20040413

It's the harmonics, stupid

Wired's shorticle on electronics forbidden on aircraft elicited the harrumpf factor.

"The operating frequency of these devices is well below the range of aircraft systems." But if these devices are digital, their myriad transistors are being turned from full-off to full-on, to signify the binary ones and zeros that trigger other transistors in turn in the construct we know as code. Whether that code drives the video of a GameBoy or the audio of an MP3, or a spreadsheet, makes little difference.

When a waveform of full-off to full-on and back is created, it wants very badly to divide itself into an infinite number of sine waveforms whose frequencies are multiples of the fundamental on-off frequency, with amplitudes decreasing as the frequency increases. If you have a lowly 16 MHz square wave clock, it will want to break apart into sine waves of higher frequency. They aren't intended, they just happen, and good design will seek to reduce the amplitude of those harmonics, or the power at which those harmonics escape the device.

The issue returns to frequency and power. What frequency or range of frequencies is being emitted by these devices, and at what power? Then, what frequencies, at what power, will disturb a commercial aircraft's systems?

It was our luck to have been employed on a project where the engineers intentionally overdrove a part so its harmonics were stronger, because one of those harmonics was desired as a clock frequency for the product. The manufacturer of the part improved it to reduce the harmonic output. It could not longer be mis-applied that way. The PCB had to be relaid, we missed milestones, the financing was withdrawn, the project cancelled.

On the bright side, this got Fûz laid off, forcing a relocation from NY to Colorado. The harmonics created by square waves have offered me a wild, sinusoidal rollercoaster ride.

No comments: