Posted On: Wed, Aug-22-2007
Reviewer: Oklahoma Wolf
Product: Coolmax CUG-700B 700 Watt Power Supply
Product Link: http://www.coolmaxusa.com/productDetailsPower.asp?item=CUG-700&details=features&subcategory=sli&category=sli
Supplied By: CoolMax USA
Price: 129.99 @ NewEgg

Introduction

     The quest for higher efficiency and environmental friendliness has led us to the Coolmax Green Power CUG-700B 700 watt power supply today. Will our review sample herald in a new era of stability and efficiency? Come inside and see.

Page 1 -

     I have a confession to make - green is my favorite color. Fortunately for me, a lot of things in this world are green. Why, take my desktop for instance: the mouse pad is green, my screwdriver is green, my wallet is green, and this cheese I got out of the fridge is green. Wait a second, maybe I shouldn't eat that.

     Something else in my house is green today, if the box it came in is any indication: the Coolmax CUG-700B. And so, with much rubbing of my hands in anticipation for all the cool greenness contained within this box, I hereby declare today's review underway.

     The Coolmax Green Power series, according to the back of the box, consists of two models... the 700W we are looking at now, and a 600W model. Both boast triple 12V rail topology, with the only differences in the specs being on the 12V rails. Since triple 12V topology is a mite unusual for a high power unit, I immediately began to look forward to getting a peek inside to see how many 12V rails it really had.

     Except for a couple of cool little logos claiming RoHS compliance and Intel/AMD dual core support, the front of the box is pretty uncluttered and tasteful, I daresay.

     On the side of the box, we can see a long statement on how this power supply meets RoHS compliance. Then, we see a great big long list on what Coolmax claims is not present in the unit. While I suppose it's nice to know what's not in the box, I was getting anxious to see what WAS in there. After that list, I wondered if perhaps the power supply had been replaced with a nice, safe, RoHS compliant block of wood or something. But I turned the box around again to see what else was printed on it, confident I would in fact find a power supply within.

     It's a little hard to read this, so I'll type out some of the highlights:

  • triple 12V output rails
  • advanced double forward circuit and double layer PCB
  • active PFC
  • Super high efficiendy [sic] maximum 84%
  • Ultimate balance between cooling and noise level
  • Dual PCI-Express power connectors fully support SLI & Cross-Fire system
  • Gold plated terminal

     There are a few more less interesting bullet points on the box, but I think I got the more relevant ones covered. I can't wait to see which terminal is gold plated. What's on the side opposite the big RoHS compliance statement? Why, another long RoHS compliance blurb I'm way too lazy to type out here. Or get pictures of. Time to open the box and unleash the greenness within.

     I call false advertising! That power supply isn't green at all! The big white box just behind the power supply is the first thing you see when you open the lid, and it contains the modular cables. The black box holds the power cable and mounting screws. Finally, we see an owners manual with so much information contained within that I'm certain you could almost build one of these Green Power units from scratch. Ok, let's have us a table wherein we can see all the numbers being promised by this PSU.

Coolmax CUG-700B +3.3V +5V +12V1 +12V2 +12V3 -12V -5V +5VSB
28A 30A 16A 25A 17A 0.5A N/A 2.5A
Max Combined Watts 180W 500W 6W N/A 12.5W
18.5W
700W

     I must admit, the triple 12V topology is weird all by itself. But the numbers up there make it all the weirder - why not even out the specs, instead of giving 12V2 so much power compared to the others? Despite the otherwise complete information in the manual, it was not immediately clear what rail went where. Some connectors were specified as being on +12VDC with no number to signify which one, while others like the SATA connectors claimed to be on +12V3DC. That's a little odd, and I think I'm going to be taking this apart after the load testing to see what the real story is.

     But first, a good shot of the modular cables and the end of the power supply you plug them into. All connectors are Molex Mini-Fit Junior type, which are fairly reliable. Those two little two pin connectors are supplementary 3.3V connectors for the SATA cabling, which uses a combination of these connectors and the four pin connectors next door to power the SATA connectors. At this point, you might be thinking to yourself, hmm... I see three different colors, could that be how the three 12V rails are split up? I'll answer that later, I promise.

     In the meantime, let's throw together a nice table with a rundown on the modular cables and how long they are. The manual helped me get really lazy here, by telling me exactly how long each cable was. I just copied it all down.

