350-450W Roundup: 11 Cheap PSUs
by Martin Kaffei on July 3, 2012 1:30 PM ESTCorsair CX430 V2 430W
Corsair was using Seasonic units for the majority of their products, but most of their cheaper offerings are now manufactured by CWT. The CX430 V2 is the lowest-end unit from Corsair, which still has enough quality to satisfy the customers. It comes in matt black and has a large Corsair logo on the fan grille. The back is perforated with hexagonal-shaped openings and a small power switch can be found above the power input.
The contents of the package are what you'd expect. You get the required four screws and power cord, naturally, along with some cable ties, a user manual with product data and safety references. Corsair prefers a large single-rail 12V design, rated at 28A (336W). The reason for the high rating of the 12V rail is the high power consumption of CPUs and GPUs.The small rails are rated at 20A each with a combined output of 120W; that's comparatively weak compared to some older PSUs, but since modern PCs usually don't need much from the low voltage rails, this will hardly be a problem.
A 120mm Yate Loon fan cools these units. It has a ball bearing and seven sharp-edged fan blades. A plastic guard blocks part of the intake area to help direct airflow.
Cables and Connectors | ||
Connector type (length) |
Main | 1x 24-pin (45cm) fixed |
ATX12V/EPS12V | 1x 4+4-pin (50cm) fixed | |
PCIe | 1x 6/8-pin (50cm) fixed | |
Peripheral |
3x SATA (ca. 50, 65, 80cm) fixed 3x SATA (ca. 50, 65, 80cm) fixed |
|
3x HDD, 1x FDD (ca. 50, 65, 80, 95cm) fixed |
The inside reveals a typical CWT design with three heatsinks, two for the primary side and the third for the secondary side. Three of the filtering caps are attached to the other side of the AC jack. The internal layout is pretty typical using a two-transistor forward converter, with a minimal number of components in the transient filtering. The primary cap is made by Samxon--just like the secondary ones. They are a slightly lower end vendor CWT uses for these units.
67 Comments
View All Comments
Phaedrus2129 - Thursday, July 5, 2012 - link
That Sinan PSU is likely made by LongYi, a small and crappy PSU manufacturer in Guangdong, China. They make knock-off, cost-down designs based on reverse-engineered FSP and similar PSUs.Martin Kaffei - Thursday, July 5, 2012 - link
Thanks for these information. I guess there are more brands who use this manufacturer so it wasn't useless to test a German product.hasseb64 - Thursday, July 5, 2012 - link
Not many need more than 400W anymore, some think it's better to buy more now to be sure about future upgrades, but that view is should be reversed, total watt consumtion in builds going down for every generatio. I hope to see Platinum/Gold PSU from 200-300W in the future.Owning:
460W GOLD ATX
300W Gold ATX
bryanl - Thursday, July 5, 2012 - link
"The Teapo models don't have the longest lifetime and lowest ESR; however, we've never detected a problem with Teapo in power supplies."Official ratings mean little when they're not honest, and while I haven't looked at Teapos in recent power supplies, they had very high failure rates in older CWT supplies, including those sold as Antec Smartpowers and Truepowers.
Martin Kaffei - Thursday, July 5, 2012 - link
True, but Teapo has several products, you can't compare old models with current offerings. And some ratings are wrong for sure, but most manufacturers in this roundup used the same series and some of this companies have less than 1% failure rate according to retail shops.Not least you can't compare old PSUs with a modern 80Plus device. Temperatures are much lower today and switching frequency is different (which is why a low-ESR model would have been nice anyway). The ambient conditions are different. Teapo had problems on mainboards for the same reason. There are no bad or good components, just the right choice for a specific range of functions.
bryanl - Thursday, July 5, 2012 - link
"True, but Teapo has several products, you can't compare old models with current offerings. And some ratings are wrong for sure, but most manufacturers in this roundup used the same series and some of this companies have less than 1% failure rate according to retail shops.Not least you can't compare old PSUs with a modern 80Plus device. Temperatures are much lower today and switching frequency is different (which is why a low-ESR model would have been nice anyway). The ambient conditions are different. Teapo had problems on mainboards for the same reason. There are no bad or good components, just the right choice for a specific range of functions. "
These were mostly Teapo SC (green w/ gold or bronze writing), as new as 2010, a product still being made. What model Teapos were you thinking of?
I believe these operated at frequencies of 60K-100K Hz and averaged temperatures of 50-65C. What are the operating conditions of the capacitors in the 80Plus power supplies?
Power supplies will often continue to operate adequately even years after capacitor failure since some capacitors are redundant, including with motherboard capacitors, and most computers draw little power.
