Fortron-Source Zen fanless 300W ATX12V power supply

It was first seen by SPCR staff all the way back at the January 2005 CES in Las Vegas, but the Zen by Fortron-Source Power didn’t make it to our test lab until a couple of weeks ago. It takes a somewhat different approach to cooling, boasts very high efficiency numbers and a more modest price. We repeat our caveats about careful usage, however, as with all fanless PSUs.

Aug 14, 2005 by Devon
Cooke and Mike
Chin

*POSTSCRIPT added Oct 22, 2005*

Fanless ATX power supplies have been on the market for some years now, and new
models are being introduced with increasing frequency. The FSP Zen is a recent entry
to the fanless category, but one look proves that
it’s not just a copycat product. While most fanless power supplies rely on hefty
external heatsinks to transfer heat away from the internal components, the housing of
the FSP Zen is as open as mosquito netting, presumably to allow air to flow through the power supply,
not over it. Large heatsinks are still used, but the internal space is not as tightly packed
as most other fanless models.

FSP claims an unheard of 89%
efficiency for the Zen, although no load or input voltage is specified. At this level of
efficiency, the heat generated inside the PSU at full 300W capacity would be just 37W ? about
half of that of an 80% efficient power supply would and less than one third
of a 70% efficient model. It suggests very cool operation, which is certainly good for a PSU that does not employ any cooling fan.

At the moment, the Zen is difficult to find in North America. An extensive
search turned up only a single online store that carries the model, and this
was by special order only. The new stock is only just finding its way into the retail channels.

The retail box is large and flashy: Good for sales but bad for the environment.

Our Zen came in a large, windowed box that invites retail sales. The marketing
material is actually quite informative: The number and type of connectors are
listed on the back of the box, as are the electrical specifications. Oddly,
our box came with a prominently displayed “Intel only” sticker. This
implies that the Zen is not compatible with AMD processors, which is, to put
it politely, hogwash. The Zen is an ATX12V power supply,
and should power AMD and Intel based systems equally well.

The number of connectors and the output specs are listed on the back of the
box.

The usual accessories are included: PSU, screws, AC cable, and a multi-lingual
instruction manual.

Although the 300W capacity doesn’t seem like much when compared to the 600W
behemoths that currently occupy the top end of the market, it’s fairly typical
for a fanless power supply. It is more than enough for most systems.
Only overclockers or heavy gamers are likely to push this unit to its limit.

EXTERIOR EXAMINATION

The majority of the casing is made from a rigid metal mesh which protects the
internal components while allowing air to circulate freely. One side effect of
this is that the internal components are visible through the mesh, which should
appeal to those who like to show off the inner workings of their PC. The corners
and edges are solid metal, to prevent the casing from being easily
dented or deformed.

Aesthetically, the Zen is a tranquil blue color. Personally, I prefer this
look to the highly polished titanium finish that so many other power supplies,
but this is really only significant in the retail showroom; there is no practical
benefit to the blue paint. There is a dim blue LED behind the power switch but
no other light sources, which should make the Zen bedroom safe, visually as well as aurally.

The wire mesh that encloses the Zen makes it possible to see the internal
components of the power supply — mainly heatsinks.

Five sides of the power supply use the mesh: The top side is solid steel.
This is unlikely to affect the airflow much, because the internal PCB that sits
against this side of the power supply would block the airflow anyway. Furthermore,
when it is installed in a typical case, this side of the power supply is flush
against the ceiling of the case.

Only the top of the unit is not covered in mesh.

The front and rear faces of the Zen both have vents of metal mesh. These vents
seems quite small, roughly 7 cm x 7 cm, and the mesh obstructs nearly 50% of
the surface area. In a fanned power supply, this would be a bad sign, as the
airflow would be limited by the small size of the vent. It is probably not so good here either. It seems a design error not to stretch the grillwork across the entire back panel, especially when FSP already uses such designs in their 120mm fan PSU cases.

