99% of Power Supply Reviews Are Wrong
- Thread starter ennogs
- Start date
That's a particularly ridiculous statistic. It's not all that difficult to find a technical review of a power supply if you want one. Most people don't, they're looking for the words "quiet" and "modular". There's a considerable difference between wrong and irrelevant, very few of the reviews their article is condemning are wrong for the simple reason that they don't say anything testable.
If you buy from one of the high end brands they actually quote specifications and adhere to them. 7 year warranties and rating their supplies at 50 degrees ambient is a good sign as well. 6 months warranty and not actually specifying operating temperature is a bad sign. If you buy based on the cheapest per W you deserve everything that's coming to you.
edit:
I'm not completely convinced the author knew what he was talking about either. That or he is deliberately misleading. Indeed a multimeter power won't tell you a great deal, but it's fine for checking if the rails are within the 5% range the atx spec requires. It'll also show the rail drooping with load, and will demonstrate cross loading with no problems either.
Motherboards are designed to run on dirty power; it'll never be a perfect 12V when there's several things drawing various amounts of current from it. It has to assume a psu which is barely atx compliant to avoid huge numbers of returns. Obsessing over ripple is largely a waste of time as a result, it may be worth a couple of mhz, but then again it might not be.
"99%" of reviews are like that because it's all people want to know, and probably all they need to know. I don't think insisting every review include a tear down of the supply and discussion of the component choices would lead to fewer people buying 1200W supplies for £30, it'll just alienate people further.
If you buy from one of the high end brands they actually quote specifications and adhere to them. 7 year warranties and rating their supplies at 50 degrees ambient is a good sign as well. 6 months warranty and not actually specifying operating temperature is a bad sign. If you buy based on the cheapest per W you deserve everything that's coming to you.
edit:
I'm not completely convinced the author knew what he was talking about either. That or he is deliberately misleading. Indeed a multimeter power won't tell you a great deal, but it's fine for checking if the rails are within the 5% range the atx spec requires. It'll also show the rail drooping with load, and will demonstrate cross loading with no problems either.
Motherboards are designed to run on dirty power; it'll never be a perfect 12V when there's several things drawing various amounts of current from it. It has to assume a psu which is barely atx compliant to avoid huge numbers of returns. Obsessing over ripple is largely a waste of time as a result, it may be worth a couple of mhz, but then again it might not be.
"99%" of reviews are like that because it's all people want to know, and probably all they need to know. I don't think insisting every review include a tear down of the supply and discussion of the component choices would lead to fewer people buying 1200W supplies for £30, it'll just alienate people further.
Last edited:
well, none of the examples were from reputable sources... I expect that it is fair to say, 99% of COMPUTER reviews are useless, cos they're written by the 99% of the population who don't know the difference between a Jeantech and a PCP&C, or an Athlon and a toaster...
That's why its 99%
Didn't even read the article sounds like bs to me, as said if your looking for reviews you will find a good source for you personally, if you don't you have yourself to blame end of..
edit: Just read the article, is this news of course people want to play with hardware if they can get it for free, didn't like the author, has a point though..
Last edited:
News flash: there are lots of rubbish review sites run by people in their bedrooms You can generally tell which ones they are because the review is cursory, full of grammatical errors and accompanied by photos which look suspiciously like they were taken either on an iPhone, or on an über SLR camera in the hope that pretty pictures will provide a distraction from the awful prose.
You can generally tell which ones they are because the review is cursory, full of grammatical errors and accompanied by photos which look suspiciously like they were taken either on an iPhone, or on an über SLR camera in the hope that pretty pictures will provide a distraction from the awful prose.
Last edited:
Associate
- Joined
- 27 Jan 2005
- Posts
- 402
- Location
- Worcestershire
TBF that article is years old, I am pretty sure it predates JonnyGuru's site, it's just been updated a few times since. I've seen it pop up a few times since people started paying attention to PSUs and actual load output ratings.
It really needs a title change now, but in essence it is kinda right. The sites that are better reviewers of PSUs are not the most common visited unless you use somewhere like here or XS maybe, something like that. CustomPC started using scientific reviews (reposted on B-T), and the aforementioned JG, but it's not uncommon for me to visit a site and see morons giving their rail readings for 3 lines with a LOADED CPU and GPU and 4HDDs attached (as if it matters...)
It really needs a title change now, but in essence it is kinda right. The sites that are better reviewers of PSUs are not the most common visited unless you use somewhere like here or XS maybe, something like that. CustomPC started using scientific reviews (reposted on B-T), and the aforementioned JG, but it's not uncommon for me to visit a site and see morons giving their rail readings for 3 lines with a LOADED CPU and GPU and 4HDDs attached (as if it matters...)
Soldato
- Joined
- 1 Apr 2009
- Posts
- 9,952
I like their PSU reviews actually. But yeah, grain of salt and all that.
anandtech used to be regarded as the best review site years and years ago , i cant say i really keep up with review sites unless im planning on buying something so i have no idea if they are still actually good or not.
toms hardware was always regarded as a plonker ^_^
Soldato
- Joined
- 9 Jan 2008
- Posts
- 6,501
- Location
- London/Camberley
I read that article a while back and while I agree that PSU reviews are best left to people with the right equipments, it doesn't explain why his own reviews are poor. All his motherboard are just first looks and even then, he might as well have just copy pasted the specifications from the manufacturers' sites and added a few photos on top...
Thank god somebody put a clicky word,otherwise I never would have went to the site.