.
Detailed reviews are defined as having : 10-110% load, crossload voltage regulation tests, ripple measurement for all voltage regulation test points"
 
lack of info

Oh. Ok. Well, it's based off of HEC's TPK and there's a few reviews of that. If you need component analysis for 650W and 750W I have that.

But I noticed the Deepcool DQ-M in tier 1 without gray. So you've seen a review that makes you comfortable with the components used there?

I mean.. if we're going to put TX-M in tier B and put the EVGA GT and Deepcool DQ-M in Tier A, we might as well be honest with each other.
 

Juular

Respectable
BANNED
Mar 14, 2020
1,061
258
1,940
Oh. Ok. Well, it's based off of HEC's TPK and there's a few reviews of that. If you need component analysis for 650W and 750W I have that.
We'll gladly take that.
But I noticed the Deepcool DQ-M in tier 1 without gray. So you've seen a review that makes you comfortable with the components used there?
Yes, Aris did a review for Kitguru on 850W SKU and there are a whole bunch of reviews for lower wattages on other sites (all listed in the spreadsheet).
I mean.. if we're going to put TX-M in tier B and put the EVGA GT and Deepcool DQ-M in Tier A, we might as well be honest with each other.
We had to paint a line somewhere on ripple, TX-m just happened to be out of limits by 3mV, and it's not the only one. If we'll see that EVGA GT also has >50mV ripple it would be detiered too, that's why we require detailed reviews for tier A units to be white/normal priority (not to mention gilded), but do not for tier B and down (since it's not that important there) where just knowing the platform is enough. As in, we assume that units on the same platform perform comparably if they have similar components choice so as long as we know the breakdown for EVGA GT and have reviews for platform in general it would be a good ground to not gray out it. We considered requiring detailed reviews directly on the unit (or even multiple to account for differences between low and high wattage SKUs) for tier A but things got messy as we don't have enough reviews for nearly as much units we though we have, like a half of Seasonic lineup for example.
Edit: that said, Great Wall's E Gold seems to use the same platform as Corsair TX-m so probably we should detier it too.
 
Last edited:
I'm already looking into the FEDs with an unnamed reviewer... seems like they sadly used the same feds, so we're likely to detier unless changes occur

Tear down is already here: http://www.jonnyguru.com/forums/showthread.php?17496-DEEPCOOL-DQ-850-M-V2L-amp-CWT-GPX

Already blew up two of the 750W using the same test program I use to test RM and RMx (both are GPU). And yes... I mistook a PFC MOSFET for a PWM MOSFET. I don't admit to being the smartest guy in the room.
 

Juular

Respectable
BANNED
Mar 14, 2020
1,061
258
1,940
That's not the one I'm talking about. A multi-rail GPU is a good thing.
I'm talking about this one in the single +12V rail list:
Deepcool / Gamer Storm | DQ-M-V2L
But you were talking about regular DQ-M ...
But I noticed the Deepcool DQ-M in tier 1 without gray.
Either way, we have review on DQ-M-V2 from Fantastic, that's why it isn't grayed out and ripple is good enough to keep it in tier A and i have a feeling that it would actually clear ripple limits for gilding too ... Like, i get what are you saying, it doesn't look good components wise, but how formulate that into methodology ? Because sure, we can just quote you, but i'm sure it's not the only PSU out there that performs good in reviews but has components ratings barely enough for it's needs so that possibly could result in some issues, but we can't ask you about each single PSU to determine if they're good enough, can we ? And even in this particular case you don't seem to be overly confident on what CWT's components choice actually resulted in failures you observed either ...
Just not sure how I would want to share that info. Wouldn't want you guys to realize that the GT is on par with TX-M. :D
As far of components choice or performance ? Because again, TX-m has 53mV ripple, there are possibility that this were under performing sample but we work with what we have, if GT would endup having >50mV ripple it would go in tier B too, until then we assume that it would be comparable to Cougar GX-F until proven otherwise.
Edit: i went through Cyben reports on TX-m and there are two reports with ripple data listed with <50mV ripple in both, <30mV even in one so i guess TX-m is getting moved to tier A after all, what's up with Great Wall consistency tho ?
 
I think, and I'm not really sure how you're going to factor that in, if at all, aside from using the known quality of components, but it seems like ripple is your PRIMARY tiering factor and I think that skews things tremendously. I'm not going to point to any specific units, but we already know there have historically been quite a few models/series that had good or at least passable ripple etc. on initial reviews but degraded to much worse or even failure levels after a season or two. So poor longevity because the components were not high quality and can't hold up until extended or continous or frequent demand.

It makes no sense to put a unit in tier 1 because it has great ripple, if you know that because of the capacitors or other internal components, or lack of a sufficient cooling profile, that is isn't going to be getting that same result if it is retested in six months to a year after being hammered for a while. And we have seen a number of units like that in the past. I'm sure it's not just a historical precedent either. Again, not sure exactly how to fix that, but I think it should be a factor and when somebody like Jon is telling you that a unit contains crap, regardless that it tests well initially, I think that's something you have to at least figure out a way to incorporate to some degree or other.
 
Status
Not open for further replies.