This is how a friend and I were trying to drive a DC motor and ended up making our first wireless broadcast without even knowing it! It took me one day after to figure out what really happened …

That day, we went down to Al-Abdali, bought some electronic components, and came back to my place to try them up. We were trying to drive a DC motor and control its speed using a PIC microcontroller. The circuit we connected was something like this:


The circuit was basically a common-collector (buffer) arrangement that is intended to switch the motor on/off using a low-current microcontroller logic signal. It worked perfectly for still logic values.

As soon as that worked, we reprogrammed the microcontroller to generate an oscillating signal in an attempt to drive the motor at a lower speed. We used Pulse Width Modulation (PWM) which is a technique that involves changing the duty cycle of a rectangular signal in order to change its DC average. Supplying a motor with a rectangular signal of amplitude 5 volts and 50% duty cycle is like giving it 2.5 effective DC volts. Using the microcontroller we can easily change the duty cycle (0 to 100%) of the control signal and thus change the effective amount of DC voltage delivered to the motor.

It sounds cool but unfortunately it didn’t work using our design (I found later that the switching times for the BJT transistor we used were too long and couldn’t match the frequency of the control signal, we should have used MOSFETs).

Anyway, after we re-programmed the microcontroller to use PWM, the motor became motionless and we started testing the pins with the voltmeter. My friend connected the voltmeter probe to the base of the transistor and there it happened …

We heard a “beep” going out from somewhere!

It wasn’t the circuit … it wasn’t the voltmeter … and it wasn’t the power supply. The sound actually came from the computer speakers that were right next to us.

It was a bit weird … the computer had no connection whatsoever with what we were doing and the sound was produced only when the voltmeter probe was connected to that specific pin.

I was too tired and exhausted that time that I just didn’t bother to find out what that was. We ignored the sound, and kept experimenting with the motor for some time.

The next day I paid it a little thought and found out what was that about …

Whenever the voltmeter probe was connected to the base of the transistor, it was connected to the PWM signal. Since tiny currents flow into the base of the transistor and inside the voltmeter, the probe picked up the entire signal and worked as an antenna! It broadcasted a rectangular periodic signal which, according to Fourier, is an infinite sum of sinusoidals at integer multiples of its fundamental frequency. The frequency we used was 500 Hz, which is pretty low, and so the resulting harmonics were at 500, 1000, 1500 Hz and so on. Periodic signals with low frequencies have denser frequency spectra. The frequency we used, along with its harmonics, was in the audible frequency range and was thus easily picked by the computer’s sound card!

It sounds weird but that’s what actually happened!

It’s not like we broadcasted anything meaningful, but as far as we went, that was an official monotonic test signal transmission :P