Experimental 136kHz Vertical Antenna at G3YMC

Introduction

Having used my loop antenna very successfully for a couple of years on 136kHz I was curious to see how a vertical antenna would work at my QTH.  Most of the active operators on the band use either a vertical or Marconi style of antenna, and many consider this is the best way to get out on the band.  As you will have read, I am unfortunate in not only having a small and amateur radio unfriendly QTH, but also very lossy sandy ground which makes the performance of any antenna depending on a low earth resistance of questionable performance.  My experience in the past on 160m confirmed this feeling - a relatively small vertical which worked wonderfully at my previous QTH was next to useless on Top Band at this QTH.  Would 136kHz be any better?

I was very aware that the presence of nulls in my 136kHz loop was limiting my ability to have QSOs with some of the operators, particular towards the continent.  It was not readily feasible to rotate the loop to change the situation, and it was hoped that the vertical, if it did not increase performance in the directions already covered, might fill in these null gaps.  I also had a lot of encouragement from other operators to at least give a vertical a try.  Some were rather more blunt than others, and there still seems to be a sense of disbelief that I am able to get out at all with a loop antenna, such is the scorn these antennas are held in some quarters!

Thus it was that in summer 2001 I worked out a suitable vertical design and set to work to construct it.  Progress was rather slow because of other commitments, but by the autumn it was complete and I was able to have some QSOs.
 

Configuration

G3YMC Vertical The vertical at the G3YMC QTH is based on my old Butternut HF2V 40/80m vertical.  This forms the vertical section of the antenna, and is free standing at 10m high.  Additional top loading increases the capacitive impedance.  It is resonated at the base with a 6.5mH loading coil, and impedance matched to the 50 ohm coaxial input using a matching transformer wound on a 58 mm OD 3C85 core ferrite ring.

The Butternut as used on 80/40 is of tapering aluminium tube construction, and quite flexible in its upper sections.  It is mounted on a ground spike and insulated from this with a composite plastic base insulator.  It amazes me that this insulator supports the whole antenna, but it has survived many a storm in the 15 years it has been up and apart from a bend in the direction of the prevailing wind, is in remarkable condition.  Existing loading coils for 80/40 are mounted at the 1m level, with a further insulator at that point.  These coils are retained, in series with the main loading coil, but are relatively small (c 100uH).  The 80m matching capacitor and and its coupling strap have been removed.

Calculations showed that I would need some form of top loading in order to resonate the antenna with a reasonable size loading coil. The bare Butternut has a self capacitance of around 90pF and would require 15mH for resonance, an unrealistic value.  However it was feasible to attach a 12m horizontal wire at the top and this would increase the self capacitance.  Theory indicates around 5pF per metre for such a top loading wire, however the exact figure depends on many conditions.  It was thought that the Butternut was too flexible for just this single wire, so a couple of short back guy wires were also used.  Sloping loading wires are not normally recommended however as they can reduce the effective height, so these were deliberately kept short.  Initial measurements found that 6.5mH were required for resonance, giving an effective capacitance of 210pF.  It seems in my case I am getting rather more than 5pF per metre!
 

Loading Coil

Various configurations of loading coils have been used by 136kHz operators.  Considerable information on loading coils and their construction is to be found on GW4ALG's site and I made much use of the information there.  A convenient program, solonoid2.exe, to calculate the inductance of coils can be found on Reg G4FGQ's site.

It was decided to construct the loading coil on plastic domestic 2 gallon buckets - these have tapered sides with a mean diameter of 230mm and a usable winding length of 220mm.  Using solenoid2.exe it was readily established that 135 turns of 16/.2mm pvc insulated wire would fit on a bucket and have an inductance of 3mH and a series resistance of 3.8 ohms - conveniently the amount of wire required is just under one 100m reel.  So just over 2 'buckets' of inductance would be required  (Like one used to measure capacitance in Jars, you can now measure inductance in Buckets!).  There seemed little point in using thicker wire to reduce the series resistance, as this was expected to be swamped by the ground losses.

Two 'buckets' were constructed.  One of these was tapped every ten turns to provide a means of coarse adjustment, the other was untapped.  In order to increase the inductance of the series combination the two buckets were loosely coupled by placing the upper bucket slightly inside the lower one - by moving the bucket in and out of the lower one the combined inductance could be varied.  In conjunction with the taps on the bucket, the correct inductance could be easily obtained.  The buckets were held in position with a compressed wad of paper - Heath Robinson approach.
 

