Correct, but the rating power on the box most of the time refers to the actual power delivered by the rails (which can be found on the label), not the power drawn from the socket. As a general rule, power drawn from the socket greatly varies regardless of the 80+ rating (actual output value of the socket, quality of the wirings, caps, etc.). Very few power supplies actually deliver what is written on the box.
Comment has been collapsed.
This is the part that matters slightly more, but much less than who is making it and if that particular model is good:
bronze, silver, gold, platinum and titanium
So, your best bet is going to look at some of the tier lists and determining trusted OEMs then seeing how well their individual products are reviewed.
There are enough guides on it.
Comment has been collapsed.
Manufacturer means much more than imaginary watts or even the brand, so buy one made by Seasonic or some other top-tier you prefer. They are often labeled with less watts than they can actually deliver to be extra safe. Like you said, it's the most important part to not be cheap on.
Comment has been collapsed.
I've used Corsair for a number of years and without doubt, they have certainly delivered and i will look into them again for future purchases.
It seems to be a mixed bag in reality as not everyone will receive that level of satisfaction they want. I remember buying a Western Digital hard drive back in 2006. The first day, it went, full of bad sectors, took it back and opted for Seagate, and ever since then, no problems.
Only each person can make the right and wrong decisions and we risk it when we are basically experimenting with different parts and peripherals.
Comment has been collapsed.
Efficiency has nothing to do with power output, you're making a thread with intention to educate users while providing not only wrong but outright harmful information in the OP. 600W bronze and 600W gold both provide 600W power, the only difference is their efficiency, or in easier words, how much heat they waste during delivering that amount. You'd know that much if you at least bothered reading wikipedia.
The reason why you buy higher-watt PSUs than needed is not to match their efficiency, as you wrongly described, but to give some breathing room to the PSU, as the closer you are to its maximum output, the bigger wear and tear and a potential to have issues like power down at the minimal fluctuations of power output.
I'd suggest to educate yourself to the point of having an idea what you're talking about before trying to educate others. Later on people like me have to clean up all of that misinformation caused by posts like those.
How to choose the right PSU, then?
Compiling all points above will lead you to a very simple answer of what PSU should you get: "the one with around 20% more watts than I need, from good manufacturer, with 10 years warranty, at least 80+ bronze certificate". Then you can get a list of PSUs in your local prices and decide which one is the best based on both your local price for getting it, and its 80+ cert. Finally you can apply your personal preference and stick to that one.
What do you have then?
I have EVGA 600B 600W that I bought 1 year ago in august. I also had a lot of seasonics, corsairs, silentiumPC and a lot of other brands, it really doesn't matter, as long as it has a decent warranty and is already in production (and use) for a long time. Using any sort of "PSU tiers" and buying what somebody subjectively said is "the best" is horribly stupid. You won't get any meaningful data due to the amount of people that use various PSUs, and bad units happen everywhere. The best choice is sticking to something a bit older than latest release (i.e. not model that was released a week ago), that is already used by a lot of people without huge amount of complaints. Exactly the same what you should do with every other PC component, be it HDD, SSD, CPU, GPU or motherboard.
Comment has been collapsed.
Using any sort of "PSU tiers" and buying what somebody subjectively said is "the best" is horribly stupid.
They are useful for determining which manufacturers are good, or at least which product lines. You can always go and compare datasheets since as far as you know some might not be trustworthy.
Lists are usually pretty old, but can give an idea of the best brands.
Comment has been collapsed.
Having 100 failures compared to 50 isn't worse if it's 100 failures across a million of units, rather than 50 across 100k. A good manufacturer is one that has a long history of producing PSUs and passing all certifications, with PSUs being used by a lot of people in all sort of workloads.
Rankings are based on factors that are biased, irrelevant, impossible to verify or compared at a wrong scale, very likely all 4. I looked at example one and I can clearly say that first 5 tiers are of same quality to me, tier 6 is what you'd avoid because all of the above exist, and tier 7 you'd never buy, with last two being excluded already due to my points I stated above.
So yes, if somebody wants to see what manufacturers (not PSUs) are good, he can pick any from first 5 tiers. He'll also notice how it's almost the same manufacturers over and over again, which makes you think how it's possible for the same manufacturer to have tier 5 and tier 1 PSU at the same time (pro tip: perhaps because comparison of those PSUs is stupid, and tiering them is outright idiocracy?).