It is
difficult to judge what effect this will have on a fanless power supply, where
the main sources of airflow are convection and peripheral airflow from other fans in the case. The direction of the airflow will change
depending on how the fans in the case are set up. In an
negative pressure system typical in quiet computing, the rear vent would act as an intake, drawing fresh
air through the power supply. A positive pressure system (such as our test box)
will use the rear vent as an exhaust.

With such low airflow, the amount of airflow does not matter as much as where
the airflow is, so the relatively small exhaust vent may not be as important
as its position. We can only speculate about the actual engineering intention
of these vents, but the thermal portion of our testing should give some indication
of its success.

The rear exhaust is fairly small.

The front end of the unit has another patch of mesh, plus three small intake
slots.

INTERNAL LAYOUT

Like most other fanless power supplies, the interior of the power supply is
dominated by large heatsinks. In keeping with the open design, there is room
for air to circulate between the various components.

The vents on each end of the unit are aligned with each other and seem to be
designed to ensure that the airflow current is kept to the side of the power
supply. The open casing should also allow the air inside the power supply to
mix with the air in the larger case, preventing heat buildup inside the power
supply and keeping the ambient temperature close to that of the case interior.

Unless you look closely, all that can really be seen is heatsinks.

Vertical aluminum plates transfer heat to the finned heatsinks.

The heatsinks all have heavy fins that run lengthwise though the power supply.
The fins face downwards, and the spaces between them should help create a convection
current that directs heat to either end of the power supply, where the heat
will be evacuated from the power supply.

The fins form channels that direct airflow to either end of the unit.

Heat is guided to the end of the power supply where it can escape through
the exhaust vent.

CABLES AND CONNECTORS

There are a total of five cable sets and two splitters that increase the number
of IDE power headers. All cables are sleeved with the exception of the two splitters.

• 14″ sleeved cable for main 20+4-pin ATX connector
• 15″ auxiliary 12V connector
• 21″ cable with two 4-pin IDE drive connectors
• 26″ cable with two 4-pin IDE drive connectors and
  one floppy drive power connector
• 21″ cable with two SATA drive connectors
• 2 x 6″ two way splitters for IDE drive connector

The cable lengths are pretty standard.

On the whole, the Zen has fewer and shorter cables than most
power supplies we’ve reviewed recently. This is not necessarily a bad thing: Most systems do not require
the large number of connectors included with most retail power supplies, and
unused cables can impede airflow if there is no space to hide them. There are
enough connectors to power any low to mid-range system, although the lack of
a PCI-e header and the modest capacity on the +12V line makes it a poor choice
for use with a high powered VGA card. The limited number of SATA connectors
may also pose a problem for those with a lot of SATA drives, although most drives
on the market (even SATA drives) can still be powered with the legacy IDE power
connectors.

TEST RESULTS

For a fuller understanding of ATX power supplies, please read our article Power
Supply Fundamentals & Recommended Units. Those who seek source materials
can find Intel’s various PSU design guides, closely followed by PSU manufacturers,
at Form Factors.

For a complete rundown of testing equipment and procedures, please refer to
the article SPCR’s Revised
PSU Testing System. It is a close simulation of a moderate airflow mid-tower
PC optimized for low noise.

In the test rig, the ambient temperature of the PSU varies proportionately
with its output load, which is exactly the way it is in a real PC environment.
But there is the added benefit of a precise high power load tester which allows
incremental load testing all the way to full power for any non-industrial PC
power supply. Both fan noise and voltage are measured at various standard loads.
It is, in general, a very demanding test, as the operating ambient temperature
of the PSU often reaches >40°C at full power. This is impossible to
achieve with an open test bench setup.

Great effort has been made to devise as realistic an operating
environment for the PSU as possible, but the thermal and noise results obtained
here still cannot be considered absolute. There are far too many variables in
PCs and far too many possible combinations of components for any single test
environment to provide infallible results. And there is always the bugaboo of
sample variance. These results are akin to a resume, a few detailed photographs,
and some short sound bites of someone you’ve never met. You’ll probably get
a reasonable overall representation of that person, but it is not quite the
same as an extended meeting in person.

REAL SYSTEM POWER NEEDS: While
our testing loads the PSU to full output (even >600W!) in order to verify
the manufacturer’s claims, real desktop PCs simply do not require anywhere near
this level of power. The most pertinent range of DC output power is between
about 65W and 250W, because it is the power range where most systems will be
working most of the time. To illustrate this point, we
recently conducted system tests to measure the maximum power draw that an actual
system can draw under worst-case conditions. Our most powerful P4-3.2
Gaming rig drew ~180W DC from the power supply under full load ? well within
the capabilities of any modern power supply. Please follow the link provided
above to see the details. It is true that very elaborate systems with SLI could
draw as much as another 150W, but the total still remains well under 400W in
extrapolations of our real world measurements.

INTERPRETING TEMPERATURE DATA

It important to keep in mind that fan speed varies with temperature,
not output load. A power supply generates more heat as output increases, but
is not the only the only factor that affects fan speed. Ambient temperature
and case airflow have almost as much effect. Our test rig represents a challenging
thermal situation for a power supply: A large portion of the heat generated
inside the case must be exhausted through the power supply, which causes a corresponding
increase in fan speed.

When examining thermal data, the most important indicator of cooling efficiency
is the difference between intake and exhaust. Because the
heat generated in the PSU loader by the output of the PSU is always the same for a given power level, the intake temperature should
be roughly the same between different tests. The only external variable is the ambient room temperature. The
temperature of the exhaust air from the PSU is affected by several factors:

• Intake temperature (determined by ambient temperature and power output level)
• Efficiency of the PSU (how much heat it generates while producing the required output)
• The effectiveness of the PSU’s cooling system, which is comprised of:
  • Overall mechanical and airflow design
  • Size, shape and overall surface area of heatsinks
  • Fan(s) and fan speed control circuit

The thermal rise in the power supply is really the only indicator
we have about all of the above. This is why the intake temperature is
important: It represents the ambient temperature around the power supply itself.
Subtracting the intake temperature from the exhaust temperature gives a reasonable
gauge of the effectiveness of the power supply’s cooling system. This is the
only number that is comparable between different reviews, as it is unaffected
by the ambient temperature.

Because the side of the Zen is open, the top and the side of the test rig was
covered up with cardboard to simulate the walls of an actual case. This ensured
that only the rear vent is exposed to the external air, as it would be in a
real system.

Ambient conditions during testing were 25°C and 20 dBA, with
input of 120 VAC / 60 Hz measured at the AC outlet. It was a couple of degrees
warmer than usual in the lab, and the Intake Temp readings in the measured data
table below reflects this.

ANALYSIS

1. VOLTAGE REGULATION was excellent, within ±1% on the +12V
and +5V lines in any combination of loads. The +3.3V line strayed by a maximum
of 2%. This is very impressive performance.

• +12V: 11.89 to 12.04
• +5V: 4.96 to 5.00
• +3.3V: 3.35 to 3.38

2. EFFICIENCY was excellent, although we did not quite reach see the claimed 89%. It is likely that this efficiency could be reached if a 240VAC
input voltage was used. Most impressive is the efficiency at low output: This
is the first power supply we have tested that measured 80% efficient at the low 65W
load. As the load increases, efficiency also goes up. Peak efficiency was achieved
at 300W output.

This performance is on par with the best we have measured: The
Antec Phantom 500. Although the Phantom is slightly more efficient under
higher loads, the Zen is more efficient below 150W ? the range in which
it is most likely to be used. These differences are minor, however, and are
always within a percentage point or two. The differences may lie within the resolution capability of our test setup.

3. POWER FACTOR was lower than we usually expect of active power factor
correction, although still very good compared to passive or no PFC. Once the
total load rose to 150W and above, power factor quickly approached the ideal
value of 1.0.

4. NOISE

The Zen is fanless, so the noise generated by the power supply was minimal.
No electrical noise was noticed during the course of the testing, which cannot
be said of some of the other fanless models on the market. There was no hum, buzz or squeal that could be heard from any distance. A trace amount of something was barely audible to Mike when he pressed his ear right up against the cover. For all practical
purposes, the Zen is silent.

5. HEAT

The temperature rise within the power supply was in line with the other fanless
power supplies we’ve tested. Although a 20°C rise at 300W output is hardly
stellar thermal performance for a fanned power supply, it’s fairly typical for
a fanless model. Warm air could be felt coming out of the rear vent throughout
testing, which shows that there was airflow through the power supply even
though it has no active source of airflow. We can attribute most of this airflow to the internal airflow within our test
rig. The overall stability of the power supply gave us no
reason to think that its internal cooling was inadequate.

REAL SYSTEM TESTING

We generally do not test the PSU in a real system due to the constraints of time. With fanless PSUs, however, we feel compelled to try it in a real system because inability to stay cool in a real system can be disastrous. At least we can verify that a fanless PSU will work in our low airflow, quiet system.

The sample Zen PSU was installed in a system with the following components:

• AOpen i915Ga-PLF motherboard
• Intel 660 3.6 GHz processor, rated 115W TDP by Intel, overclocked to 3.8 GHz
• 1 GB OCZ Gold PC4000 DDR SDRAM (2 x 512 MB sticks)
• Scythe Ninja CPU heatsink
• AOpen Aeolus 6600GT-DV128 PCIe video card
• Antec SLK3000B mid-tower case

The system had two fans: One 120mm exhaust fan on the back panel and the Nexus 120 fan on the Ninja heatsink blowing in the same direction.

Windows XP Pro ran fine with the PSU during typical tasks. When the system was loaded a bit of odd, buzzy ticking from the PSU that had not been present on the test bench. We began running some benchmarks to stress the system a bit harder, and the noise became a bit louder. The total AC power draw was 240W, which means it was delivering around 200W DC to the components. It’s a fairly high load but well within the capability of the Zen.

Mysterious Failure

After about 15 minutes of CPU stress testing with CPUBurn, the system stopped running and refused to boot. There was no dramatic flash or bang; it simply turned off .

Subsequently, the system was tested with a different CPU, different RAM and with just the on-board video. At no time did the motherboard even post. We pulled the PSU out of the system, and tested it on the bench to find it completely dead. A second check several days later confirmed that it was still dead. (We had had the odd PSU mysteriously come back to life a day or two after some kind of failure.)

Later, we installed a known working PSU with all the other original components. When the power was turned on, the system refused to post, then moments later, a small flame burst forth at the end of the VGA card. It burned out almost as quickly as it started, within seconds. It was such a shock, I felt like I was moving in slow motion to pull the AC plug off the back of the PSU after the flame burst forth. A close examination of the VGA card revealed the burned piece to be a small 8-pin IC. Mike said he’d never actually seen a flame on a PCB before. Sparks, yes, and a flame off melting plastic from wire insulation, but not on a PCB component. The acrid smoke took a while to clear.

At this point, we can say little about the cause of this system failure. The fact is that we have a dead PSU and a dead VGA card, and their failure seems related. All the other components survived. We suspect the PSU more than the VGA simply because it made that buzzing noise before failure, but there’s no way to be sure. This is just conjecture.

FSP in the US moved quickly to supply us with a replacement sample. They asked for the dead sample to be returned; as this article is being posted, the sample is packed and ready to go back to FSP in Calfornia for a forensic diagnostic by their technicians. Incidentally, AOpen has asked to examine the VGA card that died, as well.

A Good Second Sample

The second sample measured pretty much identical to the first on the test bench. We installed it in a different system this time:

• Intel 520 P4-2.8 GHz processor for socket 775
• 512 MB Corsair DDR2 533MHz SDRAM
• Intel D945GTP mATX motherboard w/integrated VGA
• Scythe Ninja CPU heatsink, cooled
  with a Nexus 120mm fan at 12V
• Western Digital Caviar SE WD2500JD 250GB hard drive

The Fortron-Source Zen powering a modest Prescott P4 system.

Admittedly, this system is much less demanding than the first. A higher power system was not available at the time, and we were short of PCIe VGA cards, so we ran with what was on hand. Still, it’s closer to the kind of lower power system that most quiet system builders would favor with a 300W PSU. The total power draw running CPU stress and system benchmark software was 160W in AC, which translates to about 130W DC output from the PSU. The FSP Zen worked perfectly, never making any trace of the odd noise we heard from the first sample.

The CPU temperature was a bit higher than normal in this setup, reaching 65°C (in a room ambient of 25°C). The voltage regulators on this Intel 945 chipset board have been noted to run fairly hot; this might be a factor. The absence of a fan in the PSU was probably a factor as well. Being situated so close to the CPU, the heat from the PSU has to have an impact on CPU temperature.

A Hotter VGA Card

Some time later, an AOpen Aeolus 6800GT DVD256MV PCIe VGA became available, and it was installed into the above system for a higher load. With the nVidia 6800GT GPU, the system idle went up to 135W AC, and running CPUBurn and 3DMark2005, the maximum AC power peaked at around 250W. The latter translates to over 200W DC output.

The CPU temperature reached the slightly higher temperature of 67°C at load, and the board sensors also read a bit higher, but stability was unchanged. This was expected with the ~90W increase in total AC power caused by the 6800GT vidcard. That 90W translates almost directly into additional heat in the PC. The system worked as fine as before, albeit considerably speedier with the AOpen 6800GT VGA. Still, in the long term, this kind of power hungry video card may not bode well for longevity due to the thermal challenge.

In the test system, the back vent of the PSU served as an intake. Outside air was pulled in through the PSU grill opening by the case exhaust fan mounted below the PSU. It means that rather than follow any convection path of rising heat, the air was being forced in the opposite direction. The fan was the stock Antec 120 TriCool at the lowest of its 3 speeds. At this setting, it blows 28 CFM in free air, and it is quiet, measuring 20 dBA at 1 meter.

A point of concern is that because the bottom panel of the Zen PSU is so wide open, the air could be sucked in from the PSU exhaust grill and down to the 120mm case fan without moving across much of the heatsink fins, without any real cooling action being caused by the airflow. Thermal engineers refer to such airflow paths as short circuits. This might lead to the heat being trapped inside the PSU. Or perhaps that fan creates enough turbulence to keep air moving around in the PSU. Without much more sophsticated testing tools, it’s not possible to know for sure. It may be worthwhile to actually block the part of the cover of the Zen so that the air is forced to flow across more of the heatsinks before being pulled out of the PSU.

The back case panel exhaust fan ended up pulling air through the
PSU back vent.

CONCLUSIONS

FSP’s approach to the fanless PSU is different from its competitors: Instead
of turning the PSU casing into a gigantic external heatsink, FSP has opted for improved ventilation to better use to existing case airflow. The heatsinks are certainly beefer than in any normal PSU, but other than that and the super-perforated cover, the mechanical design is conventional. This means much reduced tooling and manufacturing costs, which is a good thing. It’s reflected in the selling price, which is substantially lower than that of the fanless Silverstone or Antec PSU models.

We have questions about how well the mesh casing works to direct existing airflow, especially the limited mesh area on the back panel. Although we are unable to examine real world airflow paths in detail, we suspect that cooling in the Zen could be improved.

A more obvious down side to the Zen is the short length
and sparse selection of cables. If you are working in a large case and cable
management is important to you, you may be frustrated by the conservative cable
lengths. On the other hand, a small case may benefit from the shorter cables,
as there will be less slack that needs hiding.

The Zen should not be run in a completely fanless case without special cooling arrangements. In a conventional case without any fans, the Zen would need to
rely on convection alone to evacuate the heat, and that will not be enough except with a very low thermal system, perhaps a Pentium M setup with a modest VGA card. We’d recommend a CPU fan at the very least, preferably, a CPU fan and a case exhaust fan. Care should be taken when using the Zen in any system that draws much more than ~200W AC, especially in warmer weather, say >25°C.

The Zen is a viable choice for use in a quiet computer, and its more conventional
design makes it more affordable than the competition. The electrically important aspects — efficiency, stable power delivery and voltage
regulation — are excellent, proving that you don’t
have to sacrifice performance for low noise. All in all, FSP has a solid fanless PSU in the Zen.

* * *

Much thanks to FSP
Group USA for the opportunity to examine this power supply.

Discuss this article in the SPCR Forums.