Is Cooler Master’s 1300 W unit not large enough? How does a 1500 W unit strike you? That’s enough to start your car if you drive a compact car (Note to self: Try this sometime). It’s also a touch more than most users need, you’d best have four top-end GPUs and a serious CPU if you’re contemplating this PSU. Alternatively you could be a bencher running a couple GTX580s on LN2. In any case, 1500 W is a lot.
(Courtesy Cooler Master)
Built for quality and performance, the Silent Pro M2 1500W utilizes dual 12V rails that boasts a significant increase in available power output per power supply model over its predecessors. It translates into increased total amperage, high quality components to increase efficiency, environmentally friendly RoHS and ERP 2010 Certifications, and a long life 135mm fan with a durable Hydraulic Dynamic Bearing that decreases noise.
Nothing sticks out hugely here, the usual PSU manufacturer confusion about the word “silent” is evident. (Side note: “Silence” and “Silent” are absolute words, something is silent or it isn’t. If it makes any noise at all, it is not silent)
I like modular cables, the flat ribbon style that are included with this PSU are my favorite type. The Erp Lot6 business I also like a lot, roughly translated it means “Off means OFF”. The 70 A and 55 A 12 V rails are quite juicy, I’m curious what they’re connected to and what the OCP point is. Lots of protections are claimed, I expect that they exist as well. I’ll check if possible of course. Lastly the heatsinks, large slabs of heatsink aren’t exactly uncommon in the PSU world, nor would I call them aerodynamic. Copper is rare however, I’ll be interested to see what bits have the distinction of being cooled by a copper heatsink.
|Type||ATX 12V V2.3|
|Dimension||150 x 220 x 86mm (5.9 x 8.6 x 3.4 inch)|
|Input Voltage||90-264Vac (Auto Range)|
|Input Current||18 – 9A|
|Input Frequency Range||47 – 63Hz|
|PFC||Active PFC (>0.9)|
|Power Good Signal||100-500ms|
|Hold Up Time||>17ms|
|Operation Temperature||0-40°C(Nominal Input Voltage)|
|Certification||TUV / CE / UL / FCC / BSMI / GOST / C-tick / KC / CCC|
|Fan||Silent HDB 135mm|
|Connector||M/B 24 Pin Connector x 1
CPU 8 Pin x 2
PCI-E 8 Pin x 6
PCI-E 6+2 Pin x 6
SATA x 12
4 Pin Peripheral x 5
4 Pin Floppy x 1
Nothing really jumps out here other than the wattage (huge). Being rated for operation down to 90 V is nice, officially speaking modern PSUs have to be able to operate at 90 V, but many manufacturers put 100 V as the minimum anyway. I lack the ability to drop the voltage on this much of a load, so testing to see if this unit can do 1500 W with 90 V input cannot be done.
The connectors are somewhat interesting as well. You don’t get as many SATA / Molex connectors as I would expect. The NZXT HALE82 N series has the same number despite being half the wattage. On the other hand here we have two 8-pin CPU power connectors (if your motherboard doesn’t have at least one 8pin CPU power connector you have no business buying this power supply), six 8pin PCIe connectors and six 6+2pin PCIe connectors. Again, this is a top end serious PSU. If you don’t have at least three top end GPUs you’re going far overkill. Top-end GPUs all have at least one 8-pin PCIe connector on ’em, so this is fine. In my ideal world one of the CPU power wires will be modular, we’ll see how that goes. Lastly, the motherboard connector is listed as a straight 24pin, this is okay for the same reasons the other solid connectors are okay.
Photos Part One: The Box
I’m sure that the lightning on the box is supposed to signify the staggering output power of this PSU. I can’t help but think about the staggering arc you could manage if you bypassed OCP/OPP/SCP and played Arc Welder with it. Other than the lightning pictures, the box is fairly standard Cooler Master PSU business, white and purple are the colors of choice. There are a few bits of marketing saying what we read in the previous section, as well as a couple of charts. The charts are remarkably honest looking. This is refreshing.
I suspect the box can be opened, let’s check.
That is not a small PSU. That’s not even a large PSU. That’s a HUGE PSU. We’ll look at it shortly. First though here are the accessories you get with this 1500 W $400 MSRP (retail is lower) monster:
Yes, that’s all. Four black screws. Nothing more. I can’t help but feel like there should be a couple hook+loop straps or something in here too. I’m glad they included screws at least, I’d feel awfully silly using silver screws to install this thing.
Photos Part Two: The PSU and Cables
I’ll press onward, I expect this will be inspiring after the disappointment of the accessories pack (or lack thereof).
