Thursday, October 17, 2019

WSPR - Turning off MW Band Attenuation for 630m

About three weeks ago I decided to turn off the MW band attenuation in the Icom IC-7300 - this is a sub menu option in the settings that is on by default. Wow what a difference! I started spotting stations much further away from Australia, Hawaii, Alaska, midwest to the east coast.

I've been spotting these stations semi-individually on and off, finally got a good selection of them in one night:
  • VK4YB - Roger's 630m antenna is something else.
  • K9FD - on a former AM broadcast site with radials intact on an Island, dream location!
  • K3MF
  • K5DNL - holds the distance record for a QSO on 630m with VK4YB!
  • KL7L

The antenna I'm using on receive is the ZS6BKW with no additional matching, just an HF 1:1 current choke at the bottom of the ladder line, remote ATU (bypassed) then 10ft coax drop to ground where it runs around to the other side of the house. The other RX settings I configure when running WSPR with the IC-7300 are NB and NR off, notch filter off, AGC fast set to 0.3 seconds.

Since turning off the MW band attenuation, I would start spotting the closer stations in California as soon as it got dark through until sunrise like clockwork. First I thought it was propagation changing with the D layer absorption which quickly disappears at sunset and quickly returns at sunrise.

Turns out it's more than the D layer at work, there are several strong AM broadcast stations in my area, during the daytime they are very strong and overload the receiver in the IC-7300. At night these stations lower output power and/or change their antenna pattern, this reduces signal levels at my location to a level where the receiver can cope.

Screen capture of the strong signal on 850 kHz during daytime, S meter maxed out at +60dB. There are several more at +40 to +50dB.

630m band WSPR frequency during daytime, the S meter hovers around S9 +15dB. Band scope shows the "mess" these strong AM broadcast stations create in the receiver.

630m band WSPR frequency at night, the S meter hovers around S1 when AM broadcast stations reduce power and/or switch their antenna patterns.

Note: The time displayed on the IC-7300 clock top right is UTC.

Sunday, September 22, 2019

43ft Vertical 80m/40m Matching & K9AY Loop

Note: This post is being updated/edited as the project progresses. Once complete, and if successful I plan to re-write it.

For this coming winter my 80m DX antenna system will be base matching the 43 ft vertical for transmit and using the K9AY loop for receive.

Reviewing 160 and 80 Meter Matching Network for your 43-foot Vertical by Phil Salas - AD5X. Phil provided a parts list in his article which I found to be very useful.

I decided to keep it simple and just have matching for 80m. I ordered the coil stock, 4:1 unun (not sure I even need that for my setup), ceramic feedthru from MFJ.

Did a quick throw together of the components to verify that things will work as expected:


Rainy day hence the umbrella, screwed the coil to some wood, current choke is 7 turns of RG-8X through two mix 31 clamp-on ferrites. Not having micro alligator clips made this a bit fiddley, but I got fairly close resonance and SWR wise.

My cunning plan is to use the vertical as a support for the K9AY loop, mount the K9AY relay box inside the 8x8x4” electrical junction box containing the matching network for the vertical.

With a simple control unit in the shack, when the rig is un-keyed, power will be supplied to the relay in the junction box which will open circuit the vertical (making it non resonant), unground and connect the loop feed-points to the K9AY relay box. Figure this is safer way to do it, if the system loses power I minimize the risk of sending TX power into the K9AY relay box which will let the magic smoke out!

At the shack end I will have one coax for transmit and one for receive. Since I have an Icom IC-7300 which does not have separate receive antenna jack, I got an INRAD RX7300 receive adapter for the Icom IC-7300. This allows you to use a separate receive antenna or insert things like filters into the receive chain for example. If I loop it via an antenna switch, I can choose when to use the loop on receive.

Update 29 Sept: I got the matching network in basic form built and tuned:

K9AY relay box top right - not yet wired up with the isolating relays.

