It has been 16 months since our last power supply review, but the long wait is finally over. We have a new PSU and cases editor, and this is the first of many new PSU reviews to come. The first PSU to hit our new testing lab is the S12G 650W from Seasonic. Seasonic is a company that hardly requires any introduction; they are one of the oldest and most reputable computer PSU designers, manufacturers and retailers. Today, their products are held in very high regard by experts and enthusiasts. The company has their own retail range but they also serve as an ODM for many other well-known brands.

The target groups of the S12G series are advanced users, enthusiasts, and generally those who are not afraid to pay a bit more for advanced performance. It has impressive specifications and features, including an 80 Plus Gold efficiency certification and a "Smart and Silent Fan Control", as well as a five-year manufacturer's warranty. Despite that, it does not really break the bank, as the 650W version that we will be testing today retails for $89.99 plus shipping, a price that appears quite reasonable for such a product.

A year or two ago, you would have had a hard time trying to find a high quality 80 Plus Gold certified unit for that kind of price; however, today Seasonic's retail range has tight competition from several other manufacturers, such as Antec, EVGA, Rosewill, and others. As the Seasonic S12G is not a modular unit and doesn't really include any special features, the company clearly relies on quality and performance above all else; middling quality and average performance cannot be tolerated when several other manufacturers are breathing down your neck. Let us see if the 650W version of the S12G can live up to Seasonic's reputation.

Packaging and Bundle

Seasonic supplies the S12G PSU inside a fairly well designed cardboard box, with a black/blue color theme. The front of the box is tidy, with just a few badges with the certifications of the unit on display. The sides and rear of the box however are littered with information about the unit, in several languages.

Companies rarely supply anything noteworthy alongside their PSUs, especially the non-modular models, but the Seasonic S12G appears to be an exception. Besides the usual power cord, manual, and screws, Seasonic also includes a few cable ties, three blue/black velco cable straps, and a case badge with the company's logo.

The Seasonic S12G 650W PSU
Comments Locked


View All Comments

  • tynopik - Saturday, March 1, 2014 - link

    > if I short the rectifier bridge

    I mean on the component side. Cables can and do short and it should NOT cause permanent harm to the PSU (or to other components on other cables). If it does, the PSU is poorly designed.

    > if the distibution grid around your area shifts like "a DJ is rotating the knob", which is not really possible (infinite bus theory)

    The request is based off an incident that actually did happen to me. The power dipped 3 times in quick succession and the power supply blew and took out everything connected to it. Granted this was before I knew better and it was a crappy unit running right at its limit, but PSUs do have to deal with power 'variances' in the real world.
  • E.Fyll - Sunday, March 2, 2014 - link

    Shorting a component = OCP protection. I always perform such tests before proceeding to testing the unit. I don't it that to fail catastrophically with $12.000 connected to it...

    I do understand that this might have had happened to you. It most likely was due to the catastrophic failure of a distribution grid's component. Still, that is not a fault of the PSU, it is an issue with your energy supplier. Your energy supplier is under contract to provide you with electricity of specific quality. If you have such severe problems with your distribution grid, you do not need a better PSU, you need a lawsuit.
  • HangFire - Monday, March 3, 2014 - link

    I'm glad to hear you test OCP. Will you be testing it on all 3 voltage rails?

    As for voltage variances, they are not typically due to catastrophic failures. Quite mundane failures such as tree and animal shorts happen often, and there is a high voltage protocol to burn them off and resume normal power that sounds just like the poster's experience.

    Power in proximity to manufacturing areas can show the effects of large electric motor startup and shutdown. Grids switching in and out contingent power sources can introduce phase issues, and so on.

    Blowing all of these off as unusual or the result of large failure is not realistic.
  • HangFire - Monday, March 3, 2014 - link

    Here is a good overview:
  • E.Fyll - Tuesday, March 4, 2014 - link

    It is entirely realistic, simply because a) they are unusual and b) are not a problem of the PSU itself but a problem caused by your grid.

    Large motors and appliances that could cause a voltage drop are not being fed by the consumer distribution grid. Anything above a medium industry is being powered by polyphase distribution. So you being in the same grid as a massive motor, no, it is not going to happen. You are connected on a distribution bus, which is nearly infinite (hence the term "infinite bus") and will no bulge even when huge generators start. If there is an appliance capable of causing a voltage surge or drop connected into the residential grid near you, that is a problem.

    Trees and animals are a) not mundane failures and b) definitely not just being "burned off and resume". If an animal shorts two distribution phases and no safety kicks in, the current inbalance will easily destroy a large transformer in milliseconds.

    The faults that are being described in APC's white paper certainly are correct and possible. Well, to an extend; after all, it is a paper meant to advertise how good having a UPS is, not really a scientific document. This does not mean that they are frequent, nor that they can be replicated by "rotating the knob". See these ups and downs the waveform does in their examples? Your household grid is doing 50 or 60 of them, depending on where you live, *per second*. The faults rarely have a length of more than a few milliseconds. Rotating a knob 100.000 times per second is definitely not possible, specialized equipment is required to replicate any such test.
  • HangFire - Friday, March 7, 2014 - link

    I'm not the one suggesting the knob methodology.

    I would have agreed with you when I lived in the suburbs. Since moving to a rural area, I can assure you that brownouts, tree related outages and crossed power lines are quite common here. Besides trees the big threat here is drivers hitting utility poles. Very common this time of year, the combination of ice and hill country is quite hazardous, and many of the poles here carry two sets of voltages. In one event just a few years ago, every lightbulb in my house lit up very bright. I lost 5 lightbulbs and 2 small appliances.
  • HangFire - Friday, March 7, 2014 - link

    I guess MY point is that Active PFC power supplies are supposed to take a range of voltages; but I don't see anyone testing that. I test it every few months, when I get out the 100 foot extension cord and use a PC with 5.1 speaker system for party tunes. (My guests are not picky I guess). I use quality hardware, usually a Seasonic so I'm OK but I wonder what the limits are for the cheaper stuff.
  • tynopik - Saturday, March 1, 2014 - link

    I can see where 'component side' could be confusing. I mean 'external to the power supply'. No of course I don't expect you to go digging inside a unit and shorting things out.
  • tynopik - Saturday, March 1, 2014 - link

    > the performance of a PSU in situations that will almost never arise should not influence a buying decision.

    it depends on what you mean by 'almost never' and what the consequences of failure in those cases are.

    New Orleans is 'almost never' hit by a hurricane. Does that mean they shouldn't worry it?

    Saving a few dollars a year by being more efficient in the usual case is great, but if it blows your system during a random power blip, was it worth it?
  • CeceliaAFolger - Sunday, March 2, 2014 - link

    this is a goor site to earn $145 per Month ..........

Log in

Don't have an account? Sign up now