If somebody wants to argue how PSU X is better than PSU Y because one is in tier 2 and second in tier 4, then he's a fool. A person who puts PSU X above PSU Y is also a fool, because all usual arguments are flawed and biased, such as the most common one of "better 80+ cert" while at the same time entirely forgetting about "efficiency/price" ratio. Likewise with newer releases forgetting about potential of being problematic in long-run, and all other usual "reasons" that people use.
This is because objective ranking doesn't exist. Objectively you can't tell a difference in quality of two PSUs from first 5 tiers. You can only compare their technical differences and try to defend your choice based on them, but in practice you couldn't tell which PSU would blow up faster, and this is the main factor that matters. This is also why I say that any pick from first 5 tiers is good, and it absolutely doesn't matter which one precisely, because they're not better and worse, they're different.
Comment has been collapsed.
Having 100 failures compared to 50 isn't worse if it's 100 failures across a million of units, rather than 50 across 100k.
Sorry, how is that relevant to what I said? Are you referring to customer reviews seeming more negative to products with less sales?
Rankings are based on factors that are biased, irrelevant, impossible to verify or compared at a wrong scale, very likely all 4. I looked at example one and I can clearly say that first 5 tiers are of same quality to me, tier 6 is what you'd avoid because all of the above exist, and tier 7 you'd never buy, with last two being excluded already due to my points I stated above.
Still better to start with some slightly biased information than with no information; obviously, you can't take them as gospel.
Disqualifying some PSUs with issues and giving some credibility to good manufacturers, is how you should use them, particularly if you've never built a PC before and have nothing to go by otherwise.
This is because objective ranking doesn't exist.
Yeah.
Comment has been collapsed.
Still better to start with some slightly biased information than with no information; obviously, you can't take them as gospel.
And this is why I'd appreciate people linking good and trusted PSU manufacturers, rather than tiering single models and trying to make up ideology why one is above the other. I totally agree with you that this should be taken into account and every person who wants to avoid problems also should make use of such list. Sadly, a lot of people do not understand that, and instead will argue how PSU X is better than PSU Y, with me questioning their sanity, considering that both are from the same manufacturer. They'll happily reject PSU X because some fool put it in his subjective "tier 4", while picking overpriced and much worse for his setup "tier 2" just because it was put higher in irrelevant biased ranking. This is what annoys me, because people see PSUs as better and worse, rather than different depending on their usage, with manufacturers having a better or worse record of failures in the past, which is what should be used for making an educated choice.
Comment has been collapsed.
Might be worth mentioning that as long as they are well built and designed the differences shouldn't be too great. They are more and more standardized as they try to match the other's features, so the differences are often minimal.
As a hardware engineer, given enough information, one could probably pick out slightly better and worse designs among ones that would seem the same, but those small differences and production variations might effectively cancel each other out, so a "good" one is usually good enough.
Most should be fine since people are going to use them somewhat overdimensioned, as long they can actually regulate voltage and don't have much parasitic inductance. So "secondary" things like price, modularity, availability, and warranty can make a big difference.
Comment has been collapsed.
Do remember some lines of PSU's have a tendency to short out. A perfect example is the old CM line from Corsair. I work with building PC's and the general conscious and my own anecdotal evidence shows they die way too often.
But something like the CX line from Corsair is high tier and I have only had a few of those returned over the years. Whereas the CM line I've had over a thousand dead or dead on arrival returned to me.
Comment has been collapsed.
But something like the CX line from Corsair is high tier and I have only had a few of those returned over the years. Whereas the CM line I've had over a thousand dead or dead on arrival returned to me.
This is the effect of scale you're missing out. A few out of 100 would equal a thousand out of 100k. I agree that older lineups tend to be in general of worse quality and this is why I do not recommend grabbing PSUs that got EOL'ed 10 years ago, of course not. But at the same time I'd strongly doubt that "high" tier and "low" tier of some PSU from trusted manufacturer with similar release date would have that huge fluctuation in stability.
You should always aim for latest versions, even with discounts on older ones, but you should also refrain from trying out brand new ones that got released just now, until they're confirmed to work fine in long-run.
The effect of scale also works for your advantage, as getting a very popular PSU with a fraction of potential issues is a far better pick than something brand new with yet-to-be-seen stability.
Comment has been collapsed.
Definitely, agree on the new lineup wait for it to come proper reviews. The do not pre-order always come into effect regardless of what kind of product it is.
But saying 2 different line up coming out at the same time from the same producer and 1 can't be shit and the other is great is a false positive.