Martin Kaffei - Friday, July 6, 2012 - link
Even then it might be that Teapo changed some ingredients.However, let's assume that this is not the case. Current PSUs switch much faster than 100 KHz in order to reduce component size. E.g. 250 KHz is a common value these days, some of them are even faster. This is just one out of many different conditions:
"averaged temperatures of 50-65C."
No matter where--it will be lower, but where did you measure these temperatures? On which outputs or circuits did they fail? Did they all fail? Is there more or less space for cooling than in most PSUs in this review? How high was the average ambient temperature during operation? Do these PSUs have the comparable ripple values? How high was current? How high was load?
This and even more is what we need to evaluate the position.
For sure PSUs can operate without some of the output capacitors, but not for ever and ever. There are signs. In many cases you can hear noises when capacitors fail (or measure something). Moreoever I've never seen faulty Teapo capacitors in old FSP units and I used plenty of them for years.
The source of such problems can be elsewhere. I had a PSU with an oscillating controller and all capacitors on 5VSB died. Does this mean the capacitors are the problem? No, after replacing a single resistor the controller was stable and the caps didn't fail again.
bryanl - Friday, July 6, 2012 - link
"RE: Capacitors in Rosewill power supply by Martin Kaffei on Friday, July 06, 2012 Even then it might be that Teapo changed some ingredients.However, let's assume that this is not the case. Current PSUs switch much faster than 100 KHz in order to reduce component size. E.g. 250 KHz is a common value these days, some of them are even faster. This is just one out of many different conditions:
"averaged temperatures of 50-65C."
No matter where--it will be lower, but where did you measure these temperatures? On which outputs or circuits did they fail? Did they all fail? Is there more or less space for cooling than in most PSUs in this review? How high was the average ambient temperature during operation? Do these PSUs have the comparable ripple values? How high was current? How high was load?
This and even more is what we need to evaluate the position.
For sure PSUs can operate without some of the output capacitors, but not for ever and ever. There are signs. In many cases you can hear noises when capacitors fail (or measure something). Moreoever I've never seen faulty Teapo capacitors in old FSP units and I used plenty of them for years.
The source of such problems can be elsewhere. I had a PSU with an oscillating controller and all capacitors on 5VSB died. Does this mean the capacitors are the problem? No, after replacing a single resistor the controller was stable and the caps didn't fail again. "
250 KHz has long been common for motherboards, but which PSUs oscillate near that frequency? I wasn't able to find anything operating above approximately 150 KHz at websites that presented oscillographs (Jonnyguru.com, Hardwaresecrets.com) where the fundamental could be identified. Temperature readings were taken with direct contact thermistors or thermocouples touching the metal bodies of the output capacitors while the PSUs operated motherboards without high performance video cards.
In well designed power supplies, few components fail nearly as much as capacitors, and there's a fairly common +5VSB circuit made of 2 discrete transistors that overvolts when it fails, almost always due to a bad capacitor, often 1-2 10V, 1000uF Teapo SCs.
Bambooz - Sunday, May 5, 2013 - link
Forget about Teapo. C(r)apxon is the biggest epidemic right now. They're in basically every Fortron and Fortron based PSU (Bequiet anyone?)If you have a PSU with Crapxon on the secondary side it's almost guaranteed it won't make it to it's 2nd birthday without the crapxons puking their guts out and in turn, sh*tloads of ripple (I've had client computers where shitty cheap FSP OEM PSUs killed HDDs due to their caps failing)
Instability, whining and sometimes (for example Bequiet E5 series and FSP Bluestorm I & II) power-up problems (because of bad caps on the 5VSB circuitry) are pretty much the norm with these..
No matter how good fortrons may sound.. avoid them at all costs unless you can recap them BEFORE putting them into service.
Onus - Thursday, July 5, 2012 - link
I too am interested in lower-wattage units, so this article was a good read. I've been more impressed with FSP's recent offerings. The way the Raider's efficiency quickly ramped up was excellent. Maybe not down to the Picos, I am also interested in physically smaller units, like TFX and SFX models. Seasonic and FSP have some good ones there.My "goto" PSU in budget builds for years has been the Antec EA-380D. I've never had one fail. I buy Seasonic or Antec for my own systems, although I'd consider FSP, Corsair, or Enermax also.
If you go through the assortment of reviews Gabrial Torres at HardwareSecrets has done on CM PSUs, you will understand why I call them "Crappermaster." The "Extreme ..." lines in particular are near-junk, most with liar-labels and/or claimed protection circuits actually missing. That's consumer fraud. One of my favorite comments though comes from a HardOCP review, in which the GX650 was summarized as a "polished turd in a box." Friends don't let friends buy Crappermaster (even their good stuff, because that is supporting a dishonest business).