Zalman opened up shop in 1999 with a focus on bringing peace to the computer world - in the form of silent computing. Today we have Zalman's ZM750-HP PSU on deck. Is it all that and a bag of chips? We're gonna find out!
Internals
Now, you really shouldn't open a PSU. It voids the warranty and it can be dangerous. But this one has HEAT PIPES!
NOTE: Opening up a PSU not only voids your warranty, but also presents a serious health risk. Even after turning off and unplugging a unit, there can be enough charge stored to fatally shock anyone poking around inside. Regardless of what you may have heard about “draining” circuitry and other methods of discharging a PSU, these are simply risk reducers – not risk eliminators. In short, let us at HardwareLogic be the ones to stand barefoot on a wet copper plate while snapping shots of the internal componentry, and keep the cover on at home.
We open up heat pipes to get a feel for the build quality and see if anything is obviously odd - or special. The initial view on opening reveals a few heat sinks/spreaders, a single primary capacitor, and HEAT PIPES! Like any modular unit, there is a second PCB that is used as the header for the modular connection system - and also contains the three blue LEDs. The fuse is soldered in and heat shrink wrapped so if it blows it keeps the damage to itself. Soldering work appears clean in the unit as well.
(
The main capacitor in this unit is an OST (or Oster) out of Taiwan. It's an SPS series 420V, 470uF cap rated at 85°C. There are better capacitors out there (in fact OST even offers some) and we would have preferred to see a Rubycon, Hitachi, or Chemi-con. However, with a 3-year warranty, Zalman seems comfortable with the choice and OST has enough pride in this product to not only put thier name on it, but list the specifications publically and openly on thier website. Other capacitors on the board are Teapo and CapXon also out of Taiwan. Hmm, seems to be a bit of a Taiwan theme here.
Keeping with the Taiwan theme, the mainboard in the unit was printed by the Chian You Company of Taiwan. The board is also solely designed for the ZM750-HP and not generalized for a range of PSUs like many are. As far as who actually built the unit, we were somehow guessing it might be a Taiwan based company. We weren't sure why we would jump to such an unfounded conclusion, but we did. There are no obvious UL numbers that point to anyone but Zalman and Chian You, but the little "SPI" printed on the large transformer kinda gave up the secret. If you don't recognize SPI, you might recognize thier subsidiary - FSP (aka Fortron or Sparkle). FSP built or no - we're pretty sure Zalman gets the conceptual design credit. A quick browse of FSP didn't come up with any heat pipe units.
The fan in the ZM750-HP is an ADDA AD1212MB-A71GL. ADDA hails out of...um...Taiwan, and is a popular choice in power supply fans. This model is a is 120mm, 12V, medium speed ball bearing fan. At a full tilt of 2050 RPM, this fan will push 80.5 CFM at 38db using ~4 watts. That's a good amount of air and the maximum noise would be less than someone talking very quietly from about 3 feet away. It's a good fan choice, although we hope we won't be hearing much from it.
Performance
Our performance tests are done with a load testing system. This system allows us to configure a series of known loads to put on a PSU. More information on our PSU testing methods and philosophy can be found here.
Test Setup
| Test |
5V
Rail |
3.3V
Rail |
12V1
Rail |
12V2
Rail |
12V3
Rail |
12V4
Rail |
-12V
Rail |
5V SB
Rail |
3.3V+5V
Watt. (Est.) |
12V Total
Watt. (Est.) |
Total
Wattage |
| 1 |
4.8 |
4.8 |
3.6 |
3.6 |
3.6 |
3.6 |
0.00 |
0.83 |
39.8 |
173 |
217 |
| 2 |
12.0 |
12.0 |
9.0 |
9.0 |
9.0 |
9.0 |
0.26 |
0.83 |
99.6 |
432 |
539 |
| 3 |
15.3 |
15.3 |
10.2 |
10.2 |
12.0 |
12.0 |
0.80 |
0.83 |
127 |
533 |
674 |
| 4 |
19.2 |
19.2 |
10.2 |
10.2 |
13.8 |
13.8 |
0.80 |
0.83 |
159 |
576 |
749 |
| 5 |
3.0 |
3.0 |
10.2 |
10.2 |
13.8 |
13.8 |
0.80 |
0.83 |
24.9 |
576 |
615 |
| 6 |
13.5 |
13.5 |
6.0 |
6.0 |
6.0 |
6.0 |
0.80 |
0.83 |
112 |
288 |
414 |
Test Results:
*Efficiency results include wattage from the -12V and 5V SB rails.
| Test |
5V |
3.3V |
12V1 |
12V2 |
12V3 |
12V4 |
-12V |
5V SB |
T in (°C) |
T in (°C) |
Watt in |
Watt out |
Eff. % |
| 1 |
5.14 |
3.32 |
12.18 |
12.18 |
12.16 |
12.15 |
12.44 |
5.15 |
25.7 |
30.3 |
259 |
220 |
85.0 |
| 2 |
5.09 |
3.30 |
12.13 |
12.16 |
12.12 |
12.05 |
12.41 |
5.09 |
25.8 |
39.2 |
641 |
544 |
84.9 |
| 3 |
5.05 |
3.29 |
12.12 |
12.16 |
12.10 |
12.06 |
12.18 |
5.07 |
26.3 |
43.9 |
815 |
679 |
83.3 |
| 4 |
5.00 |
3.28 |
12.15 |
12.19 |
12.16 |
12.07 |
12.25 |
5.03 |
27.0 |
50.2 |
924 |
756 |
81.8 |
| 5 |
5.23 |
3.31 |
11.91 |
11.91 |
11.90 |
11.81 |
11.97 |
5.23 |
27.3 |
42.3 |
719 |
610 |
84.8 |
| 6 |
5.04 |
3.30 |
12.25 |
12.27 |
12.22 |
12.22 |
12.06 |
5.06 |
27.5 |
39.2 |
494 |
420 |
85.1 |
Our first four load tests are proportional in nature. In Test 1, we see a PSU that is fully in spec and has voltages set not too far above the assigned voltage. Test 1 is our "light" load test, and the ZM750-HP showed it can handle it. In Test 2, we up the ante to a "typical" load. Impressive to note, is that the change in voltages between Tests 1 and 2 is less than 1% for all but the 5VSB rail. That's excellent!
In Tests 3 and 4 we ramp up the load to simulate "Heavy" and "Max" conditions. We should note that on PSUs this size, we introduce some cross loading on the 12V rails. While not strictly "proportional", it is a tad more realistic and a tad harder on the PSU. We can see in Test 3 that the voltage remained rock solid on the main rails. In fact, the main rails between Test 1 and 3 all keep to less than 1% voltage except for the 5V - which still showed a very good deviation of <2%. In Test 4 we really put the hammer down on the ZM750-HP. Notice our total wattage output is slightly above the maximum rating for this unit! The ZM750-HP takes the load in stride. We see some deviation on the 5V and 3.3V rails overall, but it's still very good. The minor rails move around a bit, but still well within spec. What really got our attention was the rock solid behavior of the 12V main rails. Overall, they were still showing a voltage deviation of less than 1%! Wow!
In Tests 5 and 6 we introduce extreme crossloads. Test 5 is an extreme (but plausible) example of having very heavy loads on the 12V rails while the other rails see light loads. We have come to find that this is where we really put the hurt on a PSU. The ZM750-HP certainly showed it was under stress, but held its ground. The 5V rail popped up 4.6% (still good) and the 3.3V rail moved up just a few hundredths of a volt under the sudden lighter rail load. The main 12V rails all took a hit, but all remian at <3% deviation overall - great showing! The -12V rail took a hit, but we nearly always see this in Test 5.
Moving into Test 6, we basically reverse the crossload - just for good measure. As you might expect, the 5V and 3.3V rails drop under the sudden heavier load, while the 12V rails recover. Overall, the ZM750-HP showed very good performance. The worst deviation we saw on the main rails was 4.6% on the 5V. This is still pretty good. The 12V rails really danced on the move from Test 5 to Test 6, but all held to less than 4%. Everything was well within spec. Good show Zalman.
As for noise and cooling, the Zalman unit was unheard in Tests 1, and 2. In Test 3 we barely heard the fan spinning. In Test 4, the delta T approached 25°C and the fan spun up to max. You could certainly hear it. Ninety seconds into Test 5 the controller throttled back on the fan and that was about the last we heard from it. Test 4 is a condition you should rarely be sitting at for more than a few seconds, so we doubt you will hear too much from this unit. What about the high delta T's? After testing we realized that heat pipes are often partially gravity driven. We actually test PSUs in an orientation that is "upside down" from how they are typically installed. We are not sure the orientation actually matters, but if it does it just mean that one can expect even better noise/heat performance in a typical install.
Efficiency is also important to us. Energy costs are rising and many of us are debating how our often wasteful lifestyles are impacting our planet. We are happy to report that, at least in our test suite, the ZM750-HP has some of the better efficiency we have seen. We pulled an average of 84.1%. We have to admit, that surprised us!