Just because it does not happen often does not mean it does not happen.
Thus the best you can do is always google reviews of the product you are planning to buy.
But the general rule of thumb:
Comment has been collapsed.
in practice you couldn't tell which PSU would blow up faster, and this is the main factor that matters
I don't want a PSU that blows up later, I want a PSU that will fail without blowing up at all 👀
600W bronze and 600W gold both provide 600W power,
Yeah, I don't know where the hell this misconception comes from. 600W is the output not the input, how hard is it to get? :s
Comment has been collapsed.
All will blow up sooner or later, because nothing works for indefinite time, the only question is how soon and whether it'll affect you. You're unlikely to use the same PSU for more than 20 years, and manufacturers know that too.
And of course I do not mean direct explosion but just getting broken :3.
Comment has been collapsed.
The 80+ ratings are pretty easy to understand, but I agree that finding a quality PSU is important. You have to check reviews on every new model, even if the company has a good reputation because every once in a while, even the best brand can release junk. I would recommend before anyone buys a PSU, to check a good indepth review first from hardocp or jonnyguru.
Comment has been collapsed.
You have to check reviews on every new model, even if the company has a good reputation because every once in a while, even the best brand can release junk.
That feeling when you see that someone gets it. Brand name doesn't matter as much as what goes into the power supply, since manufacturers often use a wide range of OEMs to fill their product line, and OEMs themselves use various quality components in their own OEM lines.
Edit: And yes, review sites that actually tear into the thing and give you a breakdown of internal component quality (like the two you've listed) are the only ones worth their salt.
Comment has been collapsed.
I swear sometimes you're a mind reader on these boards lol I was just about to post something like that myself. I used to be in electrical engineering and it's so frustrating trying to tell normal users that brand or the rating sticker doesn't always mean jack because manufacturers buying OEM parts that are already using OEM parts themselves and you don't really know the quality to expect until someone gives it a good tear down.
Comment has been collapsed.
There is no perfectly objective way to measure quality in a PSU, because it depends an awful lot on what you do with it. If you buy a 1000W+ platinum rated monster from the best of the best and stick it in a desktop with last gen components and a cheapo graphics card that you never use for anything more complex than rendering animated buttons on webpages then you have gotten a seemingly objectively better PSU that will perform worse than a seemingly objectively worse shitty middle tier 500W PSU... because a 1KW PSU will perform worse at 500W than a 500W PSU will.
So basically you can consider the 80+ marking a rough guide. Don't assume a cheapo PSU with the highest 80+ marking is legit, and don't assume all 80+ markings are worthless just because a few low grade products use them fraudulently. If it's "too good to be true" then when it turns out to not be true that's not an indication of the flaws in the system, but flaws in the purchaser.
Before you look at the 80+ rating though you need to determine your need. Once you have your need you know a certain range of PSU you should be in. Too low wattage would be very bad, but too high wattage is also a mistake. If you want room to upgrade later take it into account, but count narrowly because you will be losing efficiency until you do upgrade. That's money paid on electricity (for most of us) that you wouldn't have to pay, as well as money spent on a higher wattage PSU when you could have gotten a higher quality well calculated PSU for the same price or less.
Once you have your wattage range you can use the 80+ ratings as a rough guide to quality ranges. If you are planning to upgrade and so will be using less watts than you have then a higher rating is more important because it will save you money and work without complaints for longer. Use google to find the top brands recommended by serious review sites (I'm not going to share my favorites because everyone has their own thoughts on who is biased and who isn't. The good ones will have a post detailing their test methodology so you can nod sagely and pretend you know what they're talking about.)
Once you've got a couple of top brands and a range of ratings and wattages use a price comparison site to see what's available for a good price. Find the ones that match your criteria and are within your budget and search for reviews of them. Read multiple reviews of each PSU if possible. If you can't find any reviews then you should be concerned that the brand may be a problem. Search for reviews of other PSU's in the same series but with different wattages, it can give a hint and sometimes your PSU will be mentioned in the reviews.
After all this is said and done you should have no more than a couple PSUs on a short list. Buy the one you feel best about. Don't forget to make sure it comes with the cables you need and all that. Modular systems are absolutely awesome (but do note you can NEVER re-use modular cables from one PSU for another, they do not share a standard even within the same brand at times and WILL kill your PSU, maybe motherboard, maybe more. ONLY use cables that come with your PSU or sold specifically for it.)
