That’s right folks it’s time for another Thermaltake power supply review! This time it is the top unit of Thermaltake’s Smart series at 1000 W of rated output and modular to boot! Sounds good in theory, let’s check out reality.
Did I say reality? I meant marketing, in the form of the…
Features and Specifications
Features and Specifications come to us today directly from ThermaltakeUSA.com’s product page for the Smart-M 1000 W.
- TX 12V V2.3 & EPS 12V 2.92 enables most reliable and robust power delivery.
- Guaranteed to deliver 1000W continuous output at 40℃.
- Ultra-quiet 135mm cooling fan delivers excellent airflow at an exceptionally low noise level.
- Universal AC input 100V~240V automatically scans and detects the correct voltage for different country.
- 99% Active Power Factor Correction provides clean and reliable power to your system.
- 80 PLUS® Bronze Certified: 82% or more efficiency at 20%, 50%, and 100% load.
- Robust and dedicated dual +12V output provides superior performance under all types of system loading.
- Double forward ultra-efficient circuitry design for added power savings.
- 100% High quality 105℃ Japanese capacitors provide uncompromised performance and reliability.
- Cable Management improves internal airflow by reducing cable clutter within PC to promote accelerated heat removal.
- Flat, low-profile modular cables allow for easy cable routing and increased airflow.
- Supports multiple core high-end graphic cards & CPU. (NVIDIA SLI & ATI CrossFire Ready)
- Dimension: 150mm(W)x86mm(H)x160mm(D).
- High reliability: MTBF>100,000 hours.
- Built-in industry grade protections : Over Power, Over Voltage, Under Voltage, Over Current and Short-Circuit protection.
- Safety / EMI : UL, CUL, TUV, FCC, CE certification.
My thoughts? I have a number of them, many are even on the subject at hand. Feel free to skip the following block of text if you’d like, nothing crucial is contained in it.
I think that a 40c rating for full output power is good, nothing special per-say, but good. I’ll believe the Ultra-quiet fan claim when I (don’t) hear it. I will say this: I like that Thermaltake didn’t say ultra-silent. Silence is absolute, you cannot have Extra Silent any more than you can have infinity+1. So thank you Thermaltake on that front. Universal input comes with APFC, nothing special there these days, though it is nice. 99% APFC is almost assumed these days, if your APFC bits cannot manage 99% they’re doing it wrong. 80+ Bronze is an excellent start, though these days it isn’t as snazzy as it used to be given all the gold units floating around. Same for Double-Forward switching, it’s ultra-efficient compared to older flavors perhaps, but compared to modern LLC resonant stuff? Not so much. 100% 105c Japanese capacitors is something I can get behind. They’d better believe I’ll be checking this though! (Modular) Cable management and flat cables are two things I like a lot. If a 1kw unit didn’t support SLI/CFX I would be shocked! A MTBF rating of >100,000 hours means that, in theory, you’ll get over 11 years out of this thing. Lastly we have claims of various protections and safety certifications. I’ll be checking the supervisor IC to make sure it can do what they say it can, though given that the second 12 V rail (assuming there really are two!) is 80 amps, SCP/OCP may be sort of pointless. I’m tempted to bridge some connectors and check for SCP.
Whew, that was a larger number of thoughts than I thought I had, let’s move on to connectors.
Note that the motherboard connector is not a 20+4 pin, but a single piece 24 pin connector. The picture of the connector is inaccurate. Actually the whole list is amusingly enough. It doesn’t show the second CPU power connector. Oh well. The text version is more accurate:
Connectors & Cable length
1 x 24pin Main connector (600mm)
EPS 12V 8 pin
1 x EPS12V 8pin connector (600mm)
ATX 12V 4x4pin
1 x ATX 12V 4+4pin connector (600mm)
3 x Peripheral connectors (500mm + 150mm + 150mm)
|3 x S-ATA connectors (500mm + 150mm + 150mm + 150mm )|
3 x S-ATA connectors (500mm + 150mm + 150mm + 150mm )
3 x S-ATA connectors (500mm + 150mm + 150mm + 150mm )
3 x S-ATA connectors (500mm + 150mm + 150mm + 150mm )
|2 x PCI-E 6+2pin connector (600mm + 150mm)|
2 x PCI-E 6+2pin connector (600mm + 150mm)
2 x PCI-E 6+2pin connector (600mm + 150mm)
PATA to FDD adaoter
|1 x PATA to FDD adapter (150mm)|
This is the first time I’ve seen the standard “Molex” connector (a poor name for it really, Molex is the company that holds the patent. They make a lot of other connectors too) referred to as a PATA connector. It’s probably more accurate than calling it a Molex though.