Type of connector: Coolmax
CUG-700B
ATX connector (550mm) 20+4 pin
2 x 2 12V connectors (550mm) 1*
2 x 3 PCIe (550mm) 2
8-pin Xeon/EPS connector (550mm) 1*
6-pin Xeon/AUX connector 0
5.25" Drive connectors (550mm + 200mm + 200mm) 8
3.5" Drive connectors (+200mm) 2
SATA Drive power connectors (550mm + 200mm) 4
Fan only connectors (thermostatically controlled 12V only) 0

*8 pin EPS connector splits apart

     Now that the preliminaries are done, it's time to make the doughn... er, test the power supply. Yeah, that's it. We don't need no doughnuts. I'll be right back. I'm not going out for pastries, honest.



Page 2 -

     Time to get started the shakin' and bakin.' As is the tradition here at jonnyGURU.com, power supply torture testing will be handily taken care of by a SunMoon SM8800 ATE, a fancy and very expensive piece of test equipment designed just for testing computer SMPS units like this one. The SunMoon is able to be programmed for up to ten tests, but we're only going to throw eight of them into gear on the CUG-700B. We're going to do this first at room temperature, get some oscilloscope shots, and then we're going to try and make the Green Power scream by roasting it alive in a borrowed incubator.

     A quick word about tests one and eight - these are crossload tests, designed to see how well the power supply copes with first high 3.3V and 5V loads but unusually low 12V loads; and then high 12V loads with the 3.3V and 5V barely loaded. A really well designed supply won't flinch at this treatment, despite the fact that only Mom's old Duron based rig still uses 5V to power the CPU.

Results from Coolmax CUG-700B COLD load tests
Test # +3.3V +5V +12V1 +12V2 +12V3 DC Watts/
AC Watts
Eff. P.F. Intake/
Exhaust
Simulated system load tests
Test
1
20A 20A 1A 1A 1A 203W/
250W
81% .97 25°C/
28°C
3.29V 4.66V 12.37V 12.36V 12.36V
Test
2
2A 4A 2A 2A 2A 114W/
141W
81% .94 25°C/
27°C
3.35V 5.02V 11.83V 11.84V 11.85V
Test
3
4A 7A 4A 5A 5A 229W/
282W
81% .97 25°C/
28°C
3.33V 4.97V 11.82V 11.80V 11.83V
Test
4
8A 12A 7A 7A 7A 349W/
425W
82% .98 25°C/
29°C
3.30V 4.88V 11.83V 11.82V 11.87V
Test
5
11A 15A 9A 9A 10A 455W/
538W
85% .99 25°C/
29°C
3.28V 4.81V 11.82V 11.81V 11.87V
Test
6
13A 18A 11A 12A 13A 560W/
678W
83% .99 25°C/
31°C
3.26V 4.75V 11.82V 11.79V 11.88V
Test
7
18A 22A 13A 14A 14A 675W/
897W
75% .99 25°C/
36°C
3.23V 4.62V 11.91V 11.88V 11.98V
Test
8
1A 1A 13A 13A 13A 467W/
564W
83% .99 26°C/
28°C
3.31V 5.09V 11.34V 11.32V 11.41V

     Our cold testing results tell us several interesting things. I was stunned to see that not only was the efficiency living up to the claims on the box, it was actually exceeding the 84% claim on test five. That's awesome! Finally, I get to review a half decent power supply! I'm ignoring the numbers on test seven, you say? What test seven, I see no test seven. Oh... right. THAT test seven. Efficiency took a nosedive here, leading me to think the power supply was, to put it bluntly, past its limits at this point. The 5V readings here lend some more weight to this conclusion.

     Another item of interest can be seen in the voltage readings on tests one and eight. Quite simply, this power supply doesn't like crossloads one bit. Test one saw the 5V bottom out, while test eight saw the same trend at 12V.

     In general, voltage regulation was rather mediocre. 5V takes the biggest hit in the progressive load tests, while the 12V rails start out below 12V and never really rise to meet it until, to my surprise, test seven where the 5V rail takes a dive. These voltage readings tell me one thing above all else - this power supply does not have independant voltage regulation like many competitors at this power level. Instead, it's an older group regulated design, much like the FSP Epsilon design. It's Epsilonilicious. PFC readings are just what they should be for active PFC.

     The power supply apparently thought its job was also to entertain, for it started to buzz midway into test one. The buzzing didn't last though... it was gone the moment test two began. However, the buzzing combined with the rather awful efficiency numbers for test seven was an ominous sign. But, on we went to the hot box testing anyway.

