Wednesday, August 25, 2010

Numerical modelling of VLF propagation in the Earth-Ionosphere Waveguide

This video shows the E field strength in dB for the TM modes as function of distance from the transmitter and height above ground. The simulation show the effect of the sunrise terminator as it moves from right to left, lowering the effective reflection height from 87 km to 75 km. The calculation is not based on solving for modes but uses FFTs to propagate a scalar field forwards in small steps, correcting for the Earth's curvature at each step. The model uses a perfectly conducting ground (i.e. sea water) and neglects the effect of the Earth's magnetic field on the refractive index of the ionosphere.

Friday, November 27, 2009

I decided a blog would be a useful way for SoftPAL users to share ideas and information related to using SoftPAL to measure VLF signals, so my first example describes a way to easily determine the frequency offset required for MSK signals not controlled by an atomic standard.

Normally, I don't bother monitoring the Japanese MSK transmitter JJI at 22.2kHz because it is not phase stable. Recently I thought I had better check it once more, since it might be useful for SoftPAL users in India and perhaps Japan.

The default settings for JJI in SoftPAL (22.203kHz and 100 baud) were set some years ago and I suspected things might have changed, so last night I set SoftPAL to run two JJI signals - one at 100 baud and the other at 200 baud. The amplitude at 200 baud was several dB stronger than at 100 baud. This suggested the baud rate had changed to 200 baud. However, at 200 baud the amplitude had a strange square-wave variation, while the phase had a saw-tooth variation, which you can see in the screen shot below. The flat tops of the square-wave amplitude corresponded to the rising slope of the saw-tooth phase, so I suspected that JJI had drifted to a frequency beyond SoftPAL's lock capture range.

I decided to measure how far off frequency JJI was by using the Data Pad Slope calculation. First, I selected a good rising part of the phase saw-tooth. I then opened a Data Pad Mini-window using the tool bar. Clicking on the mini-window brought up the Data Pad Column Setup Dialog, which let me measure the Slope of the selected data on the JJI phase channel. This gave a result of 35.44 deg/sec, which is about 0.1Hz. So I concluded that JJI was now transmitting at about 22.21kHz. Sure enough, when used the SoftPAL Setup Dialog to change the frequency for JJI to 22.21kHz, the square-wave and saw-tooth variations disappeared!

The screen shot above shows the JJI phase and amplitude with SoftPAL set to 22.2116kHz. The phase is still too unstable for any thing but short time scale studies (fast Trimpis), but the amplitude is now much less noisy.

Prompted by the above results, I decided to measure the SoftPAL capture bandwidth for MSK signals. For strong signals, the measured value turned out to be +/- 0.06Hz, which is consistent with the filter time constants used by the MSK demodulator.