|
This file
1mhz_000.zip (51096990 bytes)
is a zipped version of 1mhz_000.wav which was recorded
June 2011 by John, OZ6AGQ with a Perseus and a magnetic
loop antenna.
The center frequency is 1.0 MHz and the sampling speed 2 MHz.
The file contains strong interference from a 32" LCD TV. The characteristics of the interference is obvious from figure 1. |
|
|
| Figure 1.High speed waterfall graph from the 1mhz_000.wav recording. |
|
The interference is an overtone-rich frequency modulated oscillator.
Presumably some kind of switch power.
The lowest frequency is about 88 kHz.
Only the odd overtones are visible at 264, 440, 616, 792, 968, 1144,...
The highest frequency is about 160 kHz.
It is obvoius that the interference is deterministic.
It can be modelled in the computer and subtracted with great precision.
This class of interference can be described with a limited set of parameters. A function that describes the fundamental frequency as a function of time. In this particular case the FM modulation is a 50 Hz waveform that is dominated by its overtones. The 100Hz component is strongest. The 50 Hz component is small but varies with a period of about 1 Hz. That variation is nearly a sine-wave. Each overtone has its own amplitude and it varies probably a little with the frequency. In all something like 50 parameters would probably be enough to describe the interference over long times. Once a model of the interference has been set up its parameters can be fitted to the input waveform. Then the model waveform could be subtracted. That would most probably eliminate this kind of interference completely. Setting up suitable algorithms is probably difficult. The recording on this page is made available to all clever programmers as a challenge for those interested in advanced SDR algorithms for interference fighting. To SM 5 BSZ Main Page |