The original article comments on the fact that the cooling fan can be heard running, and knowing that this is an annoyance to some operators I decided to try and reduce the noise. Believing the 33 cfm rated airflow of the fan to be significantly more than what is needed to cool the amplifier I decided to reduce the fan speed by reducing the voltage applied to the fan motor. This technique works well with brushless DC fans but there is a minimum voltage limit below which the fan motor won't start running.

       To find out just how much the voltage could be reduced I connected the fan to a variable-voltage DC power supply and gradually increased the supply voltage until the fan began running. The applied voltage was approximately 3.5 volts for a clean start every time, but the airflow at this voltage was not deemed sufficient to provide the cooling required. I gradually increased the voltage and found that about 7 volts seemed to give a more reasonable cooling effect but with a significant reduction in fan noise.

       With a shack supply voltage of 13.6 volts, a fan voltage of 7 volts meant that 6.6 volts would have to be dropped across a series resistor. I measured the current drain of the fan at 7 volts and found it to be about 60 mA, so applying Ohm's Law:

                              6.6 volts/0.060 amps = 110 ohms

        This is the value of series resistance required to drop the necessary 6.6 volts at the observed value of current flow with 7 volts applied to the fan motor itself. The power rating of this resistor can be found from the Power Law, P=I²R, as:

                              0.060 amps x 0.060 amps x 110 ohms = 0.396 watts

        At this point I decided to try an experiment, so I connected two 100 ohm, 1-watt resistors from the junk box in parallel, connected them in series with the fan, and cranked the power supply up to 13.6 volts output. The cooling improved noticeably but the noise increased only slightly, so I decided to use this combination. The amount of voltage dropped across the two resistors was measured at 4.83 volts, which meant that 8.77 volts was applied across the fan. The parallel combination of the two resistors measured 49.7 ohms on the DVM so that meant that the current was about 97 mA (use Ohm's Law to verify!). Using the Power Law as before, the required wattage rating calculated to 0.47 watts. Since the two1-watt resistors in parallel can safely dissipate 2 watts there is no likelihood of the resistors overheating. If a single 47 or 51-ohm resistor is used I recommend that it have at least a 1-watt rating to provide a safety factor.

        One note about tinkering with small DC fans: The amount of current drawn by the fan is not linear with applied voltage, so assuming that if you halve the applied voltage that the current will be halved is a useful starting point but it isn't really accurate. Therefore, unless you have a variable-voltage power supply handy you will have to determine the exact resistance required for a specific voltage drop by the empirical method, i.e., cut-and-try. Once determined, measure the actual current flow to determine the wattage rating you'll need (HINT: Use high-wattage resistors while you're experimenting, otherwise they may turn into firecrackers if you're not careful).

        I inserted the two parallel resistors in the positive lead to the fan and covered the work with a piece of heat-shrink tubing. The resultant airflow provides adequate cooling for the amplifier. While it does run a bit warmer than before it is still just comfortably warm to the touch after 45 minutes of 30 seconds on, 30 seconds off operation running FSK441A. Within 5 minutes of cessation of transmitting the chassis & heat sink are cool to the touch. The fan cannot be heard running unless you put your ear close to it and the fan in my IC-706MkIIG normally masks the sound.