The GM3VLB "Special" - a short multi-band vertical antenna

See also the GM3VLB Multi-Band Inverted-Vee Dipole
the GM3VLB Mini-Delta
the SCOTIA Bandhopper Multi-Band Mobile
the SCOTIA Poor Man's Antenna Analyser

and the GM3VLB Expedition Equipment List

Most of the author's operating takes place either /M or /P, using homebrew antennas ranging from his multi-band inverted-vee dipole to a variety of relatively short verticals. His primary mobile antenna is a Webster "BandSpanner", acquired second-hand, and free of charge, over forty years ago. This antenna has adorned many different vehicles and is still in regular use on his current Ford Sierra Estate.

André believes that it is arguably one of the best mobile antennas ever produced and rues the fact that the company appears to have ceased trading in the late 1960s. He will happily pit the "BandSpanner" against any other commercial amateur mobile antenna (some of which cost a small fortune), not only in terms of performance but also in its ability to sit quite happily, hour after hour, giving no cause for concern, even on European motorways at speeds sometimes well in excess of the UK legal limit . The author finds it amazing that no other manufacturer has produced a mobile antenna of similar basic engineering design, which performs equally well (with no coil-changing needed) on all WARC bands from 10m to 80m. VK and ZL have often been worked on 80m on this antenna whilst 'early morning mobile' heading north to activate Scottish islands. [The lack of a remotely comparable mobile antenna has recently prompted  Andre and Alex GM0DHZ  (winter 2006/7) to design and successfully test an innovative mobile antenna, the proto-type being called the SCOTIA "Bandhopper". Like the BandSpanner, it uses continuous tuning, but in a completely new and ingenious way - details elsewhere].

The BandSpanner screws into a standard 3/8" UNF base, mounted level with the lower edge of the rear windscreen and is held fairly rigidly in position by a plastic "arm" to a single rear-mounted roof bar. A small ex-CB meter on the transmission tunnel permanently monitors SWR A ratio of 1:1 is easily obtained on all bands at resonance, by adjusting the length of the telescoping high-tensile stainless steel whip top section (which has band markers etched into the steel). Fine tuning is by means of a home-brew toroidal base-matching transformer mounted inside the tailgate. This provides twelve impedance transformations (although, in practice, six would be adequate) in the range from a few ohms to 50 ohms, selected by a switch whose shaft protrudes unobtrusively through the black plastic moulding housing the tail-gate lock. André does not believe in using lossy ATUs if they can at all be avoided. With resonant antennae in a /M situation, they should be totally unnecessary. At best, they provide under-carpet heating. With the afore-mentioned method, band changing or tuning takes place in moments. Good earthing is essential, especially in the above estate-car situation where the tailgate hinges and locking mechanism normally provide the only earthing. Additional substantial, flexible copper braid is used for proper bonding (as he had to in the past, when antennae were often mounted on metal bumpers secured to the vehicle body by rusty bolts / brackets). There is virtually no noise pick-up and such noise as there is, is easily removed by the excellent noise blanker on the Kenwood TS-50S.

However, the primary purpose of this article is to show that simple yet efficient and effective vertical antennas suitable for either /M or /P, can be home-brewed with minimal cost and effort. When André first became involved in activating Scottish islands, amateur rallies were a good source of 8' ex-CB fibreglass whips, which were easily converted to single band, centre-loaded amateur band antennas. All of these easily equalled or out-performed miscellaneous well-known commercial mobile antennas acquired or tested at various times, and which, as a result, have all long-since been disposed of.

Anyone contemplating the installation of a mobile antenna would be well advised to read the articles by Mike Grierson G3TSO, which appeared in Electronics Digest (winter 1987) and RadCom (February 1988) and in which he challenges some popular myths. These articles were clearly based on solid research and experiment and GM3VLB considers them to be definitive articles on mobile antennas. Many, including himself in the past, have put too much faith in the Mobile Antenna section of the ARRL Handbook - this seems to never have been updated and has been reproduced word for word, from revised edition to revised edition, and presented as gospel for half a century or more! Furthermore, the information is very limited (as indeed is that in the RSGB's Radio-Communication Handbook). It is worth quoting part of the last sentence in G3TSO's RadCom article: "Commercial antennas vary from good to bad, and in general will only work as well as a home-made antenna . . . "

In contrast with the Webster "BandSpanner", the principal disadvantage of typical centre-, base- or indeed helically-loaded "whips" such as the "G-whip", "Hustler", "Pro-am"-type antennas, is that they are generally designed for single-band use. This results in much screwing or unscrewing of whip sections, loading coils or "resonators". What was needed was a single vertical antenna that could not only be mounted either on the vehicle or at ground level on a simple mount, but would be capable of efficient radiation on all WARC bands from 10m to 80m.