Nice looking unit though, huge. There’s a very important warning label plastered over the power plug, let’s check that area out in detail.
Not only do you need to have an abnormally powerful outlet to plug this thing into, it has a special C19 power cord. It has this cord for two reasons, one is that the C19 plug can deal with more amps, the other is that C19 cables only come with heavy gauge wire in the cord. If this unit had a standard plug it would be possible to plug it in with a 18 gauge cable. If you loaded the unit down to 1500 W with that cable it would likely melt/burst into flame. Not good! The C19 plug comes on 14 gauge cable at the thinnest, which is perfectly capable of shuffling enough power for this unit. Don’t lose the cable.
Attached to the unit are the 24pin motherboard power cable and the two CPU power cables, plus three PCIe cables that each one one 8-pin and one 6+2pin plug. I’d prefer that the second CPU power cable was modular, but oh well.
The hardwired cables are quite long, 22-24″ easily. The second PCIe plug extends another few inches. No sleeving on that part unfortunately.
The modular cables are indeed the flat ribbon sort. The insulation feels like silicone, it’s very flexible.
See that character second from the left? That is a 8-pin CPU power to 4-pin CPU power adapter, it may count as an accessory and make the accessories section a little less depressing. It’s here in the cables section because it is a cable. Just in case you have one of the motherboards that has one 8-pin and one 4-pin. Or I suppose to use this PSU with a motherboard that has a single 4-pin, though I still think that would be awfully silly. Here we have another four PCIe cables, three with one 8-pin, one with a 6+2pin connector, and one more cable with two 6+2pin connectors. If you’re thinking “Wait, that’s 14 plugs!”, you’re right. But only three modular cables can be plugged into the PSU at a given time, resulting in 12 plugs. Why they spent the money on a fourth cable I don’t know, but the 2×6+2pin cable will be nice for anybody who wants to run a bunch of 2x6pin cards. It really does have three SATA cables with four plugs each and two Molex cables. One with three Molex plugs and one with two Molex and a FDD plug. All of those cables can be plugged in at the same time.
Last for this section, a few angled shots.
The fan looks mean, I like it! I’m not sure how the modular connector board and any components mounted in that end are supposed to get airflow, but I’m not an engineer, and I suspect at least a few of the people that put this unit together are engineers. That’s a good thing in theory. In any case, it’s time for testing!
Load Testing, Cold
A decent load test of a PSU requires a decent load. Contrary to what some may believe, that means you need a known load that can fully stress the PSU. Computer hardware does not cut it. Worse if the PSU fails during testing it might take out the computer hardware anyway. Commercial load testers cost a lot of money. I do not have a lot of money, so I built my own with juicy power resistors and a Toyota cylinder head. It works great. I’ll be using it to load this thing down fairly severely and will check voltages and ripple (more on that later) at various points. The down side to my tester is that the loads it can put on PSUs are fairly coarse, they go in increments of 48 W for 12 V, 50 W for 5 V and 22 W for 3.3V. Those wattages assume the PSU is putting out exactly the official rail voltage, a PSU putting out 12.24 V rather than 12 V will be at 49.9 W per step rather than 48 W. I file that under the “tough beans” category as I figure if a percent or two of load makes that much of a difference the PSU manufacturer should have hit the voltage regulation more squarely. It does make calculating efficiency difficult at best. however, given that the input power is read via a Kill-a-watt, the efficiency numbers are dubious to begin with. Kill-a-watts not being known for extreme accuracy on things with automatic power factor correction (APFC). For these reasons I am not listing the efficiency.
The ATX spec says that voltage regulation must be within 5% of the rail’s official designation, regardless of load. It doesn’t actually mention that the PSU shouldn’t explode, though I expect they figured it was implied. Exploding is a failure in my book regardless.
I had to upgrade my load tester again to deal with this monster, it has 200 W on the previous largest unit tested.
No crossloads this time I’m afraid, other than the the maximum load is rather of a crossload all things considered (50 W of 5 V and 1392 W of 12 V!). I’ve completely run out of room for more load resistors at the 1392 W 12 V mark. Here are the results cold:
|Wattages (total)||12 V Rail||5 V Rail||3.3 V Rail||Kill-a-watts|
|1392/50/44w (1486w)||12.10||5.07||3.34||(Umm, no)|
Yes, there is a big gap in the loads. No, you didn’t miss anything. The 12 V rail’s regulation is excellent at 0.82%, the 5 V isn’t bad either at 0.98%, the 3.3 V rail can’t keep up with the other two, but it still manages 2% regulation. The kill-a-watt was very unhappy at full load and actually started beeping at me and telling me that it was over. I had to unplug it before the thing exploded in flames. A rough guess at efficiency, pretending that the loads are exact (they’re actually higher than listed due to the unit running the rails over the nominal value) and pretending that K-A-W units are exact, as well as pretending the last test the K-A-W survived puts efficiency above 88% from test 2 onward. That beats 80+ Silver requirements easily, but falls short of Gold.