Can see the coax tap point on the front side of the inductor, the antenna end enters through the top and is tapped on the rear. In practice I got a strange SWR curve with two dips, one around 3.6 MHz, and another at 3.8 MHz. Seems I accidentally made it broad-banded somehow.

Once I got in the shack to see how things looked with a radio connected to the vertical with matching network, the noise floor on 80m was quite high, nearly S9, will need to get the receive loop up and running.

When I reconnected the ZS6BKW antenna (near by), the noise floor on 80m was lifted near three S-units - safe to say there is some interaction occurring. After disconnecting the matching network from the vertical the noise floor returned to its usual daytime level of S2.

For now I disconnected the matching network, when the relays arrive I should be-able to get this completed, and remotely switch out the matching network to eliminate the interaction between antennas.

Update 05 Oct: Relays and some mini test clips arrived. I decided I should model the vertical with the K9AY loop in its presense, it has a small effect on the vertical's resonance requiring 12 uH vs 9 uH in the model to match it, but no difference in pattern or gain.

View in MMANA-GAL with a 55ft loop (35% smaller than the standard K9AY), the smaller loop maintains performance on 80m through 40m. Larger loops with better performance on 160m lose directivity on 40m.

Update 20 Oct: Some progress on the matching network. Relays mounted partly wired. Re-did the antenna connectors with a copper plated steel strip (found in the plumbing section at Lowe's) for better grounding. Top left is a SPDT relay, the blue "thing" is two 5 Meg ohm metal film resistors in parallel to bleed static via the inductor which is grounded at the bottom.

Control cable for the K9AY box top right is an RJ-45 socket mounted bottom right, I decided to standardize my control cable using shielded CAT5 cable. Hopefully the AWG size isn't too light, a test with 150ft of very cheap CAT5 seemed to work ok. Under the RJ-45 socket is a DPDT relay to isolate the loop from the control box when transmitting.

Making the square hole for the RJ-45 and marking the location of the holes for the mounting screws. Actually got it all spot on, which is quite rare for me!

Whats left? Quite a bit..
  • Finish the relay power wiring.
  • Make a mounting bracket for the loop top support.
  • Find the tap points on the inductor with everything in place and put appropriate gauge wire in place.
  • At the shack end I need to build a keying interface for the T/R switching.

Update 27 Oct:

For the last couple weeks I keep thinking about adding 40m to this system, and modeling it all together to see if it'll work. If this works out it might make a nice compact low band DX antenna system covering 40m and 80m with reasonable performance that'll fit in a modest back yard..

Over the weekend I worked out how to control the relays such that if I'm receiving on 80m with the loop, the vertical's matching network is set for 40m making it non resonant to the loop. Inverting that when going to transmit, and reversing the logic of the system when receiving on 40m with the loop, and via a simple control in the shack e.g. a two position switch. And "failing safe" in such a way that if power is lost, the loop isolation relay defaults to isolating the loop and vertical's matching defaults to 80m.. If that makes any sense..

Anyhow, it can be done with a relay, switch, and a solid-state relay driven by the Icom 7300 PTT accessory line (already using one in the keying circuit for the AL-80B amp).

Circuit diagram I threw together using an online tool. RY1 and RY2 are shown energized as they would be per the control unit when in receive mode on 80m.

Reviewing the idea of using CAT5 for the control cable where the loop control needs 6 wires (3 pairs), leaving a spare pair to individually control the loop isolation relay and vertical matching network relay, 24AWG can carry around 1 amp based on what I can find. The closing current of the relays is 100mA each, the calculated voltage drop over 75ft run is a couple of volts, if I use a 15V supply in the shack for this we should be good there.

Lastly, modeling the whole system as one to see if resonating the vertical via its matching network on the band not in use while receiving (e.g. if receiving on 80m with the loop, select 40m matching on the vertical and vise versa) is sufficient to have no interaction with the loop's performance. Short answer yes, appears this will work! Model file: MMANA-GAL 43ft TX Vertical with K9AY RX Loop.

