6 June 2017
n WITH KRISTEN A. McINTYRE
A Fan of Quiet Fans
QI have a nice surround sound stereo system inside a closed cabinet. The cabinet has a built- in cooling fan, but I find it too noisy for regular TV viewing. So, I usually leave it off. However,
I also forget to turn it on before I watch some blockbuster
movie with booming bass swelling crescendos. It’s a real
bummer to have the receiver go into thermal shutdown
mid-explosion, so I was hoping you might be able to design
a circuit that would sense the temperature in the cabinet
and turn on the fan when things get toasty.
AI also like quiet fans. My employer has a long history of trying to either eliminate fans, or at least make them less objectionable. It’s possible to make an individual fan less noisy exploiting
the physics of aerodynamics, but I find that no matter
what, the less air they move, the less noise they make.
I had this very same issue myself some years ago.
As a ham radio operator and what’s called a net control
operator (someone who directs the flow of a round-robin
conversation — or Net — on the radio), I can spend a lot
of time on the air transmitting. In particular, I have been
net control operator for a long-standing net called the 9
AM Talk Net (see http://9amtalk.net) for many years. On
Fridays, this net can easily approach three hours or more in
length. I may spend about 30%-40% of that time with an
FM transmitter keyed down at 50W.
FM transmitters generally use Class-C non-linear
amplifiers, so they are roughly 70% efficient. This is not too
bad, but it means that 30% of that 50W is turned into heat:
50 0.3 = 15W. While that may not seem like much, in a
small space in the summertime, it gets really hot. The radio
in use has almost no cooling fins. It’s an FT-1500, for those
who know. I’ve measured the chassis/heatsink temperature
at 160°F/71°C under extreme operating conditions. This
is near the top of the operating temperature range for the
MOSFETs in this application. Since I’m talking on the radio,
I wanted a quiet fan to cool it when necessary.
Of course, we could use a computer to control
something like this, but being someone of a more analog
mindset, I decided that I would do this only with transistors
and a few other analog components. That included using a
transistor as a temperature sensor.
As you might know, Vbe, or the base-emitter voltage
of a transistor is quite sensitive to temperature, decreasing
with temperature increase. This is often used to provide
negative thermal feedback when biasing a class-A or class-AB amplifier.
There’s a great little circuit called a Vbe multiplier that
can be used to both sense that voltage and also amplify it
with a variable resistor. This forms the basis of the circuit.
A little experimentation showed that the rotational
speed of commonly obtainable muffin fans is best
controlled with current — not voltage, as you might guess.
So, we can use a transistor; in this case, a Darlington pair
to change voltage into current. This controls the fan’s air
I’m sure it’s highly non-linear, but
we’re using negative feedback, so only
the error signal is what counts as long
as the system is stable (i.e., it doesn’t
oscillate). So as the device (in this case,
a radio) heats up, the fan spins faster
until it reaches an equilibrium. That
is why I call this the homeostatic fan
The schematic is presented in
Figure 1. It’s a fun circuit because it
accomplishes so much with so little. This
schematic is more complete than my
usual drawings because I actually made
a printed circuit board (PCB) for it.
In this column, Kristen answers questions about
all aspects of electronics, including computer
hardware, software, circuits, electronic theory,
troubleshooting, and anything else of interest to
the hobbyist. Feel free to participate with your
questions, comments, or suggestions. Send all
questions and comments to: Q&A@nutsvolts.com.
• A Fan of Quiet Fans
• NiCd to NiMH
• Clean Automotive Power
Q & A
n FIGURE 1. Homeostatic fan controller.