Then after you bought it, keep the box and receipt and make sure you can test it fully before the return period is up unless you live in and shop from a country with good warranty protections for the consumer. Never throw away the box for any expensive things you buy until at least two years (or the entire warranty period, whichever is longer) have passed - being able to send it back for warranty repairs in the original box increases chances of a good repair/replacement experience dramatically.
Oh and if the PSU has a voltage chooser make sure it's set right, and then when you finally have put everything together and the thing doesn't start and your heart sinks and you worry you broke everything - make sure the power switch on the PSU (if it has one) is on. I swear I forget that fucking thing every damn time.
Comment has been collapsed.
The fun thing is reading who makes the things, there are many manufacturers who actually don't...manufacture...
One of the big names behind the scenes makes high quality PSUs for other companies, while their own range isn't always as good as those they make for others.
Comment has been collapsed.
Marking this for later, since we are about to build new PCs for the first time again.
Comment has been collapsed.
I think you got something heavily mixed up there buddy. That cert stands for the actual power consumption efficiency (electrical current not being wasted), where proven brands/ OEMs claim it's usually at the ~50-60% utilization; rather than the amount of wattage being delivered to your rig.
your psa is misguiding at best. . .
Comment has been collapsed.
13 Comments - Last post 11 minutes ago by Chris76de
73 Comments - Last post 29 minutes ago by romana1994
1,341 Comments - Last post 1 hour ago by devotee
15,412 Comments - Last post 1 hour ago by Aradiel
2,067 Comments - Last post 2 hours ago by Wok
103 Comments - Last post 3 hours ago by cartcollector
46,826 Comments - Last post 6 hours ago by FranckCastle
81 Comments - Last post 2 minutes ago by Chad78
3 Comments - Last post 10 minutes ago by Devirk
800 Comments - Last post 25 minutes ago by Mdk25
1,636 Comments - Last post 27 minutes ago by Mdk25
43 Comments - Last post 38 minutes ago by Exodust
65 Comments - Last post 58 minutes ago by Andrewski
112 Comments - Last post 1 hour ago by ThePonz
Since i've seen a rise of threads about builds recently and saw some common misconceptions about PSUs i've decided to write up a small guide that hopefully will help some users when it comes into picking an adequate PSU for their newest builds.
So, about 80+ certification, i've seen multiple times users suggesting this or that PSU based on their certification alone, because if the manufacturer has that sticker on the box it must be at least of better quality than those who don't, right? Well, the actual reality is different. But what is this certification exactly?
The 80+ rating is a supposed trusted way of checking the efficiency of a PSU, rated on different levels based on standard, bronze, silver, gold, platinum and titanium. That is, the amount of total output power that the PSU can supply on its rails. For example, a bronze rated PSU can, on paper, output 81% to 84-85% under different load values of the total rated wattage on the box. Let's say we have a 500W PSU rated bronze, at best it will actually output around 420W. No PSU can be 100% efficient because of thermodynamics; this is the reason why, despite a tipical gaming build needing 350W under full load at most, you still need to leave a bit of a room when considering a power supply.
More specific infos can be found here: https://www.tomshardware.com/news/what-80-plus-levels-mean,36721.html
This measurement isn't really accurate because testing a power supply in a lab is fundamentally different from normal use. Besides, some manufacturer can send specifically-made lucky units just to get the certification and then produce worse (and therefore cheaper) models for sale. Vultech PSUs are a prime example of silver rated power supplies that almost never reach their rated efficiency. This also means that there are some non-rated PSUs that actually are quite efficient and do not have the certification simply because they didn't ask for a rating. And yes, the rating is not mandatory nor is reflective of the actual build quality of your unit. Some gold rated PSUs can employ crappy chinese components that will likey fail within a couple of years, and you wouldn't want to waste your precious and expensive 1080ti, right?
Now, don't get me wrong here. As a general rule of thumb higher rated PSUs are actually built with good components simply because it's difficult to reach higher efficency rates without putting some efforts in build quality and reliable parts. I'm just saying that it's not an objective way of measuring the quality of the PSU and that you shouldn't put your faith in it, especially when powering high grade parts. After all, you're spending 1000+ €/$ on a PC, why cheaping up when it comes into the power supply? That 10 extra dollars can make a lot of difference between a flawless PSU that will work for years and a bad one that will die along the rest of the computer.
How to choose the right PSU, then? This is another story that i will elaborate if there's enough demand in this thread.
Comment has been collapsed.