Update 12 Nov:

I've had the loop operational for a few weeks, so far things don't look very promising at my location:
  • 160m: With a 55ft loop vs an 85ft (common size), I've found several occasions where the loop is noticeably better vs a horizontal wire dipole in terms of SNR meaning they sound louder on the loop, often it makes the difference of hearing vs not. The horizontal wire is a ZS6BKW antenna at 45ft, which is horrendously short for 160m at 93ft so it's perhaps not a fair test.
  • 80m: Not found a case where the loop was better than the horizontal wire antenna, when I've found a station on the west coast working DX, the loop makes no difference, either way I still don't hear the DX. Several times in the evenings I have gone searching 80 and 75m for situations where the loop shows an advantage or even pulls out further away stations that cant be heard on a horizontal wire, no joy there either.
  • 40m: Horizontal wire always better. Listening a couple stations in California working VK early one morning, out of several I could only hear one VK6 who was not quite readable on the horizontal wire, was worse on the loop. I think it's safe to say on 40m the K9AY loop is of no advantage. Frustratingly the stations in Cali were giving the VK stations 5/5 to 5/9 reports.

At this point I'm pretty much ready to shelve the project, too much local noise, not enough space for good antenna arrays, and pockets not deep enough for 60+ ft towers and antenna systems to match. Another couple points I've realized are:
  • My preferred time to actively operate is in the mid afternoon on weekends, this is at odds when the low bands are open. I had a bunch of fun last weekend in the afternoon on 40m chatting with a station in Utah for bit.
  • The second is after listening to stations actively working low band DX from the west coast is the QSOs are short, often just an exchange of callsigns and signal report. This doesn't excite me so much. I get a much bigger kick out of using WSPR to test the limits of propagation across a far wider range of bands as I can set it going and review the results when I have time.
At this point I think I'm going to follow the general crowd and do what works with more modest resources, 20m and above for DX, 40m and below for regional / NVIS. Time to start researching and planning antennas to suit.

Saturday, September 21, 2019

Gain vs Wave Angle

Something I've noticed is that most don't consider is the whole picture when it comes to antenna gain. Often only peak gain at the wave angle it occurs at is quoted. Sometimes a pattern is shown with no gain figures which is useless, this is often the case with verticals.

To build a picture of what one might expect when considering low band DX antennas, I feel it's better to consider the gain at different wave angles.

To illustrate this, I created a table in a spreadsheet with several different antennas and the gain they have at different angles. LiberOffice Calc has the option to fill cell background color based on the value. The table is sorted based on the gain at 5 degrees.

I used MMANA-GAL for the modeling of these antennas. Most are 40m antennas, those that can be matched on 80m are also shown.

Wire antenna heights are 40ft/12m, the inverted Vs are 15m/50ft, the delta loops have a minimum height of 1m/3ft.

From this table I can see that one of the simplest antennas gives the best bang for buck for DX on 40, thats the half square. A small surprise is the ZS6BKW on 40m looks to have an edge over a 43ft vertical for DX! The OCFD antennas are well behind.

Disclaimer: I'm no expert at modeling antennas, I perhaps know enough to be dangerous :-)

Monday, September 9, 2019

ZS6BKW Antenna

One of the antenna projects I wanted to get finished off this summer was get the 40m OCFD up in a flat-top configuration.

Got thinking about how I'd lack an antenna that covers 75m NVIS, thought about an 80m OCFD again. At 135ft long, the current situation with tree shapes, sizes and locations leave that easier said than done, and not for a lack of trying with out resorting to chainsaws. Having said that, getting one of two 80ft Fir trees closest to the house removed is on the list, so I may be-able put up longer wire antennas in the future.

What else is there that's under 100ft, at a minimum covers 40 and 20m with out the need for an ATU, and will work on 75m?