Intel ATX 12V 2.3 & EPS 12V 2.92
Max. Output Capacity
Peak Output Capacity
150mm x 86mm x 160mm
Power Good Signal
16msec (minimum) of full load at 115Vac/ 230Vac input
|Input Frequency Range|
50 Hz – 60 Hz
100V – 240V
0℃ to +40℃
20% to 90%,non-condensing
-40 ℃ to + 70 ℃
20% to 95%, non-condensing
13.5cm two ball fan, 1200W: 1800±10%R.P.M.
82-88% efficiency @ 20-100% load
100,000 hours minimum
UL, CUL, TUV, FCC, CE certification
One last chart, I promise this is the last for a while. The output table:
|Input Voltage: 100V-240V|
Input Current: 13A
|Max Output Current|
|Max Output Power|
900 watts divided by 12 volts gives us a maximum 12 V combined load of 75 amps, yet 12V2 is rated at 80 amps? Sometimes I wonder about people in marketing. I guess this unit can put out more on 12v rail 2 than it can put out overall. If that doesn’t make sense, blame Tt, not me.
Enough charting, how about some pictures of the unit and its packaging? That sounds more appealing, visually at least.
Photos Part 1, Packaging
The box looks much like the previous Smart series unit, though the decorations are blue rather than green. I like the look of the box, the blue looks better than green to me. Many of the marketing claims are given some face time on the box, including a shot of the internals that appears to show individually regulated output rails. 100% Japanese capacitors are called out again, I hope it’s true for Thermaltake’s sake.
As you can see above, there isn’t as much protection as I’d like for the PSU inside the box. Just bubble wrap on three sides, foam and bubble wrap on the 4th (bottom) with the 5th and 6th sides protected by bubble wrap and cables. Still, it arrived in good condition despite being shipped twice. The packaging seems to work well enough despite being on the skimpy side, but if it shows up on your doorstep having been shipped without an extra box and packing materials around its box, I would return it and demand a better shipping job. Especially if the box is crunched.
How about some accessories?
Why the included mounting screws are so long is beyond me, there is nothing in the PSU design that requires them to be any longer than normal. The power cord is a nice beefy 14 gauge affair and you get a number of cable ties to hold cables where you want them. Not exactly a large accessory pack, but not bad.
Why oh why must companies hard wire two CPU power connectors? Very few motherboards support multiple connectors, while on the other hand the vast majority of high power units will need a least one PCIe cable. I dislike this. The 24P motherboard power connector is an interesting choice, I think I like it. I haven’t seen a motherboard worthy of a 1000 W PSU that didn’t have a 24P connector in a very long time indeed, if I even ever have! This means no fiddling around trying to get the +4P bit to stay on the main connector. Similarly, I like that you get a straight up 8P EPS CPU power connector, trying to hold those together while in cramped quarters behind a very sharp heatsink is a major pain. I’d prefer it was modular, though. Also, what’s up with the Molex-FDD adapter? No sleeving, no ribbon cable, at least the plugs are black.
What, you want to see the actual PSU? Tough! ……….. Okay, I can’t hold out on my fans for very long I guess. Here you go:
Pretty much the same styling as the Smart 530 W I reviewed recently, but with blue lines instead of green. I like the blue better, it looks great. The modular connector bits are well designed and well labeled. Even if you want to, you cannot plug anything in where it does not belong. This is good, as plugging things in incorrectly is an excellent way to destroy hardware, especially on a nice juicy unit like this. The fan looks mean and delicious, and the grill looks quite good too. As we’ll see in a bit, that center section is flatter and shinier than some heatsink bases.
Before we head into testing, a few more shots of the unit from various angles so you can get a good feel for how it looks if you don’t spend all of your time at right angles to your case. (Hint: It looks like a black box)
(As a side note: While taking pictures on a lawn that has recently been rained on makes things look wonderful, wetness and power supplies really don’t go together well. Keep your power supplies out of the rain, or at least make sure they’re 100% dry before you plug them in, let alone turn them on. Not doing so could be rather exciting.
Other than the warranty sticker coming out of the box most of the way torn, the unit looks great. I like the look of the normal Smart series and I think the switch to blue lines for this unit is an excellent choice. The warranty sticker is 80% of the way to being broken, something of an issue for the normal user if the unit later fails. The likelihood of such a failure we will assess in the next section.
Load Testing Part 1: Ambient Temperature
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. For this reason, 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.
With the block quote out of the way, know that I will be testing this PSU at both outdoor ambient temperatures (typically between 14 °C and 18 °C here this time of year) as well as in the Enclosure of Unreasonable Warmth. TEUW is a precision engineered enclosure that I use to route the exhaust air from the PSU right back into the intake fan, it is adjustable to hold the intake air temperature at (almost) any level I want it. This way I can test the PSU’s response to hot conditions as well as cold conditions. For the hot testing I will be running the intake temp as close to the unit’s maximum rated temperature (40 °C) as possible. TEUW, in case you’re curious, is a cardboard box.
Right now it is time for cold testing, some units actually perform worse in low ambient temperatures than at their maximum rated temperature. If you like your AC and have a case with good airflow your PSU will be running in conditions similar to my cold testing. The cold testing done during summer at least.
This appears to be an independently regulated PSU. This is better than a group regulated, but generally not as good as a DC-DC unit when it comes to cross loading and general testing rudeness. For this reason (and due to the minimum step sizes imposed by my testing equipment), I will be fairly rude about testing this unit.
Here are the results of cold testing! Yes, I did over load the unit by 0.8 % and the 12 V rail by 1.33 %. If the unit cannot take this low of an overload it is rated far too close to it’s maximum for my tastes. Plus, it does say 1200 W peak on the webpage spec sheet.
|Load Wattage||12 V Rail||5 V Rail||3.3 V Rail||Kill-A-Watt|
|912/0/0w (12V CL)||12.17||5.12||3.37||1085|
This is really quite good, I’m impressed. The 12 V rail comes in at 0.9%, the 5 V rail at 1.3% and the 3.3 V rail at 2.3%. The 3.3 V rail regulation could be better, but it’s well within spec. The 5 V rail is good and the 12 V rail is excellent. This is definitely an individually regulated unit, I have yet to see a group regulated unit that could stay anywhere close to spec given this sort of 12 V cross load. During testing the fan wasn’t bad at all, it made some noise while the unit was going full blast but nothing obnoxious or out of the ordinary.
Load Testing Part 2: The Enclosure of Unreasonable Warmth
Into TEUW goes the SP1000-M (I like that model number, it appeals to me for some reason)! Full load and the 12 V cross load are re-tested hot. Here are the results:
|Load Wattage||12 V Rail||5 V Rail||3.3 V Rail||Input Watts|
|912/0/0w (12V CL)||12.15||5.11||3.36||1088|
All three rails lost 10-40mV, but regulation is still good and well within spec. No issues here at all. The fan got louder, but still very reasonable.
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.
We’ll start with zero load. Zero load is a rude test. Failing zero load does not mean it is a bad PSU as zero load will never happen in real life inside a computer. I tend to aim for the brutal however, so here come the zero load results. In the following three shots the scope is at 20 mV / 10 ms.
Well that’s disappointing. The 12 V rail is 12 mV over spec, that’s not that bad. The 5 V rail is almost three times the spec while the 3.3 V is “only” twice the spec. I am not impressed. Yes, the 3.3 V picture is almost as bad as the ripple.
Onward to full load! Still cold conditions. I’ll give you a hint towards the results, you can get an idea of them by the scope settings. The scope is set to 50 mV / 10 µs until noted otherwise.
12 V ripple at full load is now 45 mV over spec, I’m not impressed. The 5 V has dropped to only almost double spec while 3.3 V has dripped a little bit as well. None of them are anywhere close to within spec however. Are they bad enough to actually cause problems? I doubt it. Are these results putting the Approved stamp in jeopardy? Yes.
On to the cross load results, still cold, scope still at 50 mV / 10 µs.
The cross load didn’t help matters much. Maybe some heat will, it does sometimes. Into TEUW it goes. Scope still at the same settings, full power load.
Being hot did help, but not enough to get within spec. 12 V is close at least, I’ll give it that. We’ll try the cross load and see how things go. Still the same scope settings.
Hot with a cross load is pretty similar to hot at full load. Not good. The most exciting part is next though, during the Hunt At Random Loads section I discovered that while hot after running hot for a while, any load between ~200 W and ~750 W resulted in the following ripple levels. For all three shots the scope is set to 100 mV / 200 µs.
Now that is an issue. That’s just shy of four times the spec on the 12 V rail, nine times the spec on 5 V and seven times the spec on 3.3 V. While this was happening the PSU emitted a nasty whining sound and the power factor dropped substantially. Previously it was staying happily in the 0.99 to 0.97 range, in this mode it was 0.82 to 0.85. Once the unit cooled off the problem went away, it also went away under 200 W and above 800 W.
We contacted Thermaltake regarding this issue, they expressed concern and discontent with their manufacturing partner, but never sent us a second unit to test, nor did they contact us with any updates. Official policy is to allow a second unit in the case of problems, Thermaltake seems to have chosen not to take us up on that. Thus this unit will be the unit receiving the final judgement.
Perhaps there will be something obviously wrong inside the unit that I can point to and say AHA. Given the above data I expect it’s somewhere in the APFC bits.
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.
You have my assurances that I know what I’m up to and won’t kill myself during the course of this review. Admittedly, if I did kill myself opening the PSU, I expect I’d find it difficult to write this, so I must not have. Amusingly that proof is one of the stronger proofs of physics/astrophysics. That has little to do with this review however.
Upon opening the PSU these are the sights I was greeted with, a decent looking PSU and a fan hub.
I don’t recognize the layout, but it looks pretty good. Nothing is obviously damaged or defective. We’ll tour it and look at things more closely, starting with the transient filter.
The transient filter is a good setup, it’s not spectacularly large or anything but it does have all the required parts. Four Y caps, two X caps, two inductors, a fuse and a TVS diode. There’s another X capacitor after the rectifier to mop up the diode switching noise. So far everything looks good.
Next up is the rectifier and the APFC bits.
The bridge rectifier is a GBJ2506, good for 25 A and 600 V. The two switches are our old friends, 24N60C3 MOSFETs rated for 24 A at 25 °C and 15 A at 100 °C, 650 volts either way. For boost diode purposes we have a pair of BYC10600 diodes rated at 10 A and 600 V each. Controlling the operation is a CM6800TX brainbox, it does the PWM duties as well. The storage capacitor is a Panasonic unit rated at 105 °C. So far so good for design and parts wise.
For switches we have two more 24N60C3 MOSFETs (same ratings as the APFC bits, unsurprisingly). The 12 V rail is rectified by four KCQ60A06 schottky’s rated at 60 A and 60 V (and weighing 5.5 grams, according to the datasheet). The 5 V and 3.3 V rails each have two 30L45CT schottkys rated at 30 A and 45 V. The protections IS is a PS224, which has more reasonable OVP/UVP trip points than the PS232 we often see. It has an auxiliary input that can be used for OTP as well.
All told the design really looks quite good. It’s not the most modern in the world, but there are a fair number of filter capacitors, the rails are overbuilt, all the capacitors are Japanese (all of ’em!) 105 °C rated just like the box says. I was unable to find any reason for the wild ripple results while hot. The soldering even looks good:
The main PCB soldering has a few leads that could do with more solder, but no messy bits. It’s really quite good. The modular PCB is excellent soldering. Looking at the bottom of the board we see four shunt resistors for OCP sensing, one only feeds the protections IC (through a hand soldered jumper wire, not so pretty but it is well soldered), two of them are both going to the same rail and feed the modular board, while the last one feeds the hardwired cables. It looks promising for really having OCP.
The PCB has a UL number on it that belongs to HUNG HING ELECTRONICS CO LTD (UL# e327405). The PSU’s case has a Thermaltake UL number (e303666). Who Hung Hing is I don’t know. I’ve seen the heatsink design before on Sirtec units, for whatever that is worth. They may simply both buy heatsinks from the same company though.
I can find nothing wrong internally, no funny smells, no charred bits, no cracked joints, nothing. It looks like a solid PSU.
Final Thoughts and Conclusion
Thermaltake has what looked to be a good unit here. Even through the regulation/load testing phase things looked quite good. Once we hit the ripple testing the picture wasn’t as pretty and then came the issue. The ripple results were bad before the Random Hunting section and absolutely terrible at that point. 350-440 mV of ripple is far more than what is acceptable. This is a major issue.
The unit looks quite good, all the hardwired cables are sleeved, the modular cables are either sleeved or of the flat ribbon style. I really like the ribbon type cables. The marketing claims stood up pretty well, all the capacitors really are Japanese and 105 °C rated and it really can cough up a kilowatt.
The regulation was decent for the 5 V and 3.3 V rails and excellent on the 12 V, this unit did quite well there.
The internals and design look good. It’s a double-forward design with schottkys for regulation, but there is nothing wrong with that if the price is right.
Speaking of price, at $174 (the lowest I could find from a known retailer online) isn’t a very good price for a 1000 W 80+ bronze unit, modular or otherwise. There are cheaper bronze units out there, and Newegg has 80+ gold 1000 W modular units for ~$180.
I’ll summarize a bit:
There are pros:
- Good voltage regulation.
- Only contains 105c rated Japanese capacitors.
- Solder is good.
- Looks great.
There are cons too:
- Ripple well over spec at full load, cold or hot.
- Price is not very good.
- Hot ripple at mid range loads is absolutely terrible.
If the price were the only issue this unit would get an approved stamp, it isn’t that overpriced. Given the ripple results however, and especially the ripple results in hot conditions with a mid range load, I have no choice but to fail the unit. This much ripple is simply unacceptable.