Even with a sustained full load, the fan did not make much noise or spin up nearly as much as I expected. This intrigued me. The air coming out was warm but not hot.
I tossed this unit into the New Improved Enclosure of Uncomfortable Warmth (it won’t fit in the Enclosure of Unreasonable Warmth, the NIEUM is a styrofoam cooler) and raised the intake temperature to 40 °C. With about four minutes of 40 °C intake temps the following voltages were read:
|Wattages (total)||12 V Rail||5 V Rail||3.3 V Rail||Kill-a-watts|
|1392/50/44w (1486w)||12.11||5.05||3.33||(still not happy)|
Impressive! It really didn’t care much about the heat from a regulation standpoint. I mention four minutes in the leadup because at the five to six minute mark the unit shut down. I was able to convince it to fire up again after blowing some air in it to cool the heatsinks off and carefully draining the 5VSB rail (and therefor the APFC holdup caps), a second hot load test gave the same results, as did a third and a fourth. This unit is simply unable to put its full load out at the rated 40 °C. I suspect this is due to the location of the fan, we’ll look into this matter in more detail in the Dissection Section.
Ripple is fluctuation of the PSU’s output voltage caused by a variety of factors. It is pretty much impossible to have zero ripple in a SMPS computer power supply because of how a SMPS works, so the question is how much ripple is there? In the regulation testing phase we found out how the PSU does at keeping the average voltage at a set level, now we’re going to see what that voltage is doing on really short time frames. The ATX spec says that the 12 V rail cannot have more than 120 mV peak to peak ripple, the 5 V and 3.3 V rails need to stay under 50 mV.
If that isn’t complicated enough for you, there are three forms of ripple to keep track of as well. Long-term ripple from the PSU’s controller adjusting the output voltage and over/undershooting, correcting, overshooting, etc. Medium-term ripple from the voltage controller charging and discharging the inductor(s) and capacitor(s) that make up the VRM, and very short-term ripple caused by the switching itself. The first and second forms are the most important, if they are out of spec it can cause instability at best or damage in extreme situations. The very short-term (I call it transient ripple) flavor is less crucial, excessive amounts can still cause issues though it takes more of it to do so. The ATX spec does not differentiate, as far as the spec goes 121 mV of transient ripple is just as much of a failure as 121 mV of medium or long term ripple.
I test ripple in a few difference ways, first I test it during the cold load testing. It is tested at zero load and maximum load first. During the hot load testing I test the ripple at maximum load again. I have recently started testing ripple at fairly random loads with the unit still hot, it’s a bit unorthodox (a bit? maybe a lot) but has found issues in the past that did not show up with other test methods.
The previously mentioned Over-Temp-Protection issue made getting ripple results in hot conditions essentially impossible, as taking the pictures and double checking the settings takes long enough for the unit to shut down. It’s a five minute process to re-start it and I figured I’d best quit before I threw it into the bay in frustration. To make matters even more dubious I managed to break the leg off the load tester 5 V rail’s filter caps, without these caps I cannot test ripple to spec on that rail. Thus the ripple section shorter than usual!
We’ll start off with all three rails with zero load, in all three pictures the scope is set to 10 mV and 10 ms.
70 mV is more than I really like to see at zero load on the 12 V rail. It’s well within spec, but it makes me wonder what we’ll see at full load.
Speaking of full load, here it is! As suspected the 12 V ripple got worse, for that shot the scope is as 20 mV/10 ms. The 3.3 V rail shot has the earlier 10 mV/10 ms setting. The 5 V does not exist due to broken bits on my tester.
The 12 V rail did get noisier, at 140 mV it is over the maximum allowed by the spec by 20 mV. Will this cause you issues? No. Will it cause the unit issues in the final analysis? Yes. The 3.3 V rail looks great. The 5 V rail I can’t in good conscience put a picture of it up without the smoothing cap. To my eye having seen exactly what the caps change, the 5 V rail would have looked quite similar to the 3.3 V rail. Even without the cap ripple it was below spec. (As a side note, the spec calls for the caps on the tester).
Hot results are similarly picture-free due the the unit shutting down after a few minutes at full load and 40 °C intake. Before shutdown they look nearly identical to the cold ripple pictures, the only part of the unit that seems to care about the heat is the OTP circuit. It makes me wonder if Cooler Master has had the perfect storm of tolerances stacking up in the OTP bits or something.
Disclaimer: Power supplies can have dangerous voltages inside them even after being unplugged, DO NOT OPEN POWER SUPPLIES. It’s just not a good idea, and doing so could very well kill you. Don’t try this at home. Don’t try this at work. Just don’t do it.
First up, fan and overview.
The fan’s sticker is strangely informative, telling us not only that it only draws three watts (that’s it? In a 1500 W unit?) but also that Cooler Master paid for it and Young Lin Tech built it. The unit itself looks to be an Enhance design, though not one I’ve seen before. This doesn’t surprise me as not many companies make units this large and Enhance units over 1kw have a reputation for shutting down when faced with high intake temperatures. Looking at the location of the copper secondary rectification heatsink (with two thermistors attached via heatshrink, one for fan control and one for OTP) I am not at all surprised that the unit shut down. Most of the heatsink isn’t exposed to direct airflow from the fan and/or the air from the fan has nowhere to go after going over the heatsink. The only effective part is the part of the heatsink to the left of the transformer, that’s not much heatsink for 1500 W of rectification.
From this point we’ll tour the unit starting at the receptacle and following operations through it till we get to the output bits.
As usual part of the transient filter is attached to the receptacle, two X capacitors and two Y capacitors are chilling here.
On the PCB we find a huge 20 amp fuse, two thermistors, two inductors, two X capacitors, two Y capacitors and a TVS diode. That’s a good transient filter right there.
Sharp eyes will notice the lack of wires leaving the switch, that is because it is nearly impossible to get a good picture with the case and the PCB still attached to each other. Usually I de-solder the PCB end of the wires, but in this case I didn’t think I’d be able to get them re-soldered afterwards. I also notice that the tech who did the hand soldering bits had issues with them on both ends:
Obviously neither issue cause problems as the unit already passed testing by this point, but neither solder point looks very good. At the PCB end the wire got too hot and melted the insulation off, while at the switch end it wasn’t hot enough for the solder to flow fully. So it goes I guess, I’m not impressed.
Next up is the APFC unit, it takes the incoming AC voltage, rectifies it to DC and then boosts the DC to ~380 V, then stores it in no fewer than four big capacitors.
Unfortunately the APFC switches and diode are behind those capacitors. The switches consist of three 47N60C3 MOSFETs (650 V, 47 A @ 25c, 30 A @ 100c). The diode is perfectly centered behind a capacitor and eluded my attentions.
Once boosted and stored, a pair of 35N60C3 MOSFETs (650 V, 35 A @25c 22 A @ 100c) act as the main switches to send it through the (huge) transformer. In charge of both operations is a CM6802 manager IC, it shares a second PCB with the PS232S protections IC.
The protections IC supports up to four 12 V rails for OCP/OVP/UVP as well as both 5 V and 3.3 V rails. It also has an auxiliary input that can be used for OTP. I’ve never been impressed with the trip points for the OVP/UVP on the PS232 series. Below is the trip point chart from the PS232S datasheet.
I would like my power supply to decide something is wrong before the 12 V rail hits 9.5 volts, thank you very much.
Once the power clears the transformer is it rectified by seven MOSFETs, four 023N04N (40 V 90 A) and three 015N04N (40 V, 120 A). They share the copper heatsink. The heatsink is also used to connect two of the MOSFET’s output to the output of the other five. Looking at the ratings for these MOSFETs I find it surprising that Enhance would set the OTP low enough to trip, the MOSFETs do not start start to de-rate amperage wise until just before and just after 150c respectively. Only one of them, a 023N04N, was able to be captured with a camera.
Once rectified the 12 V is smoothed by a small stand of not-very-small Nippon Chemi-con 105c rated capacitors and some of it is passed onward to the 5 V and 3.3 V DC-DC regulation boards.
The DC-DC converters are essentially identical, both sport two 95N3L MOSFETs (30 V, 80 A@25c 61 A@100c) and two that are very poorly marked if marked at all. Both get a pretty red air core inductor to clean things up a bit, as well as four polymer and one electrolytic cap each.
We’re almost done with this side of the main PCB now, before we are though here are a couple overview shots.
The first picture is more of a side view than an overview, you’re just going to have to deal with it. The shielding between 12 V filter inductor and output wires is a nice touch.
Now that we have 12 V, 5 V and 3.3 V; some of it heads out of the case on the hardwired cables, the rest goes to the modular output board.