After considering a few ideas, I took a closer look at the ZS6BKW antenna which is an optimized variant of the well known G5RV. This antenna checks the boxes on these good points:
  • Light weight - no traps, coax etc hanging.
  • Low SWR on 40 and 20m, can run ~1kW with the AL-80B.
  • Respectable gain on 20m if you have the luck or luxury of orientating the broadside to favor somewhere of interest.
  • Will work on 75m with the remote ATU I already have.
  • Also covers 17, 12, part 10 and 6m with low SWR. 17m opens occasionally during the current solar minimum, and there is the Puget Sound 10m net on Sunday nights I like to take part in.

Spent a couple weeks reading everything I could find, and playing with models - one is included with MMANA-GAL, others can be found online.

Decided it was worth trying out. Initially I was going to order the parts from The Wireman and build it my self, but found a source of these pre-made cheaper and better built for $60 (normally $90) plus shipping from Amateur Radio Supplies => ZS6BKW G5RV 80, so went that route. 

Current setup:
  • Antenna apex at 50ft, ends in inverted V configuration.
  • K9YC inspired choke balun - 7 turns of RG-8X through two mix 31 clamp on ferrites.
  • Remote ATU between the choke and coax to the shack. 

How does it work?
  • Except for 15m, low SWR on six bands from 40 to 12m, parts of 10 and 6m.
  • Remote ATU dials it in on 75, 60, and 15m, brings the total to nine bands.
  • 30m is the odd one out, ATU fails to find a match in my case.

Nine bands where only three need an ATU, with 93 feet of wire in the air is OK by me. 30m might be an issue for some, but you can't always win 'em all :-) 

Links to further information:

What's next?
  1. Get the ends up higher - the test inverted V setup I have has the ends coming down too steep and into the trees, as expected resonance / low SWR favors the low ends of the bands.
  2. Maybe after the trees are "trimmed" it's possible I could move the antenna to where the vertical is with the existing radial field. This would allow me to operate it as a T (top loaded vertical) by bridging the two legs of the ladder line together at the base, and feeding it against ground resulting in an ok 80m DX transmit antenna in place of the vertical, eliminating one antenna from my setup.

What about low band DX? Not much progress there, and with winter approaching need to sort something out. Putting up the K9AY receive loop and building an 80/40m matching system for the 43ft vertical, possibly returning it to an inverted L is the "quick fix" solution.

Monday, August 12, 2019

Bay of Islands ARC Repeater Site

Following on from the Auckland VHF Group Repeater Site post, some more pics from the archives. This time of the Bay of Islands ARC Repeater Site, photos I took from around 2007 when I had the task of reinstalling the equipment after it had been removed for re-alignment.

This site was established with the help of the Auckland VHF Group in the mid 90s to get the national link extended north via a sister site between here and Auckland called Brynderwyn. NZART map here - its approximate at best. Google Map that I created here, the locations are with-in tens of feet in most cases.

Since these photos were taken when I was last there, the 2m repeater has had some changes with the addition of CTCSS for IRLP node 6398 link, there are more recent pics on their homepage and some from when the tower was erected.

Pic of the tower and antennas. Like Klondyke, the 2m repeater antenna is constructed from angle iron, welded, and hot dip galvanized, but is a simple pair of 2 half waves in phase for around 10 dBi of gain, hard to see as they are end on in this pic. At the top is a 4 dipole stack for the the 70cm repeater. This configuration resulted in the 2m and 70cm repeater performance and coverage being closely matched in most cases. The link to Brynderwyn is a set of four 7 element Yagis with around 17 dBi? of gain, the Brynderwyn end has the same antenna configuration, this is needed because the 70 mile path is not ideal.

Inside the building.
Tait T300 gear.
146.750- also had a 50W PA in the back.
439.975- had the matching Tait 50W PA.

Thursday, July 4, 2019

New Desk

After using a folding table for the last few months, 'twas time for a something better.

I had considered The Great Ham Radio Desk Project, but after looking around IKEA I found something I liked better, more so since IKEA was having a sale at the time.

The spare room/office/shack is 120" wide across the back wall. What I ended up getting were three sets of draws and two table tops, the table tops are 59" x 29", the 29" width is nice, as it allows plenty of depth for radios, computers etc with space in-font.

The total was about $270 less tax and delivery fee, took an afternoon to put together.

Left to right, Ameritron AL-80B, Kenwood TH-D72A HT on top, Motorola MCS 2000 33cm/900MHz radio, Kenwood TM-731A, Icom IC-7300, 27" iMac (2014, 5K), under the iMac is a Windows 7 laptop which runs all the ham radio applications.

Plenty of room for more stuff! Plan to add some shelves to mount the Kenwood and Motorola mobile rigs, control boxes, and antenna switches etc.

Saturday, June 22, 2019

Summer 2019 Antenna Projects

40m OCFD: I built the classical 40m version with 33% offset noted previously. For a quick "throw it up and see" test I had it in an inverted V config, the SWR and resonance results are inline with expectations.

I fired up the Ameritron AL-80B a couple weekends ago with the OCFD. On 20m propagation was the worst I've come across, managed one contact into Missouri but had to use the Utah WebSDR to receive as signals were weak. On 40m I later made two contacts into Utah, conditions were tough with static crashes from thunderstorms.

Listening to my self via the WebSDR, the 10dB gain of the amp makes a big difference. Sounded clean and punchy with the Icom IC-7300 driving at 70W with the ALC connected. The T/R switching is via a solid state relay interface I built as suggested by KV5R.

Inverted L / 43 ft vertical: Restored to 43ft vertical for now, OCFD needed the only high support I have via a tree that the horizontal wire ran to. Interesting note with the vertical, when it was last simply a vertical I had the radials laying on the ground, the SWR/match was a bit "gimmicky" in on some bands where for example on 20m, the remote ATU failed to find a match. With them buried 2" to 6" deep things are much more normalized.

What's Next?

Get 40m OCFD up in flat top config, and dialed in. At near 50ft it will cover 40m for NVIS, 20m with a good pattern and low take off angle, plus 10m and 6m. I might move the feed point to 40% and gain 15m coverage.

I got some mast sections from The Mast Company, plan to install with an apex eave mount. Put the Comet GP-15 on the top.

40m DX Option 1. The mast may provide a support for one end of a 40m Half Square. Will favor SW / NE which covers the Pacific, and Europe / Africa in the other direction. More info regarding half-squares:

Far field plot for the half square I modeled.

40m DX Option 2. Two element vertical Yagi using a single 60ft support, I got the idea from this page - I have tweaked the geometry to bring it up to 7.15MHz.

I found additional information by N6LF - Getting the Most from Half-Wave Sloper Arrays.

There are perhaps 3 ways to feed this antenna:
  • N6LF settled on resonating both elements as reflectors, and end feeding the selected driven element - the parallel tuned circuit used to match takes care of resonating the driven element.
  • Borrowing from K9YC's End-Feeding a Center-Fed Vertical Dipole, the bottom half of each element would be made from coax with good chokes at the bottom which K9YC also discusses. Leaving one side open circuit resonates the element as a reflector - the length of coax is critical, I calculated (info below) there would be enough length to form the bottom half of the dipole, the choke and connections to switching relay. Some form of strain relief may be needed for the coax half of the dipoles to stop it from stretching.
  • Using the same principal as above but center feeding with coax (just like a conventional coax fed dipole) to a mast mounted switch that short circuits the inactive element creating a closed stub which resonates it as a reflector, the length needed is much shorter hence the mast mounted switch. A balun would be needed at the feed point which adds weight.

The latter two open/closed circuit stubs will present an inductive reactance lowering the resonant frequency of the passive element turning it into a reflector. How to figure it out? W8WWV wrote a nice article on Coaxial Cable Stub Q, which lead me to an application called Transmission Line Details. Among other things this can be used to calculate the length of open or closed stubs, it also tells you how much R the stub will add at the feed-point of the passive element - this information can be plugged into MMANA to know the effect of the added R.

Another handy calculator I found is Inductive Reactance Calculator, which converts an inductance at a given frequency into an inductive reactance value. Useful in conjunction with MMANA-GAL and TLD.

It's important to note that properties of coax varies between manufactures and manufacturing runs. The length is critical, one would need to measure the velocity factor of the coax on hand using an antenna analyzer, not rely on published information. I suspect this is the reason why some struggle to get phased arrays to work properly, relying on the publish spec is not close enough to reality.

Personally I think N6LF's end fed might be the better choice as it eliminates the weight of coax, baluns, mast mounted switches etc. If using high power, capacitors need to handle several kV. The only critical part is resonating both elements as reflectors.

From the model, I like this antenna for a couple reasons, it has a nice broad beam width, and 2 dBi of gain at 10 degrees elevation, the F/B is ok at 13 dB. It can also be built using a single support.

MMANA-GAL file for 40m 2 element vertical Yagi.

View of 2 element vertical, these are half wave dipoles with the ends angled towards each other. 

Far field plot for the 2 element vertical.

3D far field plot from GAL-ANA.

What I like about this is it can be built with a single 60ft support. Has good gain at low angles, 2 dBi at 10 degrees, -3 dB beam width is slightly more than 120 degrees.

40m DX Option 3. Update 08-Aug-19, attempted to build this with my 43 ft vertical, however the vertical is too flimsy to support the tension needed on the passive elements to get them into the right shape. I'd need to rebuild the vertical with longer sections (less taper) and guy it. I did enjoy the time spent researching this antenna, modeling it, and learned a few things along the way. I'm disappointed that I cannot make a good attempt at building this antenna. Maybe another day!

Another variation of vertical Yagi from is the 3 element GP80 scaled for 40m.

Very close resemblance to the Spitfire antenna developed by K1VR and W1FV - It also appears in ON4UN’s Low Band DX book, and N6LF's Single Support Gain Antennas for 80 and 160 Meters.

What I like about this is the 43ft vertical I already have with 32 buried radials is the right length for the 40m version driven element. Makes good use of what I already have in place. All I need to do is add the passive elements, and simple matching network with series capacitance.

This antenna design is kinda neat on a few points:
  • The driven element is around 1/3 λ this raises feed-point impedance to a convenient 50 Ω, and can be adjusted as needed to be 50 Ω, series capacitance takes care of bringing it to resonance
  • The extra length also supports the passive elements and their shape, can double as guy lines.
  • The passive elements are 1/2 λ and folded (and thus independent of the ground plane), shortening one via switch/relay in the low horizontal section turns it into a director, direction could easily be made electronically reversible.

View and far field plots of the 3 element GP.

MMANA-GAL file for 3 element GP 40.

Since I have the driven element already in place, modeling it by it-self showed feed point being 70Ω, jX 108 at 7.15 MHz. I went and measured it with the analyzer and to my surprise it's very close:

Analyzer shows 72Ω, jX 104, along with the series capacitance needed for resonance - 213pF. Adding series capacitance of 200pF to the model brought it to resonance. The test and result is encouraging.

What to do for 80m due to space and tree locations, I've considered many different possibilities. The best solution I can find is a quad loop squashed into a rectangle. Feeding from one of the sides results in bi-directional low angle vertical polarization. An 160m version of this antenna gets a mention in Low-Band DXing 5th Ed - Chapter 10, pages 4 and 5.

At the other end of the spectrum, the 7 element Yagi I have in the attic for 900MHz is sub-optimal, need to relocate it or try something different.

Saturday, May 25, 2019

40m OCFD

After previously re-considering verticals or dipoles, I decided to build an OCFD using ON4AA's design and ordered a Model 4116 balun from Balun Designs LLC.