Matching Transformer

The impedance at resonance at the bottom of the loading coil will be resistive and (unless I was very lucky) more or less equal to the earth loss resistance.  As will be seen later, this was around 300 ohms.  Matching to 50 ohms was via a transformer, described in more detail on GW4ALG's site.  Transformer coupling is convenient and also serves to isolate the antenna earth from the shack earth.  At the moment the turns ratio has to be changed at the transformer itself, but in due course I shall have taps selected via 4mm plugs and sockets.
 

Loading Coil Housing

The two series connected buckets and the matching transformer conveniently fit into a standard plastic dustbin, whose lid provides easy access and is watertight.  The feed cables are fed via small holes in the sides and sealed with silicone compound.  A domestic dustbin does not even look out of place in my garden!

Initial Measurements

The impedance of the antenna was measured in the shack using the 3M Impedance Z bridge method as described by Peter Dodd G3LDO in the Low Frequency Experimenter's Handbook (RSGB) - more information on this method and a DOS program to use with it was available from Peter's site but that site is now no longer available.  By this method it is relatively easy to find the rough resonance of the system and its impedance at resonance.  Thus it was found that approximately 6.5mH was required for resonance in the 136kHz band.  The impedance however was considerably higher than desired, at around 300 ohms, so a rather large turns ratio was required on the transformer to obtain a 50 ohm impedance to the transmitter.  Most operators can obtain 100 ohms impedance relatively easily, and there are some who are lucky enough to get it down to around 20 ohms.  Clearly my suspicions about the ground conditions were justified.  Various methods were tried to reduce this impedance, including adding earth stakes and radials, and direct connection to the house water pipe, but it remained stubbonly at 300 ohms.  With winter approaching it was decided to leave it at that and get some reports.
 

Transmitter Modifications

When the antenna was connected to the transmitter it was impossible to obtain a good SWR even though the impedance measurements indicated a good 50 ohm match.  It was realised that harmonics from the transmitter were being fed to the antenna and upsetting the performance of the SWR meter - indeed, looking at the transmit waveform indicated all was not well.  It had been found that only a crude low pass filter was needed when driving the loop as that has a very sharply defined resonance - the vertical is much broader and requires a much better filter.

The output stages were redesigned to include a two stage low pass filter. At the same time the opportunity was taken to include a modified driver stage.  The circuit of the modified driver and output stages are shown here.  After these modifications had been done a reasonable (though far from perfect) match was obtained, and the output waveform was a far better sinewave.  As a bonus, it was found that the output power was increased to a remarkable 80W.
 

Initial Results

On receive performance seems to be very similar to the loop, however a few stations which were in the null of the loop are significantly stronger.  Stations such as OH1TN who were barely copyable on the loop are readable, though not strong, on the vertical.  However the noise level on the vertical, as expected, is somewhat higher, and this often means that copy of signals is not so good even though their signals are stronger.  There is not, in general, a big difference in any direction.

I have had a few QSOs with the new antenna, and have been able to obtain comparative reports.  In directions where the loop performs well, ie to the north and south, signals are either very similar on the vertical or noticeably down.  The relative signal reports seem to correlate very well with the difference in the receive signal from that station.  The only cases where the vertical performs better than the loop are those in the null of the loop.  G3OLB in Devon is quite weak on the loop, and no QSO has been possible with Tom on that antenna.  However on the vertical he is several s-points stronger and a satisfactory 449 QSO was achieved with no difficultly.

The tuning of the vertical is very broad, which is to be expected in light of the rather high earth losses. No retuning is necessary to operate over most of the cw section of the band.  However the matching varies considerably from day to day.  Droop in the top loading wire is one factor (this needs to be regularly retensioned), but there also seems to be leakage across the base insulator - matching is worse in the early morning when dew is present, and improves as it dries out.  Until I have a satisfactory method of adjusting the tapping of the matching transformer it is not easy to get round this, and I am only able to operate the vertical at certain times.

The performance of my vertical is clearly being hampered by a high earth resistance. It would seem difficult to overcome this, but I shall be trying various things over the coming months.  In the meantime I shall continue to have QSOs and continue to do comparisons. However initial results indicate that the loop remains my best antenna.
 
 

Operating 136kHz Equipment 136kHz main page  Home Page