Rosewill Capstone 450W and 650W 80Plus Goldby Martin Kaffei on March 21, 2012 8:00 AM EST
- Posted in
- 80 Plus Gold
Package Contents, Fan, and Power Rating
Rosewill offers a few extras in the package. Besides the standard power cable and four screws, you get a short user manual, and all the modular cables come in a bag (650W only). According to the manufacturer's specifications, the whole series is SLI-ready and uses a single +12V output. Again they used a "mesh sleeving on all cables for easier cable routing and better ventilation". The 80 Plus Gold certificate is another feature—nothing special these days, but still a good basis.
The 450W PSU has a +12V output rated at 37A, which is quite a lot. Both +3.3V and +5V get fed by this output and are specified at 20A each. Those outputs can provide up to 100W, which is lower than the values we saw from Seasonic (560W Gold). But these values are still more than sufficient for all modern PCs, though older PCs might need more power here. The +5VSB at 3A is also relatively strong. The 650W model can deliver 648W on +12V and up to 110W on +3.3V and +5V. -12V and the standby power are identically rated.
The 140mm fan in both models is an RL4Z B1402512M from Globe Fan. This model is used in many power supplies that are manufactured by Sirtec (High Power) and SuperFlower. It has a moderate (M) speed within the RLXX series. The fan uses 0.33A and consists of 11 fan blades with rough edges. It's a ball bearing (B), but a cheaper one. We would like to see Sanyo Denki here or an HDB fan.
Post Your CommentPlease log in or sign up to comment.
View All Comments
Southernsharky - Wednesday, March 21, 2012 - linkAre we now trying to con the public into believing that they need a high end power supply to make their system work?
I've been using the cheapest PWs I can find in my systems since oh... 1985 and I have had a few fail, but over the course of 25 years, I can only think of maybe 3 that failed. And there is no guarantee the same thing wouldn't have happened with the expensive models.
As for energy efficiency...... its over rated. Unless the efficiency saves me more money than what I spent on being efficient then its not relevant. I'm guessing the difference between an efficient card and an inefficient card is right around .50 cents to a dollar a month. So I'd have to run 80-120 months to realize a difference.
Dusk Star - Wednesday, March 21, 2012 - linkIn previous reviews, they seem to highlight the fluctuations in the power output- critical for a stable overclock. And this is largely a site for tech-savvy people; not the type to be as easily conned. I have also heard too many "horror stories" of people having their PSU fail and burn out a $300 graphics card on their way out. My point is, there may be more point to these reviews than to get people to buy more than they need.
Martin Kaffei - Wednesday, March 21, 2012 - linkOne problem is that most cheap PSUs have no real power rating. Usually you don't get any problems but some users overestimate the possible output power and use $10 PSUs in a high-end PC. That's what Dusk Star mentioned in other words.
Basically cheap PSUs have advantages (simple circuit designs) and some disadvantages (low efficiency). But peak power rating is unacceptable for me. In addition we should keep in mind that electricity costs will rise.
mpschan - Wednesday, March 21, 2012 - linkDusk Star is right.
I had a half-way decent power supply blow out a motherboard and a graphics card. $400 down the drain.
It taught me a valuable lesson: spend more money on your power supply so it doesn't destroy hundreds of dollars in other equipment.
I love seeing reviews like this. Keep up the good work!
jabber - Friday, March 23, 2012 - linkIt always makes me laugh when I read in the Power Supply section of forums someone comes on and says -
I need to run a i3/4GB/2xHDD PC with a $80 GPU. What level of PSU do I need?"
Then all the experts start pulling figures out of their backsides and start saying "Oh you need a 800W minimum!"
Fact is 85% of even the most committed enthusiasts PCs probably would run just fine with a good quality 500W.
michaelheath - Wednesday, March 21, 2012 - link"Are we now trying to con the public into believing that they need a high end power supply to make their system work?"
I would imagine the enthusiast community is capable of independent thinking, and they read more than one review before coming to a purchasing decision.
"I've been using the cheapest [PSUs] I can find in my systems since oh... 1985 and ... over the course of 25 years, I can only think of maybe 3 that failed."
That's 3 more than the number that have failed on me in nearly the same time period.
"And there is no guarantee the same thing wouldn't have happened with the expensive models."
There's no guarantee that a $100k Mercedes Benz won't break down after 100 miles of driving. What's your point?
"As for energy efficiency...... its over rated. Unless the efficiency saves me more money than what I spent on being efficient then its not relevant."
How I read your statements: Never mind the cumulative difference in power consumption if everyone ran a ≥85% efficient power supply, including being able to reduce the resources required to power millions of computers in North America alone. Forget the fact that a quality power supply would more than likely outlast a cheap PSU two-fold, thus reducing electronic waste. What matters is I don't feel like I'm saving money.
I have over 200 systems deployed in my work environment, all with ≥85% efficient power supplies. I assure you, the people who approve my spending would much rather see a smaller electric bill and lower repair/replacement rates (which, coincidentally, occurred after I started using higher-efficiency-yet-lower-wattage PSUs) than the opposite if I saved a few bucks per machine and didn't care about what PSU I used. They're especially grateful during the Summer, when an inefficient PSU produces more heat and becomes even less efficient to maintain stable power output... which doesn't happen with a good, quality, efficient, yet slightly more expensive unit.
anactoraaron - Wednesday, March 21, 2012 - linkIf you are using the cheapest 'PWs' you should measure the power draw under those PWs. My rig with an 'el cheapo' used to draw 470 watts under load and now with an 80 plus PW it draws 285 watts. That's about 200 watts difference. Efficiency isn't exactly over rated. FWIW, my rig used to draw about 30 watts when it was off with the 'el cheapo' and now only draws 6 watts when powered down. So unless you are constantly unplugging your rig or you are switching the PW off energy efficiency comes into play there as well.
JarredWalton - Wednesday, March 21, 2012 - linkHis suggestion that a "cheapest PW" would only use $0.50 more per month is also way off. $0.50 per month works out to a difference of 7W for 24/7 usage. Even if we take a lower end system that idles at 60W output from the PSU, that would use 86W with a 70% efficient PSU or 71W with an 85% efficient PSU -- a difference of 15W or around $1.00 per month. If it were a more reasonable system that used an average power of closer to 120W then that doubles to $2.00. (That's at my price of $0.10 per kWh, which is lower than the national average of $0.16 per kWh.)
Of course, even 70% efficiency is more than the "cheapest" power supplies would offer. Let's say you get a real dud and it's 50-60% efficient (which is likely if you buy a $25 PSU). Now you're looking at a 60W system load drawing 100-120W vs. 71W, or a moderate load of 120W that would pull 200-240W vs. 141W. That would be a difference $2 to $3.50 per month for a low power system, or $4 to $7 per month for a moderate system. Considering you would be paying about $25 more in power per year as a minimum, or as much as $84 per year for a moderate system, you'd more than pay for an upgrade to a better PSU in one year.
adece - Wednesday, March 21, 2012 - linkI'll take your word on "the cheapest PSUs" and link you to an article about it:
I'm pretty sure AnandTech did something similar but I can't find the article.
adece - Wednesday, March 21, 2012 - linkFound it: