proSilence PCS-350 Fanless PSU

Power
Viewing page 4 of 4 pages. Previous 1 2 3 4

TEST RESULTS

Measurements were made at 5 output power levels: 65W, 90W, 150W, and 170W. The PSU was allowed to run for ~10 minutes at each power level before measurements were recorded. The room temperature was 23C.

Table A. Load on the PSU

LOAD
65W
90W
150W
170W
+12V
24
36
60
72
+5V
20
20
40
50
+3.3V
16.5
26.4
39
39
-12V
2.4
3.6
3.6
3.6
-5V
1
2
2
2
+5VSR
1
2
2
4

Table B. On test bench, in 23C ambient temperature

AC Power
95W
127W
205W
230W
Internal Heat
30W
37W
55W
60W
Efficiency
68%
71%
73%
74%
Power Factor
0.62
0.64
0.65
0.65
Max Internal Temp*
39C
50C
58C
64C
Internal HS Temp*
29C
37C
41C
45C
External HS Temp*
27C
35C
39C
42C

*Please see text under COOLING for full details

1. VOLTAGE REGULATION was very good, within -/+3% on all lines in any combination of loads tried (somewhat at random). It was often within -/+2%. The low and high voltage seen on each of the main lines is shown:

  • +12V: 11.66 to 12.43
  • +5V: 4.75 to 5.28
  • +3.3V: 3.28 to 3.4

It should be noted that I have no way of testing line regulation, so AC conditions are steady-state, not dynamic as it would be (potentially) in a real PC; I have no way to vary input AC voltage at this time. The AC line voltage in the lab as measured by the Kill-a-Watt power meter is usually within a couple of volts of 120V.

2. EFFICIENCY was very good, measuring a relatively high 68% even at 65W, where most PSUs fall to 65% or below. By 90W output, it hits 71%, and improves to 73-74% at the highest measured levels. This is very good performance, especially at the sub-100W levels. Most desktop PC systems actually operate between 65-150W so the proSilence is well optimized for real world application.

3. POWER FACTOR was only modest.

4. NOISE was nonexistent. Even at the highest output power levels, I had to press my ear right up against the top of the unit to barely hear just a trace of electronic noise, but the level was so low that it is easily dismissed. The proSilence 350 really is silent.

5. COOLING: TEMPERATURES were measured because it is a question mark hanging over any fanless PSU.

For the open bench tests 3 temperature probes were used for the following measurements:

  • Max Internal Temp - Underneath the main internal heatsink, at a spot that seemed to provide the highest temp readings. Where the thermal sensor actually ended up is not known because it was not visible; the wire with the probe at its end was simply snaked through an intake vent slot.
  • Internal HS Temp - On the top center of the internal heatsink by inserting the thermal probe through a top vent slot.
  • External HS Temp - A thin-film sensor in between the internal heatsink and the and external heatsink.

The results of the open bench tests are shown in Table B above. The column Internal Heat is the difference between AC power input and DC power output; it is how much heat the PSU generates as a byproduct of converting the AC to usable DC. It's the amount of heat that the heatsinks are trying to dissipate.

Internal and external HS temperatures seem safe at any power level up to the maximum 170W. There is a temperature difference of 2-3 degrees indicates a bit of heat transfer loss between the two heatsinks.

Maximum internal temp was just an attempt to get an idea of how hot components might run. It's an arbitrary measurement that may have little meaning. In a fan cooled PSU, some components also run very hot, and they are rated to run safely at such temperatures.

For thermal simulation case tests, the internal and external case temperature probes were kept the same; the third thermal sensor was moved inside the case to a position about 1 inch from the PSU intake vents. Tests were run only at the maximum 170W output. As mentioned earlier, the back panel case fan used was a Panaflo 80mm low speed model (SPCR's reference standard).

Note that the case used for the thermal simulation has a 4" vent hole cut on the top panel directly over the PSU location. This case is the same one used for the quiet P4-1.6A system article I wrote about over a year ago and shown in the photos directly below. (The system has long been retired and its parts scattered throughout the testing room.)

The top vent came in very handy to consider the cooling effect of such a vent hole for the proSilence 350 PSU.

The wires going into the PSU are attached to temp probes. To simulate a case without a top vent, a manual was placed over the hole, as shown below.

The results of the thermal simulation case tests are shown below in Table C.

Table C. In thermal simulation case, at full 170W output power over 60W light bulb. Each reading after 10 mins.

Top vent
Case fan
Case Temp
Internal HS
External HS
closed
7 volts
36C
37C
33C
open
7 volts
36C
33C
29C
open
removed
41C
53C
45C
open; top CD bay cover removed; bottom PCI slot cover removed
removed
39C
51C
44C

As long as as the back panel Panaflo fan was running, the PSU temperatures remained modest. The top vent dropped the temps by 4C, which is as expected.

For a completely fanless PC simulation, the case fan was removed so that the case was cooled only by convection through the back panel fan hole,and the top vent was left open for PSU convection cooling. Case temps went up by just 5C, yet HS temps jumped very substantially.

Finally, in an effort to improve convection airflow in the case, the plastic cover on the front bezel for the top CD drive bay top and the cover for the bottom PCI slot were removed. It did not do much.

For an extreme stress test, the PSU was left running for 30 minutes without a case fan as per the last setting. The internal HS temp was measured at 64C and the external HS temperature reached 56C. The external heatsink felt too hot to keep running.

A final comment on load testing:

Full power testing of PSUs for any length of time is a very demanding test, generally tougher than what real use conditions can demand. SPCR's bench testing is steady-state and can be extended indefinitely until the PSU burns; in real world applications, PSUs in PCs rarely get anywhere close to this kind of abuse (except maybe in a serious server room, which is a different application altogether), and the power demand on them varies up and down in a much more dynamic way, with average power loads rarely exceeding 150W for desktop PCs.

CONCLUSION

The proSilence PCS-350W fanless PSU actually works! That is, it provides enough power to run most PC systems and does so without any noise. No fan noise, no buzzing, squealing or whining. The only caveat is that you need to provide adequate case cooling, because unlike fan-cooled PSUs, this one cannot provide any. A fanless PSU by definition is not really ATX12V compatible, because the standard actually calls for the PSU to provide at least some case heat evacuation. (See the case airflow drawing in the Recommended PSUs.)

This is a primary point to consider: If you buy this PSU thinking that it will make your system silent, think again. It will do that ONLY if your system has no need for case cooling fans or any other fans, and if your hard drives are silenced. If you have a higher power system, chances are, your case cooling fans will still be running the same as before. The overall noise improvement may not be large.

In typical systems, the proSilence should be able to cool itself adequately. This unit may have been adapted from a conventional PSU rated at 350W (a natural conjecture considering the model number). But attempting to obtain that kind of power from the proSilence PCS-350W without forced air cooling would clearly result in thermal runaway failure. The use of a one-piece HS to avoid the thermal loss at the junction between the separate inner and outer heatsinks would probably provide improved performance. One might also wonder whether a larger external HS might improve performance; the increased distance from the heat sources may reduce any cooling benefits.

The biggest downside of this PSU centers mostly on the effect it will have on your other components and your silent PC fantasies. After completely eliminating a major source of PC noise, many of you will undoubtedly be even more annoyed by any noise your other components make. For others, a fanless PSU is one step farther along the road to the completely silent PC.

Some of you may think that the proSilence PCS-350W lacks power to be really useful. Consider that I ran it for a while in the following system (which draws peaks of ~120W DC) without any stability problems and no change in fan speeds (all at 12V with hot weather now). The PSU it replaced was a Seasonic 300W:

  • P4-2.8G
  • 512mb PC3200 RAM
  • 2 x Seagate Barracuda IV 40G HDD
  • GF4-4800TI 128 mb VGA
  • 2 Optical drives
  • 1 Floppy drive
  • SCSI scanner card
  • 3 Panaflo 80L fans

* * *

The proSilence PCS-350W fanless PSU's strengths:

  • Heavy build quality, with massive heatsinks
  • Excellent stability
  • Very good voltage regulation
  • Good self-cooling within rated power limits
  • Truly silent: 0 dBA!

The proSilence PCS-350W fanless PSU's weaknesses:

  • Output power could be higher
  • Misleading 350W in model designation
  • Short ATX and 12V cables to motherboard
  • Price

The proSilence PCS-350W is a solid, viable fanless PSU well-suited for use in the vast majority of quiet desktop PCs. Its use does not obviate the need for fans elsewhere in the PC. A fan for case cooling is necessary to evacuate the heat generated by the CPU, hard drive, RAM and other components. Most systems will also require a fan for CPU cooling. Regardless, the proSilence PCS-350W deserves recognition for being the first fanless drop-in substitute for a ATX12V PSU.

Our thanks to Silicon Acoustics for this review sample and to SilentMaxx for their kind support.

* * *

POSTSCRIPT: May 24, 2004

Almost a year after this review, I am compelled to add this cautionary note. Over the past 11 months, I have heard about too many proSilence PCS-350 problems, both directly and in the PSU forums of SPCR. Despite the positive experiences many users have had, there appears to be a significant number problems. These problems include:

  • Buzzing or humming noise, either constant or whenever under load. Some users have gone through two replacements before getting a non-buzzy unit.
  • Early failures in just a few weeks or months.

It's good to note that the manufacturer and suppliers appear to have provided good service with warranty obligations.

I believe a large part of the problem is overheating, especially in systems that don't have good case airflow. In a typical tower case, the hot air naturally rises to the top and tends to pool there unless an evacuation path is available. In tower cases that employ the proSilence PCS-350, such an evacuation path does not exist, because the back panel of the proSilence is sealed. This is a dramatic contrast to the typical fan cooled PSU, which has a vent with a fan at the back; this fan does much in evacuating the heat that rises to the top of the case.

My suggestion here is that users of the proSilence and other fanless ATX PSUs take special care in providing hot air evacuation paths out of the case; a top panel "blow-hole" may be very effective for such systems. The stress over time in the temperature difference between a no-airflow PSU and a low-airflow PSU must be great, especially in hot weather.

* * * * *

Discuss this article in our Forums.



Previous 1 2 3 4

Power - Article Index
Help support this site, buy from one of our affiliate retailers!
Search: