Zalman ZM-MFC2: 4 x Fan Controller + Power Meter

A fan controller with a power meter built in? Zalman’s ZM-MFC2 brings something new to the endless list of things you can display on your front panel: Power consumption. Oh, and it also controls up to four fans.

October 12, 2007 by Devon
Cooke

Keeping track of system temperatures and fan speeds is an integral part of
silencing a new rig. In a simple system, setting-and-forgetting the fans when
the system is built works well enough, but a regularly tweaked system needs
re-testing every time it changes, and that adds up to a lot of work. This is
where a fan controller like Zalman’s ZM-MFC2 comes in handy; it centralizes
all the thermal and fan information, and provides a way to tweak fan speeds
— without ever needing to crack open the case. The ZM-MFC2 also provides
one extra morsel of information: system power consumption.

The ability to monitor system power is the gimmick that Zalman hopes will set
its product apart from the rest of the fan controllers on the market. While
nobody needs to know how much power his or her system is consuming, it
does provide a means of judging how hard the cooling system should be working
before the system heats up.

Besides its practical convenience, there’s no question that Zalman’s fan controller
is about bling. Chances are, most of Zalman’s customers will use the flashy
information readout to show off their m4d m0dding sk1llz just as much as they
do for practical purposes. The power meter just adds to this — it adds
yet another layer of competition to system building. Here at SPCR, we hope people
will be competing to see how little power their systems consume, but
that doesn’t seem to be how Zalman expects the ZM-MFC2 — and its 800W of
headroom — to be used. In fact, even though the power meter will not measure
above 800W, Zalman has included a safety warning to use a thicker power cable
at currents above 10A (1200W in North America, double that most other places)
— just in case you somehow manage to build a system that requires an entire
circuit breaker to itself.

Lots of wiring goodness.

The contents of the box are pretty much as expected with one exception: A PCI
plate with a single USB plug. Closer inspection reveals that, while it is indeed
possible to plug a USB cable in, doing so would be a very, very bad idea; instead,
it accepts a signal from an external pass-through brick that measures AC power.
Those like me who tend to plug rear USB cables in blind be warned: This is an
excellent way to damage both your motherboard and the CVS controller.

From left to right: Manual, PCI plate for power signal, fan cables, thermistor
cables, power pass-through.

PHYSICAL OVERVIEW

The control console fits into a standard 5.25" bay. The bulk
of the console is set aside for the display, with a large jog wheel to adjust
fan speed and a small, easy-to-miss fan selection button at the far right. The
whole package has the sleek, faux hi-fi look that so many computer cases have
adopted in recent years.

The black and silver design has a sleek home theater look to it.

Around the back is a straightforward array of connectors: Three 3-pin fan headers,
a 4-pin header, a single header for the four thermistors, a small plug for the
power meter, and a standard Molex connector for power. All of the connections
are clearly labeled and use different headers, making installation safe and
simple.

With a total of 10 potential cables, cable management could prove a challenge.

The user interface leaves a lot to be desired. Perhaps Zalman was going for
the iPod’s fewer-buttons-is-better aesthetic, but it doesn’t work so well on
a fan controller where it’s hard to imagine a more intuitive interface than
a line of four knobs. Even so, the controller is needlessly complicated. It
doesn’t adjust the fan speed until a few seconds after the jog wheel has stopped
— and it doesn’t remember what fan was selected last. This makes fine adjustment
a bit painful, as each individual tweak is a three step process:

1. Select the fan by pressing the "mode" button an appropriate number
   of times.
2. Twirl the jog wheel to dial in a new target rotation speed.
3. Wait 5-10 seconds for the fan to adjust itself and listen to see if it is
   at the desired speed.

Twirling the knob without first selecting a fan does not adjust the last-selected
fan as you might expect; it is necessary to re-select a the fan for every minute
adjustment. Perhaps Zalman meant this as a safety feature to prevent accidental
changes in fan speed.

Plenty of flashy, bright, multicolored lights to incite your loathing or
satisfy your lust for bling.

Another issue we had was the viewing angle of the display. When viewed from
above — as would be typical for a system placed on the ground — the
display appeared fully lit, eradicating the ability to read anything on it.
A quick measurement and some trigonometry revealed that the display became unviewable
27° above a straight-on view.

TESTING

Ambient conditions during testing were 23°C and 115 VAC.

Power Meter

The power meter reports the total amount of power consumed by everything
connected to the power supply, including the controller itself. The meter
itself is a little black brick that is inserted between the AC plug and the
power supply. It is similar to the Kill-a-Watt power meter that we use to
measure power in the lab, but it doesn’t report anything except wattage. It
is accurate enough for everyday purposes, reading about 2% higher than the
meter we usually use.

Beside the numerical display is a visual bar meter that supposedly tracks
the numerical amount, but the resolution is too coarse to be of much use:
It has only four different stages, with the length of the bar changing every
200W. Since our test system never exceeded 200W, this meant that the bar
meter never changed, no matter how heavily loaded the system was. Even very
power-hungry systems are unlikely to push the meter beyond the second bar.
Doing so would require a system that draws in excess of 400W!

Fan Control

Aside from the interface quirks mentioned earlier, fan control was simple
and straightforward. Rotation speed is set directly, without taking thermal
data into account. Some tinkerers may miss the ability to set fan speed according
to temperature, but leaving the fans running at a constant speed makes good
sense from a silencing perspective, as it prevents sudden noise spikes that
draw attention to themselves.

Users with fans that buzz when controlled by PWM will be happy to know that
the ZM-MFC2 uses direct voltage control to modify fan speed — with the
obvious exception of the fourth 4-pin fan channel. Zalman reports the output
range as 4V~11V, but the actual range appeared to be highly dependent on what
fan was connected, and we observed everything from 3.23~12.29V during our
tests. As a general rule, lower current (slower, smaller) fans were supplied
with lower voltages than high current fans.

Zalman advertises the adjustment range as 60~5940 RPM, and they’re not kidding
about the low end of the scale. We had no problem stalling any of the fans
we tested, so some caution is advisable when cranking down the fan speed.
This is a strong point in favor of the ZM-MFC2, as most controllers do not
permit reducing the speed to the point of stalling. However, it is also somewhat
dangerous, as it is quite easy to reduce fan speed to the point overheating.
Unlike many products on the market, the ZM-MFC2 has not been idiot-proofed!

A small degree of safety is provided by the auto-restart function that boosts
the voltage if the controller detects a stall. By default, a stalled fan will
cause an audible but surprisingly inoffensive beep to go off and the voltage
will be spike briefly to 12V to restart the fan. If the fan setting is too
low, this will cause the fan to cycle in bursts. The alarm can be disabled
if it is too irritating (or there is no fan connected), but this is probably
not advisable during ordinary operation. If something goes wrong, it’s good
to know about it!

Fan speeds are regulated based on the rotation speed reported by the fan,
so 2-pin fans and other fans that do not provide a tachometer lead cannot
be controlled. Some nonstandard fans that do not output two clock pulses per
rotation may also work incorrectly.

Unfortunately, given that most fan tachometers are unreliable below ~500
RPM, fan control below this level seemed to be equally unreliable. Below this
level, the display reported 0 RPM but fan control continued to work down to
the fan’s stall point. However, the controller clearly had difficulty figuring
out what voltage to output in this situation, as the fan speed tended to wander
slightly at very low speeds. This caused a warbling effect that was more or
less audible depending on the fan.

Thermal Monitoring

Thermal monitoring worked as expected. All of the sensors read identically
at room temperature (23°C) and accurately matched the other thermometers
in the test lab. The sensors also correctly read the reviewer’s skin temperature
as 34°C.

FINAL CONCLUSIONS

The ZM-MFC2 has two very strong points in its favor: A wide control range that
allows adjustment to below the stall voltage of most fans, and the ability to
measure power consumption. That said, they are balanced by two design flaws:
A clumsy interface, and a narrow viewing angle.

Both of the strong points are unusual enough to push Zalman ahead of the competition,
but we couldn’t help thinking that they could be improved even more. To this
end, we came up with a wish list of things we would like to see:

• A fix for the wandering fan speed by fixing the output voltage permanently
  once the target speed has been found.
• A higher resolution bar graph to go with the power meter — perhaps
  with a variable scale based on the highest power draw ever observed.
• A log function for the power meter that tracks power usage over time.
• The optional ability to set fan targets based on thermal (or power) data,
  or to set a second target once a certain threshold is reached.

As it stands, the ZM-MFC2 is a useful product that will appeal to users who
can’t find its unique features anywhere else. The power meter is certainly a
feature we applaud; every eco-savvy computer user should have one! Its usability
issues hold it back from must-have status. Hopefully, Zalman will stand behind
its product and release an updated version with the bugs worked out.

Much thanks to Zalman
for the ZM-MFC2 sample.

* * *

Articles of Related Interest

* * *

Comment
on this article in our Forums