Hi jonnyGURU,
I've appreciated reading the PSU reviews you've done and posts like this one; thanks for the good work.
Concerning the reviews in general, here are a couple of ideas I've had that I wish more independent reviews would cover as standard points of assessment; I think they're critical to know when really trying to distinguish the safety / reliability of a given PSU in one's system, yet they're rarely ever mentioned at all either by the manufacturer or by reviewers. To your credit you often measure/mention at least some aspects of these things, but a more in depth picture would be greatly appreciated.
* Does the PSU have distinct OCP on any individual (e.g. split or individually OCP protected) outputs,
if so, what's the trip point, what happens when it trips?
* Does the PSU have a general single / individual / master OCP for a given voltage output i.e. all +5V together, all +12V together, et. al. If so, what's the trip point, what happens when it trips?
* What happens when the any of the included OCP or thernal SOA circuits trip due to an significant overload but not, say, a dead short circuit? Does built in current and/or thermal / SOA protection cause a soft power off or clamp down on the outputs? Does a non-user-serviceable fuse blow?
Do any of the outputs surge or some shut down while others still work or what?
* What happens if you dead short a +3.3V lead? ... a +5V lead? ... a +12V lead? et. al. Do things melt? Catch fire? Blow up? Fuses blow? Soft-shutdown that permits the unit to function again after you power off and remove the short and let it cool down a bit?
* Most new systems well support various sleep / power throttling modes that can dramatically reduce system power consumption. However these modes can cause much more frequent and severe variances in the PSUs load than might've been typically encountered in past years. Coupled with various systems like the Intel ATOM, VIA EPIA, or even Intel Penryn one might see systems that consume anywhere from 1-10 or 10-50 watts total in one of their low power (but not in full standby / sleep) modes, but might consume hundreds of watts in their full power state. Does the PSU under review keep its voltage waveforms in specified regulation when the PSU load varies rapidly from say 0->5 watts, 0<->10, 0<->20, 0<->50, 0<->100, 0<->300, 0<->500 as one might see in a low power / high power type of transition?
* Obviously when gaming or computing in a high power state it will also be typical to see massive load changes say when finishing one level in a game and it fades to black from a previous state of 100% GPU activity, or in the case where a CPU is crunching something like Prime95 and one suddenly finishes the calculation. Does the PSU respond well (voltage compliance without excessive droop/surge) to load changes from say 20% max. rated output to 95% max. rated output in these conditions?
* Where a PSU specifies a "minimum load" on any of its outputs, does the PSU in fact regulate well AT this minimum load? And in the case where the rails are actually split at all, what is the effect of having some of these rails UNLOADED as one might do with a PSU with 2 PCIE connectors but only no PCIE attached video card, or without using the EPS8 connector at all, or without using any of the provided HDD molex/SATA connectors? Does the PSU behave well (shut down or not generate harmful over/under voltages) when operating at LESS than minimum load such as when the system begins to sleep?
* (extra credit, but good to know) What does the PSU do when its fan doesn't spin? Burn? Power off gracefully?
Getting back to the original post of the topic here, I think you've certainly expressed some good
information about the split rails issue. I just have a couple of comments:
Is it true that some PSU's that claim to be multiple +12V rails don't have the +12V rail split at all?
Yes, this is true. But it's the exception and not the norm. It's typically seen in Seasonic built units (like the Corsair HX and Antec True Power Trio.) It's actually cheaper to make a single +12V rail PSU because you forego all of the components used in splitting up and limiting each rail and this may be one reason some OEM's will not split the rails, but say they are split. Some system builders adhere very closely to ATX12V specification for liability reasons, so a company that wants to get that business but also save money and reduce R&D costs will often "fib" and say the PSU has it's +12V split when it does not.
Why don't those PSU companies get in trouble? Because Intel actually lifted the split +12V rail requirement from spec, but they didn't actually "announce" it. They just changed the verbiage from "required" to "recommended" leaving system builders a bit confused as to what the specification really is.
I was with you right up to the "Why don't those PSU companies get in trouble?" bit. There's no excuse for them to NOT get in trouble if they're falsely advertising a product's features / capabilities / architecture. Sure it may be a perfectly fine single rail PSU and if they advertised it AS SUCH then it should be reviewed AS SUCH, but IMHO if they advertise it as having split rails and it does not, that's a lie, that's fraudulent advertising, and any reviewer that is aware of such blatant deception should very prominently shame them for it in their review. I applaud you for mentioning the deceptive aspects of this kind of marketing, and pointing out the shortcomings in some of your reviews. I can't quite understand, though, why one wouldn't totally flunk a PSU in a review if it is obvious that it doesn't at all live up to any single one of its advertised benefits / specifications, though. Again, if they marketed it as a single rail PSU, great, maybe it would deserve a 7.0 rating. But if they say it has 3 rails and OCP or whatever and it doesn't, bzzt, 0.0 rating, doesn't live up to advertised specification.
After all, what else do we have to go on besides whether a product lives up to its advertised features and relevant specifications? If it doesn't do that in any way / shape / form, they company needs to correct that if they want customers.
It is true that marketing folks have told us that multiple +12V rails provides "cleaner and more stable voltages", but this is usually a falsehood. Quite frankly, the use this explaination because "offers stability and cleaner power" sounds much more palletable than "won't necessarily catch fire". Like I said before, typically there is only one +12V source and there is typically no additional filtering stage added when the rails are split off that makes the rails any more stable or cleaner than if they weren't split at all.
Agreed. IMHO if they make a marketing claim that is BS as part of the product packaging / documentation / specifications, I'd love to see more reviewers flat out call them on the BS prominently. Although they may have good aspects to their design, it doesn't excuse them lying about its benefits / features!
In fact, Nvidia SLI certification these days REQUIRE that the PCIe connectors be on their own +12V rail to avoid any problems from running high end graphics cards on split +12V rail PSU's.
Well I don't know the history of that specification, but I'd say that something is very wrong in Denmark if the consumer is to rely on this, for instance:
http://www.antec.com/us/productDetails.php?ProdID=23650
NVIDIA? SLI?-ready certified
http://www.slizone.com/object/slizone_build_psu.html
...which currently lists, for instance, the Antec TruePower Trio 650
as being SLI approved for dual 8800GT and various other SLI configurations.
Your own review clearly indicated that this PSU does NOT in any way, shape, or form have
distinct +12V rails for any of its PCIE or other outputs. So apparently nothing in this SLI certification
process caught that bit of deception on the maker's part, which leads one to question whether the
certification is really worth anything if they don't even take steps to ensure compliance with what you're
telling us is a fundamental requirement of the specification. I found it especially amusing that it and its 850W sibling was SLI certified for the 8800GT when in fact I found the 6/8 pin PCIE connector wouldn't even FIT/LATCH into one of my NVIDIA made reference design 8800GT boards.
Overall I greatly appreciate your reviewing efforts, but given what seems to be fairly rampant poor
quality, bad engineering, and false advertising in the PSU industry, I get the feeling that the deficiencies
in many of the models have yet to even be fully revealed despite excellent efforts / reviews such as your
own just given the impracticality of being as thorough as one might wish in depth/breadth.
Another issue of concern is that I've heard that many PSU companies may change the production designs
of a given model without necessarily changing the marketing naming / UPC / SKU of the unit or
or respecifying (or even retesting) it at all. So it seems not unlikely that even a PSU that gets excellent
reviews at one time might be re-engineered, maybe even becoming manufactured by a totally different OEM with totally different guts, and in the end become a PSU that is much different than the one that has been tested / reviewed due to such eventual cost reductions / engineering changes.
Thanks and keep up the good work!