The 136kHz Page.

In Europe, a LF (low frequency or LOWFER) band between 135.7 and 137.8 kHz with a maximum ERP (Effective Radiated Power) of 1 watt has been allocated to the Amateur Service. In the UK an additional LF assignment exists between 71.6 and 74.4 kHz, although this is only a temporary assignment and is due to be withdrawn in 2003.

Although LF was undoubted used by the very first Wireless Pioneers, who were in effect only Amateurs, in the very early years of the 20th Century, LF is regarded as one of the final frontiers for Amateur Radio, as Amateurs have never been able to apply what is almost 100 years worth of accumulated applied knowledge and experience to this part of the spectrum.

Despite the application of state of the art techniques, obtaining ranges of more than a few hundred miles still represents a significant technical challenge.

At LF it is very difficult to generate a significant amount of radiated power. Although the licensed power limit is only 1 Watt ERP, in practice for any amateur living in a suburban location, to actually achieve the legal limit is quite difficult as the antennas are only a fraction of the wavelength (2200 metres) and hence radiating efficiencies tend to be in the region of 1 percent or less. Thus it is normal to run a kilowatt or more simply to generate the required ERP. My local soil conductivity is particularly poor, and with my resultant ground losses, I can only generate about 660mW ERP with 1500 watts of transmit power!

Although propagation at LF is very predictable and given good ground conductivity, the signal once launched can travel many hundreds of kilometres, but due to the low ERP obtainable on this band, signal levels are correspondingly low. Nevertheless, a station running 1 Watt ERP can reliably make normal two way QSOs over ranges of 1000 to 1500km, and even stations running powers as low as 25 Watts can still expect an effective range of several hundred kilometres.

However, for real DX work it is essential that the receiver is capable of yielding the highest possible signal to noise ratio. Invariably this requires narrow filters to be used, and the Digital Signal Processor within the computer's sound card can be programmed to provide the required narrow filters. Morse (CW) is very much the ideal mode and with long element lengths, i.e. long dots and dashes in the order of seconds, tens of seconds, or even hundreds of seconds, signals can be resolved well below noise level. Such "slow" Morse is know as QRSS, after the Q code for "Please send slower" - QRS.

These pages show some photos of my rig, the antenna system and screen captures of my signal during tests undertaken over the Christmas and New Year period of 2001/2002. Tests were conducted using Morse signals where the length of the dot element was 60 seconds.

The rig used for the successful Trans-Atlantic tests.

The main transmitter is the aluminium box closest to the camera. Above the transmitter on the shelf is a Marconi 2945 test set. This has a very stable TCXO as it's internal frequency reference and yields a signal of very high stability and was used as the master oscillator for the transmitter. The output of the 2945 test set was in the 1.36MHz region, the transmitter being in effect no more than a TTL decade divider to generate the required 136kHz signal, gating to produce the Morse (CW) keying and a small push pull MOSFET amplifier to produce about 25 Watts. The unit also contains a receiver pre-amplifier.

The black heatsink to the left of the transmitter is the main power amplifier. This is based upon the "Decca Navigator" switched mode transmitter as per the RSGB's LF Experimenter's Handbook, but redesigned to operate from a 300V+ supply, and uses four IRFP460 MOSFETs in a bridge configuration. This has been tested at 2500 watts output into a dummy load. The brown wooden box to the left of the PA is a variac for output power setting. Behind the PA is a voltmeter measuring the DC supply voltage, and to the rear of the PA and transmitter is the bank of electrolytic capacitors and toroidal transformers for the DC supply.

On the shelf above the PA is a Tektronix 465 oscilloscope used to monitor the switching waveform on the MOSFETS. Viewing the switching waveform gave a very good indication as to the state of tune of the antenna, as any reactance would cause the MOSFETs to have a "slow" turn-on and turn-off time, which could lead to their spectacular destruction!

The Bench Power Supply at the back right of the photo was not part of the system.

The Antenna System.

The smaller dustbin (trash can) to the left contains the line coupling coil and an RF ammeter, and the larger one to the right contains the main antenna tuning coil.

Behind the dustbins is the ground system. Almost 1000 metres of wire is used, although it is in less than a favourable layout due to the house, garage etc. The antenna is a Marconi "T" supported between two large trees. The horizontal top section of the antenna is about 30 metres long, and the vertical section is about 23 metres.

The Antenna Tuning coil.

This is a photo taken during the Summer of 2001.

My efforts to get onto 136kHz date back to early 2001. Initially I tried to use a vertical monopole about 15 metres high which is extremely short for this frequency and this coil grew in length as I tried to get enough inductance to resonate the antenna. It is actually sections of 10" diameter drain inspection handhole from Wickes - the builder's merchants, which press fit into each other to make it as long as you wish. (B&Q sell 16" diameter versions which should give a better Q)!

Even with the coil some 4ft long, the 15 metre Monopole Antenna still would not resonate, so I gave up and erected a Marconi T strung between two straegically placed tall trees to the front and rear of my house. Now I had too much inductance so about 1/4 of the turns at the bottom of the coil were shorted out.

For the Christmas and New Year Tests, this tuning coil was mounted in the dustbin - with bricks at the bottom for added stability should a wind happen to blow! The unwanted bottom section of the coil was removed which enabled the coil to fit the dustbin even though I had to hold the lid on with a piece of string!

The Variometer.

Built into the top section of the tuning coil is a variometer to enable the antenna to be tuned to resonance. This is a few turns of wire on a tube having a slightly smaller diameter than the main coil so as to permit fine tuning of the total inductance.

 

L.F. Signal Reception.

QRSS signals are seen - not heard, and as the period of time required to send G4FTC using 60 second dots takes about 1 hour, the receiver is set to capture screenshots of the PC display at set intervals. A two way QSO using QRSS is therefore a lengthy process!

All the following screenshots were with about 350mW ERP with the transmitter output being 700 watts.

VE1ZJ's screenshot of my signal at 523.4 as taken on 29th December 2001

 

My signal as received by W4DEX - just visible on 924 on 30th December 2001

 

My signal at 135.933 kHz as received by DF6NM Nurmberg on 29th December 2001. I hadn't set a space between the repeated callsigns and so the C is run into the following G!

G3KEV's stronger signal using 10 second long dots on 135.922 kHz can be plainly seen.

 

A screenshot from OE5EEP in Vienna taken on the 29th December 2001. My signal can be clearly seen as the top trace. The YXM of the lower trace is that of G3YXM.

 

Again my signal at the top of the screen as received by W1TAG on 31st December 2001.

Results of my 2001/2002 tests.

I transmitted each night over the period 28th December 2001 to 2nd January 2002 from about 1900Z to 0800Z with the exception of the night of the 30th December.

To my amazement, on every night which I transmitted, I was received in North America. VE1ZJ being in Nova Scotia heard my signal without fail. W1TAG in Massachusetts also received me consistently. W4DEX in North Carolina also reported receiving my signal although this was not solid copy. W4DEX's local noise levels appear to be generally higher than those of W1TAG or VE1ZJ, presumable because of his relative geographical proximity to the Caribbean and the Tropics where storm systems were known to be active at the time.

Back to the G4FTC Home Page

Page last updated 8 Jan 2002