GM3VLB's design of such a relatively short vertical has gradually evolved over the last 10 years or so. It has been thoroughly tested, with considerable success, in locations that include nearly 160 Scottish islands, half a dozen IOTA islands off British Columbia, and various locations in the Pacific, the Yukon and the North West Territories. The Mark 1 version was even used by the author's son VP8NJS at Patriot Hills Base in Antarctica over the millennium, and a more recent version, which telescopes from 4m to 70cm (thus easily fitting overhead lockers on aircraft - prior to '9-11') was used as the only antenna on the 2001 round-the-world trip which included OC-121, OC-016 and OC-019.

It was used not so long ago in VE3 by Jack, G0SNV, and was again used by GM3VLB on OC-121 in 2003 and again as a "back-up" antenna on his 2005 expedition to the Pacific and VE7 IOTA islands with Alex GM0DHZ and son Niall VP8NJS.

The basic design centres on the use of a 4-section, 4m-long fibreglass fishing rod. These were originally purchased in French hypermarkets (very cheaply, at around £1 per metre for the shorter rods), but are now available in the U.K. The SCOTIA team would recommend Sandpiper Aerial Technology for both quality and value). GM3VLB has also produced various similar models for more efficient use on the LF bands, again using fibreglass rods ranging in length from 5m to 11m!

All antennas can be mounted on the tailgate mount of the GM3VLB's Ford Sierra Estate (even the 11m high 160m version, although it has not yet been used in anger, and certainly not /M!). As the majority of Scottish islands are inaccessible to vehicles,earlier verticals were mounted on the ground-mount assembly. This consists of a small base plate supporting a CB-type 3/8" UNF socket, a BNC socket and a brass wing nut to attach the radials. Using whatever ' plumber's hardware' is to hand, this is supported by, but insulated from, a ground-post about 50cm long made, in the André's case, from old tank-whip sections (light but strong) with a pointed end (e.g. a nail) to ease insertion. More recently, the 5m "Islander" has become the mainstay of all island activation.A recent development is a clip-on baseplate with provision for direct connection of the coax feed-line and the radial system. The vertical can be mounted directly on the ground, either using a  "ground-spike" as above, or held in place by a simple, triple guying system.

Although seldom required, a small, home-brew, base-matching transformer is available, mounted in a small ABS plastic box (75 x 50 x 25mm) with appropriate BNC plugs and sockets.

There can be not the slightest doubt that one of the most useful items of equipment ever produced for the antenna experimenter is the MFJ-259 type of antenna analyser , one of which accompanies GM3VLB on ALL expeditions. As this is still a relatively expensive item for many amateurs, GM3VLB has recently designed an ultra-simple/cheap  version (based on an original design by Jim Tregellas VK5JST) which it is hoped to describe shortly in RadCom and on this web-site.

GM3VLB also has available a very small, homebrew, T-match ATU consisting of two 200pF trimmer capacitors (modified to have small 1/4" shafts) and a switched toroidal inductor. This can be used, if necessary, to match an antenna that may be slightly off resonance. It was used for example in the N.W. Territories (VE8/GM3VLB) when, with an outside temperature of -20°C or lower, it was found that the 4m version, already adjusted for resonance on 15m, could also easily be tuned for use on 10m, 12m, 17m and 20m without having to venture outside!

As stated earlier, longer versions have been built, but most designs have centred on the 4m fishing rod. The original Mk1 version used one 'fixed' coil for 20m, and two separate 'plug-in' coils for 40m and 80m, whilst some later versions use a tapped loading coil that was wound on a length of 32mm diameter yellow (or blue) plastic water/gas (?) pipe which slides over the fishing rod, to sit on the first joint.

A wire, bonded to the outer tube (with two-part epoxy resin, fibreglass resin or similar), connects the 3/8" UNF screw to the base of the coil. This coil (107 turns of 20 SWG enamelled copper wire) has two taps, one at 72 turns (which, when shorted, leaves 35 turns for 40m) and the other at 100 turns (which, when shorted, leaves 7 turns for 20m). The whole coil (107 turns) is used on 80m. (The exact number of coil windings will obviously depend on the precise diameter of the coil former used).

The radiating element was initially a length of very thin (1mm), 7-strand Teflon-covered wire hanging from the top of the rod and plugged into the top of the loading coil. This was used on all three bands. Other shorter wires could also be used, with no loading coil, to resonate on the 10, 12, 15, and 17m bands, and with the 40m coil, on the 30m band.

More recently, the coils have been wound directly onto the top of the outer base section

The radiating element can also be made to be continuously "adjustable" by using bare, fairly flexible, thin multi-strand wire wound in place of the steel tape, onto the reel of a modified tape measure. The modification includes a simple but efficient sliding contact arrangement. By using bare wire, this eliminates the inductive effect one would have with a reel, or coil, of insulated wire. With this arrangement, and the MFJ-259, band changing becomes extremely simple and accurate. A mini-pulley system is used to carefully pull the wire "back up" the fishing-rod when going down in frequency.

The most recent design now uses a 5m fishing rod. The silver-plated paxolin-mounted coil was salvaged from an ex-MOD 'roller-coaster' unit and modified with plastic end-plates, with holes cut to allow it slide over the fishing rod and thus sit at the top of the base section. Some EMC 'finger-stock' (obtained at an Amateur Rally) was bent and formed into a complete 'ring', which slides up and down the coil, shorting out a selected number of turns in the process. The 'tape-measure radiator' plugs into the top of the coil. A green (more UV-proof) soda drinks bottle (also with a Canadian connection !) is adapted to keep the whole assembly dry.

Yet a later 5m version uses a small home-brew roller-coaster type coil which allows rapid tuning of all bands from 10m to 40m. An additional fixed coil wound on the rod's outer casing, provides the additional inductance needed to cover the 80m band. More recently, and as described elsewhere, the 5m "Islander" has been introduced. This uses a novel continuously adjustable coil (similar to, but considerably smaller than the ex-MOD version above), and again covering 10m to 40m but also the whole 80m band using a plug-in additional coil.

As with his island multi-band inverted 'V', GM3VLB uses an approximately 45' length of 50ohm RG58 coax that, after correction for velocity factor, is a half-wavelength on 40m, two half-wavelengths on 20m, three on 15m and four on 10m. Previously,  4 radials were used. The radial system consised of two pairs, each pair connected to a U-shaped spade terminal clamped by the wing nut on the base-plate. They were made from 5-way computer ribbon-cable, trimmed back to be a 1/4 wavelength long on each band from 10m to 20m. The same radial system was used on 30m, 40m and 80m and in most situations appeared to work satisfactorily, although the preferred antenna for these bands remains the  GM3VLB's Multi-Band Inverted-V described on this site and elsewhere (e.g. RadCom, March/April 2005). It does however require a certain amount of space - not always available.

Just as, in GM3VLB's opinion, Mike Grierson G3TSO has made a significant contribution to our knowledge of mobile antennas, then so has the work on vertical antennas by Ralph Holland VK1BRH ("Short Vertical Antennas and Ground Systems", Amateur Radio 1995) and that of Kenny Silverman, K2KW, and his team. The extensive work of the latter can be seen on his web-site, in particular "Verticals for Contest Operations" (CQ Contest Magazine, March 1998) and "DXpedition Antennas for Salt water Locations - A Study on 20m Antennas". Both works are examples of research at its best.

GM3VLB has studied the findings of these two independent groups and feels that they reach several similar conclusions that he now tries to summarise:

1. Ground-mounted radials lead to very much reduced efficiency
A significant improvement in efficiency (from almost zero up to of the order of 30%) has been noted for heights above ground as low as 0.005-lambda and up to 0.05-lambda. In recent years, GM3VLB has increased the length of his original ground-spikes and/or the position of his base-plates, to allow the radials to be raised to 40cm or more above ground, representing approximately 0.005-lambda on 3.7MHz and 0.04-lambda on 28MHz).Garden canes with nails embedded at one end, are used to support the ends of the radials. Whilst no detailed comparative tests have been carried out, we have a very strong feeling that these "raised radials" have indeed given better results....