Results from Coolmax CUG-700B HOT load tests
Test # +3.3V +5V +12V1 +12V2 +12V3 DC Watts/
AC Watts
Eff. P.F. Intake/
Exhaust
Simulated system load tests
Test
1
20A 20A 1A 1A 1A 203W/
257W
79% .99 28°C/
32°C
3.28V 4.65V 12.37V 12.37V 12.38V
Test
2
2A 4A 2A 2A 2A 114W/
144W
79% .93 28°C/
31°C
3.35V 5.02V 11.83V 11.84V 11.85V
Test
3
4A 7A 4A 5A 5A 229W/
284W
81% .97 28°C/
31°C
3.33V 4.97V 11.82V 11.80V 11.83V
Test
4
8A 12A 7A 7A 7A 349W/
428W
82% .98 31°C/
35°C
3.30V 4.88V 11.83V 11.82V 11.87V
Test
5
11A 15A 9A 9A 10A 455W/
539W
84% .99 32°C/
37°C
3.28V 4.81V 11.82V 11.81V 11.87V
Test
6
13A 18A 11A 12A 13A 560W/
688W
81% .99 32°C/
39°C
3.26V 4.75V 11.82V 11.79V 11.88V
Test
7
18A 22A 13A 14A 14A N/A N/A N/A N/A
FAIL FAIL FAIL FAIL FAIL
Test
8
1A 1A 13A 13A 13A 467W/
569W
82% .98 30°C/
36°C
3.31V 5.09V 11.32V 11.31V 11.40V

     Hot box tests gave us pretty similar numbers to the room temperature ones, with a few exceptions. A couple voltages went down a little bit more, nothing too unexpected. Test eight saw 12V numbers dive well below spec, while test one saw the 5V doing the same. Voltage regulation was even worse than it had been at room temp. Efficiency took a hit, dropping below 80% on a couple of tests.

     The real fun happened, yet again, in test seven. The SunMoon was only about a minute into the test, when the Green Power decided it was time for a spectacular light, sound, and smoke show. BANG! After airing out the lab, we took the CUG-700B apart to see what had happened. But before we get into that on page three, let's see some scope shots from the room temperature tests, back when the Green Power could kinda sorta almost try to handle test seven.

     We used a USB Instruments Stingray DS1M12 right at the load to get these pretty little shots.

Coolmax CUG-700B

+3.3V

+5V

+12V1

+12V2

+12V3

test 1

test 2

test 3

test 4

test 5

test 6

test 7

test 8

     As we can see, ripple stays pretty well in hand until test seven, when the power supply decided to tap a keg and get partying. Ripple shot up to about 90mV on the 12V rails. In addition, we got some other funky scope readings back in test one when the buzzing started, and every rail went close to or out of the ATX spec.

     Our test seven ripple results can be considered a further indication that the power supply was getting to be way past its limits. Apparently, the OCP was asleep on the job. Too busy thinking about getting up and making doughnuts, maybe. I'll... just be right back again. Excuse me.



Page 3 -

     Mmmm... iced tea. What? Did you think I was really after doughnuts? Never happen. No, that's not a cruller. Let's get the lid off the Green Power and see what crawled down its throat and died.

     While this shot is too far away to tell much, we can see the large single input filter cap that is the hallmark of an active PFC design.

     T&T. Dynamite. I'm simply thunderstruck. Now that we can see the fan, we're back in black. I'm going to keep a stiff upper lip and look for clues to see who made who. That is, who made this power supply. But before I do, see that smudge on that there fan blade? That's a result of the light show we got in the hot box. I'll tell you what made that smudge in a few minutes.

     The UL file number on the label proclaims that Coolmax apparently has their own number. Terrific. Too bad the huge "Sirtec" silkscreen on the PCB gives away who really made it. Nice try, Coolmax. Who else uses Sirtec? Yesico does, and Thermaltake used to, before that company switched OEM's a year or two back.

     Coolmax has traditionally used a number of different OEM's, but most often used is ATNG power. These Sirtec models are a change for them. Who else uses ATNG Power? A whole lotta Rosewill. Get it? Hahahahahaha! Ok, I'll quit with the AC/DC jokes now... just don't throw that rotten fruit at me.

     This is the secondary side of the Green Power. Teapo caps a-plenty. While some questions have been raised on the wisdom of using Teapo on computer mainboards, this reviewer has never found reason to doubt them when used in a computer SMPS. Here we also see the two large filter chokes that signify group regulation.

     Of particular interest are the two solder points where the yellow 12V wires are soldered in. You see, 12V1 and 12V3 are the same rail. Not just bridged together, they share the same PCB trace. Seriously. This is really a double 12V design. And what's odd about it is that the two actual 12V rails appear to get their own pi filter, but share the same source. So, this is really a single 12V design with only filtering to separate the two rails. This explains why our load tests were so close between 12V1 and 12V3, with 12V2 being a little further off.

     A good shot of the primary side. Remember the double forward mumbo jumbo from the box? That actually refers to the fact that this PSU uses a double forward converter in the primary side. Like Seasonic, and the FSP Epsilons. That's why we were doing better than 80% so easily. Again, Teapo caps are used. While I'm at it, the whole thing's done in Teapovision, just to save some typing.

     A picture of the modular connector PCB. Here, we can finally see what blew. Take a close look - can you guess what went up around all that fuzziness in the center of the picture? C'mon... guess. Seriously... take a guess. Hey! Rotten peaches really hurt!

     The brown fuzz marks the location of what used to be a Teapo capacitor, which had been launched straight off the board and into the fan. This capacitor, to be precise:

     Moving on, let's have a look at the modular PCB and finally puzzle out what 12V rail does what. Keep in mind that there are actually only two 12V rails, separated by no more than filtering, so it doesn't really matter a whole lot what goes where on this bad boy. In my examination of the above board, I can tell you that:

  • 12V1 goes to the 24 pin ATX connector and the upper PCI-E connector
  • 12V2 goes to half the big blue EPS 12V connector, the lower two black four pin connectors, and the lower PCI-E connector
  • 12V3 goes to the other side of the EPS connector and the upper three black four pin connectors

     In addition, only two sets of 12V wires are actually soldered to the modular PCB. 12V2 on the mainboard goes to 12V2 on the modular, while 12V3 on the mainboard (which is electrically combined with 12V1 remember) goes to 12V1 on the modular PCB, where it is again electrically combined with 12V3 on the modular board. Kind of an oddball way to do things, but there you go.

Methinks its time for some scoring.

Performance (40% of the final score)

     Efficiency was wonderful until we asked the supply to do full power, but the voltage regulation left something to be desired. That something was stability. OCP decided to take a nice long nap, and failed to wake up in time to keep the supply from launching a capacitor into outer space. But, at least the active PFC circuit behaved itself.

     Still, when taking everything into account, it didn't quite do as badly as the other units I've tested to date. I'm going to lighten up some and give the Green Power a 6.

Functionality (20% of the final score)

     First things first, the owner's manual is terrific. Even though it lacked power distribution information (which doesn't really matter anyway), it does a very good job saying what cable goes where. The color coded connectors on the back of the unit make it a snap to fit the right one.

     The modular cables on this are terrific. They're sleeved well and have plenty of length to them. You can even remove the main ATX cable if you're in the mood to run this as a secondary PSU. But, the box was right, the connectors are gold plated. That may sound like a good thing, but it's going to cost the unit here. You see, metals of two different types don't always get along, and it's not too likely your mainboard or hard drive connectors are gold plated. I know the ones on the back of the PSU itself aren't. While this typically won't result in poor connections for years, it's something I cannot ignore either. This is not a high end audio preamp, Coolmax! We don't need gold in here! But I won't ding it too hard for my connector paranoia. 7.

Value (30% of the final score)

     These can be found around the net right now for around the $125 mark. No, really. All the performance of an FSP Epsilon 600W combined with unreliable OCP for the low low price of $125. Extended warranty? How can I lose! 5.

Aesthetics (10% of the final score)

     I like the way this unit looks, with its matte black finish and gold looking fan grilles. I even like the color coded modular connectors on the back. And I especially like the sleeving. The Coolmax is getting a 9 outta me.

And so, our total is a rather dull 6. No recommendation this time, I'm afraid.

Performance 6
Functionality 7
Value 5
Aesthetics 9
Total Score 6

Summary

     The Coolmax CUG-700B is a highly efficient but uninspired performer, choosing more to be like the competition over at FSP rather than rise above it. Efficiency was great, but the lack of stable voltage regulation and adequate protection circuitry really counts against it. I would consider taking a hard look at Seasonic's offerings before considering this particular power supply, and even the true competition in the FSP Epsilons and OCZ GameXStreams are better options to consider.

The Good:

  • very efficient
  • active PFC
  • adequate build quality

The Bad

  • poor voltage regulation
  • a little expensive
  • modular cable connectors are gold plated

The Mediocre

  • it 'sploded
  • OCP was nowhere to be found




This review was provided by : JonnyGURU

The URL for this review is :
http://www.jonnyGURU.com/modules.php?name=NDReviews&op=Story&reid=68