We get some extra electrolytic capacitors from Suncon to re-clean the various rails. On the rear of the PCB we see good soldering and spaces for a lot more capacitors. Please note that the standoff mounts are both attached with nuts AND soldered. This is perfect in my opinion.
On the note of soldering, let’s look at the main PCB too.
Both good and bad are to be found under the PCB. Good is a set of current sense resistors, there really are two 12 V rails here. Bad is a very strange looking solder joint. There are a few bits of messy hand soldering and quite a bit of rather nice machine soldering.
I was unable to find a UL number on the PCB. The case has a number (e320127) that traces to Cooler Master. There is a Cooler Master logo on the PCB for whatever that is worth. I’m sticking to my guess of Enhance as the designer however.
Final Thoughts and Conclusion
There are a couple elephants in the room here:
- The first is that very few people actually need a 1500 W PSU, that puts it in real danger of being considered a niche product. I’m going to let it off the hook somewhat on that front, mostly due to the fact that almost all X79 motherboards support 4x GPU configurations. If you’re running four top-end cards and overclocking, you really do need a massive PSU. The fact that a decent number of people build quad-SLI and quad-Crossfire boxes keeps this unit out of the niche category for me. The question of whether anybody needs four GPUs is an entirely different ball of wax.
- It shuts down when asked to deliver its rated power at 40 °C, its rated maximum temperature. What this means is that if you have a big stack of big GPUs and a big CPU to go with it, you had best have big case airflow as well, or a case that allows the PSU to gather cold air from outside. The fact that it shuts down rather than going into thermal runaway and exploding is a plus, but it really shouldn’t be shutting down to begin with. It is my opinion that the primary cause of the shutdowns is a combination of a relatively low powered fan and the placement of the secondary heatsink where it doesn’t get much airflow. An alternative possibility is that the PS232S, thermistor and resistors that make up the OTP circuit were all on, or close to the accepted variation in resistance and trip points. If you look at the PS232S datasheet you can see that it has a rather wide variation in trip points. Unless each PS232S chip is tested individually, and the resistors on the board are sized to calibrate it (expensive!), every one of these PSUs will shut down at a different temperature. Either way, it’s an issue.
There are good things too though, like the fact that it really can put out 1500w for instance.
I also am very pleased to see a C19 plug on the unit, nice and safe.
The modular cables are very nice to work with, the hardwired cables look great, though it would be nice if the last bit was sleeved. They’re also quite long, you won’t have issues with length even in a huge case like the Cosmos 2.
The fan is nice and quiet, possibly too quiet as we talked about above.
With an MSRP of $400 this unit is far from inexpensive. Actually, it’s possible that this should go up in the Real Issues section, as there are two other units out there in this wattage range for much less money. A Silverstone unit on another Enhance 80+ Silver platform for $300 and a Thermaltake 80+Gold 1475 W unit that runs ~$350. All three units do share a $400 MSRP, but the Cooler Master M2 1500 W sells for $360. Then of course there is the MAXREVO 1500 W, which gives Gold efficiency and runs $370.
The build quality is decent. All the capacitors are Japanese. The machine soldering is excellent, but the hand soldering is dubious at best. Everything is attached at least, even if it isn’t pretty.
I’d like to see more SATA power plugs, 12 is a lot but I’ve seen (and reviewed) 750 W units with 12 SATA plugs. It isn’t like this unit couldn’t power more devices!
The voltage regulation is very good, I’m quite happy with that.
I’ve never liked the PS232 series protections ICs, their trip points vary wildly and tend to be uselessly low, and almost as uselessly high.
The ripple control I am not happy with, as it violates spec on the 12 V rail at full load. It’s not a major issue, but it is an issue. The other rails are well within spec.
The looks of the unit itself are nice, it’s all matte black with understated (painted on!) logos. Of course it also resembles an over sized black brick, but there isn’t much to be done about that.
Including an eight pin to four pin CPU power adapter is a nice touch.
All told there are some nice pros:
- Really can put out 1500 W (at normal ambient temperature, anyway).
- Very quiet fan.
- All the cables are nice and long and easy to work with.
- Plenty of CPU and PCIe power plugs.
- Great voltage regulation.
It isn’t without its cons though:
- Shuts down after a few minutes of full load at the rated 40c intake temp.
- 12 V ripple breaks spec by 20 mV (16%).
- More expensive than the competition.
- Dubious hand soldering is to be found.
All told the unit is decent but the price is too high for the performance (and shutdowns hot…), if Cooler Master drops prices to the low $300 range then it isn’t a bad choice at all. As it stands at $360 I cannot in good conscious give it an approved stamp, as there are better options for less money.