2. Large numbers of radials are unnecessary
K2KW and his team have primarily used only two radials on their CQWW winning verticals, even on 160m. Here again, the SCOTIA team have not carried out detailed tests, but have adopted the principle that there is little or no advantage to using any more than two radials, the same two radials being used on ALL bands. Whenever posible, these are mounted at 180 degrees to each other.

3. Short verticals, ground-mounted by the sea, can out-perform full-size mono-band Yagis
For angles of radiation below the so-called pseudo-Brewster angle (~12º), the sea-reflected waves are in phase with much of the "direct" waves, thus giving rise to reinforcement. This results in appreciable gain at low angles over the Yagi, which even if several wavelengths high and multi-element, has a relatively deep null at such low radiation angles. In practice, an expedition-type multi-band Yagi is unlikely to be mounted at a height greater than 1/2-lambda, even on 28MHz, resulting in the angle of radiation of the main lobe being above the pseudo-Brewster angle (in fact, over certain paths, K2KW reports that on ". . . many occasions . . . a signal was S-9 on the verticals (on the FT-1000-MP) and S-0, and almost unreadable, on the horizontal Yagi").

4. Shorter radial lengths may be adequate
The effect of radial length is less clear but one interpretation is that short radials may be satisfactory for short verticals. Radial lengths in the range 0.1-lambda to 0.15-lambda are suggested, with 0.25-lambda offering little advantage, except perhaps with 1/4-lambda radiating elements. It had been the intention to carry out comparative tests, but once again, considerable experience with a single pair of short radials tends to confirm the "short radial" idea. A 0.2-lambda radial on 14MHz for example, will be approximately 0.1-lambda on 7MHz. It will be 0.4-lambda on 28MHz. It must be said that as conditions on 10m have been very poor since we implemented the "short twin radial" concept, we have not been able to thoroughly test the performance of radials longer than 0.25-lambda. Indeed, Alex GM0DHZ had some difficulty (during his pre-3V8SS expedition tests) tuning a "Traveller" using 0.4-lambda radials. Maybe we need to do some further tests before 10m begins to open up again.

5. Distance from the sea/land boundary affects gain
There is evidence of an increase in gain of about 3dB as the vertical antenna is moved back to about 0.25-lambda from the sea/land boundary. At 0.5-lambda the gain drops to -2dB, rising again to +2dB at 0.75-lambda No figures are given for greater distances, but in terms of choice of site, the interpretation is obvious (even taking the tide into consideration, operating close to a cliff-edge above deep water, would seem best!).

The proof of the pudding
In 1997, the 6Y4A CQWW CW team, using ONLY verticals, narrowly missed the N. American record with their claimed score of over 31 million points! The team subsequently operated, not only with great success as KH5K from Kingman Reef, but also more recently from Jamaica with QRP.

Although the results obtained with the various versions of the GM3VLB multi-band vertical have always been extremely gratifying, despite the fact that the radials were initially always simply laid on the ground, we have now adopted the "raised radial" technique. Whilss we cannot irrefutably prove the results now obtained are better, we have a strong feeling they are - they are certainly no worse. With Andre's back-ground, he appreciates scientific method, but regrets to say that once on an island, the high QSO-rate takes priority over setting up a proper antenna test-site! The intention is always there, but basically, if we are being called, the antenna is working!

We should perhaps admit that there have been one or two occasions when, inexplicably at the time, it proved virtually impossible to load up the vertical antenna on the 80m band. One instance was on Isle Martin (CN31). On that occasion, the MFJ-259 analyser was then not yet available, but it was found that several "temporary turns" added to the loading coil, allowed some power to be radiated. K2KW reports similar problems. If K2KW is correct, then raising the radials might have solved the problem. More recently on Uyea (SI24), serious de-tuning on 80m was eventually put down to unavoidable proximity to a large corrugated iron shed, as a few hours later on the next island, the same antenna behaved perfectly normally!

Conclusion
It would seem the advice is simple. When operating /P from islands, forget about climbing to the highest point. Forget about cumbersome Yagis, rotators, amplifiers etc. Select a site as close to the sea as possible, ideally surrounded by water (Kingman Reef would seem ideal!) and use a vertical (0.25-lambda or less) with a couple of radials, preferably raised and no longer than the radiating portion of the antenna.

Looking back, GM3VLB now realises why certain sites have provided nothing short of outstanding results. In all cases, they were either the ends of long piers or narrow spits of land jutting out into the sea. Next time you are /M on Jura, try the end of the pier at Craighouse, or the one on W. Burra in Shetland, or at Broadford on Skye - or the many other similar sites available in Scotland. Frequently in the past, André has found that his multi-band verticals performed better when vehicle-mounted than when ground-mounted. Could this be because the vehicle body is in effect a raised radial system, whereas he has was always used to laying his radials directly on the ground?

GM3VLB recently acquired a 'monster' 11m long roach-pole. If a 4-metre rod can be used effectively on 80m, might an 11m one offer reasonable performance on Top Band? After all, many "mobileers" obtain quite pleasing results on this band with the ubiquitous "8-foot loaded-whip". Noting that a recent CQWW CW 160m contest was imminent, a makeshift support was crafted, allowing the 'monster' to be attached to the fence between his gable-end and that of his neighbour. For most of its length, the rod was less than 3m from both gable-ends. A temporary "roller-coaster" type coil was suspended at the base of the rod that was about 2 metres above ground level. Following K2KW's suggestions, only two radials were used (each about 10 metres long, the same length as the vertical section), one running to (and finally down) a 2-metre pole at the pavement end, the other running along, and in contact with, the gable-end wall. Both were over tarmac - hardly the ideal DX antenna!

The operation was spread over 3 short sessions, one early on the Saturday morning, and one on each of the Saturday and Sunday evenings, and totalling about 8 hours. The object was to contact as many countries as possible on 160m without recourse to the DX-cluster. All "new ones" were to be found by listening (as we used to do in the old days!!). The station was a Kenwood TS140-S, a mini-ATU and home-brew keyer. Output power was less than 100W

The conditions did not appear very good, the same two-dozen or so "big guns" being heard each "session". A total of 57 QSOs were made in 26 different DXCC countries. No doubt some, with their sky-high monster antennas and their kilowatts of ERP, will think this is pretty pathetic, but André was very satisfied (back in the 60s and 70s, it took him years, even as 5Z4KL, to work 26 countries on 160m and 3 years to make WAC 160m!). The pleasure was all the more when on the Sunday, Alex GM0DHZ (his regular island partner) informed him that the "SO6Y" which he had got "first call" and thought (in his DX ignorance) was Poland, might be Western Sahara! Hearing that station make frequent, unanswered, lengthy CQ calls on the Sunday evening, GM3VLB called and worked him again! 7 other countries (including the USA and a UA9) were also heard but not worked. Next time this antenna is tested, it will be under island conditions. GM3VLB has high hopes that this very simple, very light, very easily erected 160m vertical will then give an even better account of itself.

In 2003 André activated OC-121 and OC-019 again, prior to teaming up with son Niall (VP8NJS/GM6GMZ) and friend Alex (GM0DHZ), to activate several VE7- Islands / NA- Groups. Despite extraordinarily poor propagation, the vertical antennae used on the expedition worked well. However, there remains plenty scope for further experimenting with these and other antennas. Please look for us, especially on 160m and on CW, which Alex intends to give a bigger airing to in the future.

Details of all known intended activity will be posted on this web-site, or via e-mail via our extensive mailing lists (if you wish to be added, please e-mail 5z4kl@ukonline.co.uk). In the meantime, please feel free to contact him regarding any aspect of these and other antennas described elsewhere (see the links below), and also for details of any particular item of associated home-brew hardware.

See also the GM3VLB Multi-Band Inverted-Vee Dipole
the GM3VLB Mini-Delta
the SCOTIA Bandhopper Multi-Band Mobile
the SCOTIA Poor Man's Antenna Analyser

and the GM3VLB Expedition Equipment List