An Investigation of Propagation on the N.Atlantic path by monitoring Canadian Naval Station CFH.

The Canadian Naval station CFH transmits on 137.00kHz with an approximate power to the aerial of 20kW. It has been noted that when it is active, signals are received in Europe from this site during the night hours. Because dark path was the most likely time to produce two-way communication with N.America on the Amateur 136kHz band, it was decided to attempt to monitor the station regularly in an attempt to understand the vagaries of the propagation at this frequency.

Some initial logs were made by recording the scrolling screen of a waterfall FFT display ( FFTDSP4 by Mike Cook ) which provided at signal-to-noise measurement. Whilst this proved interesting it was not sufficiently accurate to derive really useful information. Also the background noise level varies quite noticably over the period of interest. After discussions with Vaino Lehtoranta OH2LX , I determined to set up a system using a simple commercial A/D converter to measure the signal level and provide a graphical display of the results. It did not seem worthwhile using the AGC output from the receiver as this could be greatly affected by noise, 'static' crashes and other signals, and time constants were not ideal for long term monitoring in noisy conditions. After a couple of false starts the following system was utilised.

The aerial, which is a 60m-top inverted 'L' at about 10m (approx 600pF ), is tuned by a 2.3mH solenoid coil with a movable ferrite core, before being fed to the 50 ohm input of an AOR 7030 receiver. The receiver is run with the preamplifier off ( 0dB position) is tuned to 138.100kHz. in the LSB mode. The 500Hz IF filter (later 300Hz) is centered over the frequency of CFH by using the Pass-Band tuning facility, and the receiver is operated with the AGC off. The 455kHz IF output (from the accessory socket) is fed to a Bruel & Kjaer Model 2426 wideband millivoltmeter. This allows easy selection of time constants and peak, average or RMS detector response. It also allows a dynamic range of about 50dB before saturation. The chart recorder ( DC ) output from the millivoltmeter was fed to a Pico Technology ADC-42. The logging and scaling of the signal is carried out in PICOLOG, a proprietry program supplied with the converter, and run under DOS on a Toshiba 1850 laptop (which is quiet at 136kHz).

The calibration was accomplished by feeding the receiver from a Marconi TF2002B calibrated signal generator, and making a table of the voltage measured by the ADC against the signal level applied to the receiver. This calibration table is stored in PICOLOG and is used to process the measurements by piecewise linear interpolation between the measured points (approximately 6dB intervals). The end result is a plot of the signal as received at the input of my receiver in dBu in 50ohms. I have continued to monitor the signal on the FFT waterfall display to allow investigation of unexpected effects (the appearance of amateur signals in the pass-band in the early morning and the occasional drifty TV) (0dBu = 1uV )

The system could be considerably improved with time, but with the level of the Amateur activity on the N.Atlantic path, I felt I must leave it as it is at present so that I am able to provide comparative data about the current conditions. I will also continue until I have a substantial number of results that I can attempt to correlate our successes with the strengths of CFH at the time. I also wish to attempt to correlate the signal strength profile with the Solar activity reports. I have noted that none of the agencies seem interested in publishing forecasts for multi-Megometre transmission paths on low frequencies.

Be careful in comparing the graphs as they do not all start at the same time. The time scale is UTC. The calibration has been checked from time to time and seems reasonably stable, though the accuracy is probably a little open to question due to the temperature coefficient of the reference in the ADC. As an estimate I would think it is probably stable to +/-1dB, but it is certainly within 2dB.

CFH (137.0kHz) ceased transmitting regularly just before Christmas 2000. When it became clear it would not return, I switched my attention to the Greek Naval station SXV in April 2001. SXV ceased transmitting regularly in August 2001, and was relocated onto an island to the south of Athens. It has only transmit sporadic test periods from this location using its old 135.8kHz frequency. During the summer of 2001 Brian CT1DRP started to monitor DCF39 and several other channels using DL6YHF's SpectrumLab software into which Wolfgang has built a text-based data logger. This allows the data to be readily inported into Excel or other spreadsheets for analysis and ploting. The majority of the results on X-ray flare enhancement are in main due to the 24 hours datalogs of the DCF39 signal. Brian is some 1950kms from Magdeburg whereas I am only about 700 kms distant. This results in Brian receiving a substantial skywave (ionospheric) signal component, which shows ionospheric enhancement is normal at this distance in daytime all year round. In Contast I receive just a big steady groundwave signal during the day. During early 2002 CFH has transmit rather more often, sometimes with the old FSK signal and sometimes with just a carrier (or maybe sub-carrier) about 25Hz off their nominal 137.0kHz. These transmissions have allowed us to compare Transatlantic propagation on 136kHz, at two very differnt locations. At May 2002 we have nearly 500 sets of data from the two sites. It is all archived and can be made available unprocessed.

See Brian's (CT1DRP) site for details of his system

Graphs for CFH September - December 2000 and SXV April to August 2001 are available in .GIF zipped form on Rik's (ON7YD) web site.

Alan Melia G3NYK (locator JO02PB )

return to Homepage