To start with, I have decided to keep things simple by comparing verticals and dipoles on 40m. I could put up a simple 40m dipole, but had this balun already.

Looking at 40m OCFD designs, none seemed to be that great. SWR over 2:1 on 40m, or different offsets to get 15m resulting in worse matches on other bands.

Having recently acquired a used Ameritron AL-80B from VA7ST, I also now need an antenna with a good match on 40 and 20m that will take some power - up-to 1kW PEP. My LDG remote ATU is rated to 600W PEP, at high(er) power levels I'd prefer to have properly matched antennas available.

I also bought An Introduction to Antenna Modeling with my $10 birthday coupon from the ARRL, this has good info for getting started with antenna modeling. With some ideas in mind from the book and what ON4AA described with their 80m center loaded OCFD design, I set a design goal of a good match (under 1.8:1) on 40 and 20m..

The process I followed:
  1. MMANA-GAL, using real ground, and a height of 12m / 39 ft (my situation).
  2. I started by finding the resonant length for a 20m dipole at 14.2MHz = 10.45m.
  3. Double the length to 20.9m.
  4. Tried various offsets between 17% and 40%.
  5. Found 33% is the best compromise between 40 and 20m.
  6. An added bonus is low SWR on 28.5MHz and 50.1MHz.


07.15 MHz, R 144, jX 74.6, SWR 1.72, Ga dBi 5.58, Elev 51.1
14.20 MHz, R 122, jX 2.39, SWR 1.64, Ga dBi 8.12, Elev 24.5
28.50 MHz, R 162, jX 14.8, SWR 1.25, Ga dBi 9.44, Elev 12.3
50.10 MHz, R 202, jX 32.1, SWR 1.17, Ga dBi 11.2, Elev 7.00

I have rounded the numbers above slightly for better formatting. Now, time for some pretty charts:


MMANA-GAL file for the 40m OCFD with 33% offset.

Next step, build it and see!

Side Notes:
  • The averaged R across the four bands is 157 ohms, using a 3:1 balun would get [in theory since this is a model after all!] a very good match.
  • On 40m, resonance is just below the band. Borrowing from ON4AA's center loaded design using a capacitor in the center to resonate the antenna on the lowest band, 425pF did the trick in the model.
  • MMANA-GAL file for 40m OCFD with 33% offset and center loading.
  • MMANA-GAL file for 40m OCFD with 40% offset, this gains 15m band coverage at the expense of higher SWR on 40 and 6m.
  • More experimentation may improve things further..

Saturday, April 27, 2019

WSPR - 40m, ZL5A Scott Base Antarctica

WSPR pulls out another surprise.

Adam Campbell / ZL5A from Scott Base Antartica:

Wednesday, April 17, 2019

Vertical or Dipole?

Vertical or horizontal dipole? Excellent question, since I began my low band DX quest, I had assumed horizontal dipoles are NVIS or short hop only antennas, due to the effort required to elevate them 1/4 wave length or more above the ground.

A few days ago I started looking at phased verticals to make some gain, and had settled on a plan to build two 80/40m trapped verticals based on W8WWV - Hex Array - 80/40 Meter Vertical, spaced 33 feet apart - 1/8 wave on 80, 1/4 wave on 40, and setup phasing for 135 degree phase shift on 80, and 90 degree phase shift on 40 with the option to reverse the phase to switch directions.

While scouring the internet for information, I ran across Verticals: got two? by N4JTE. Ok I thought, Bob had success with this, then I took at look at Bob's QRZ page and discovered that Bob had discovered something even better, the same thing but horizontal!

A while ago I noticed something interesting when comparing radiation patterns for verticals and horizontal dipoles, on first glance one concludes vertical has all the power down low. But when I compared the actual gain figures at 10, 20 and 30 degrees elevation, the dipole, even when under 1/4 wave length high has more gain. I dismissed it at the time thinking, I must be overlooking something.


I started looking into this again and found:

Above from comparing antennas on 15m:
  • A very highly optimized vertical. It uses 16 elevated radials 28 inches off the ground.
  • Horizontal fan dipole (0.3 dB better than regular dipole) at 15 feet above ground, 15ft on 15m band is between 1/4 and 1/2 wave length.
What's eating the vertical's energy? That null up top should be pushing more out at a low angle.. Ground losses against the return currents to the base of the vertical? Only when verticals are over salt water does the low angle lobe "push" right out.

Now what? I'm not tearing down the inverted L just yet, but I'll be building an OCFD based on the ON4AA design with the center loading network for 80m. The balun I'll be using is Model 4116 - 4:1 Hybrid Balun 1.5 - 54MHz 3kW from Balun Designs LLC. I have built OCFDs before with Balun Designs OCF optimized baluns, and had pleasing results.

The fun part is I'll be-able to compare the two antennas for a while.

After taking a ZL-Special model in MMANA and dropping it down 1/8 to 1/4 wave length above ground, it suggests the gain will be several dB better than a vertical at low angles.

Ultimately I'd like build something similar to Bob's double whammy with 40 and 80m capability. From my location in the Pacific North West, orientating it for South West / North East it'll favor the Pacific, ZL/VK, Europe and Africa.

Another advantage of horizontal polarization = less local noise pickup, will the RDF exceed small receiving loops? Since I have a K9AY loop control box, I'll also be-able to compare these too.

Saturday, April 13, 2019

The 900MHz Club

After getting to know a new group in my area on the local N7OEP 70cm/6m linked system, there is occasional chatter about "the 900MHz club". Otherwise known as the 33 centimeter band.

ZL has an allocation from 921 to 928MHz, commercial two way radio gear is non existent that I know of except for the digital Tetra system around 860MHz. If there was analogue FM stuff around to modify, the commercial 25MHz repeater splits are too wide for the band. As far as I know, there is no activity on this band in ZL bar maybe a mad scientist or two frankensteining something together in their basement shacks. Those into things above 70cm start at 1296MHz and up into the microwave bands from there.

The US has 902 to 928MHz allocated, and here in Western Washington we have several 900MHz repeaters and some activity. Getting on 900MHz requires sourcing, reprogramming and sometimes making physical modifications to commercial 900MHz two-way radios.

I ordered two Motorola MCS 2000s from Used Radios, they shipped quickly, were well packaged and the two I got appear to be in good condition. The pair I got are the version 3 model that also make 35 Watts, 35 Watts at near microwave frequencies is rather respectable. Here in the Puget Sound area there are a couple of people happy to program the Motorola rigs. As for antenna's plenty of cellular stuff around that covers this band, cheap Yagi's can be found on e-bay and Amazon, or in the spirit of ham radio, make something! I ordered a couple 7 element 900MHz Yagis to try out at $16 each.


See Getting Started with 900Mhz (33cm) Ham Band by WA6AER for more info about whats what. The largest hurdle with this stuff is getting the radios programmed, either your self or by someone else.

Whats it like? Well like FM anywhere else except using a commercial Motorola radio gives things an industrial feel, no S meter, basic controls - on/off, volume, memory up/down, and a bunch of buttons that do nothing.

Time for something bigger and better than a folding table.

Monday, April 1, 2019

Feed Line Routing and Grounding

As with any antenna project, appropriate cable routing and grounding is fairly important.

After reading some very educational and informative articles written by W8JI:

I realized I had some work to do, originally I had run my HF antenna feed line up the wall on the other side of the house, through the attic and down an internal wall into the shack. I had installed a ground rod at the far end of the feed line where the remote ATU was located when I had the Carolina Windom up - with that setup I had created my self a ground loop, a lightning protection problem and an electrical safety issue - House Ground Layouts has several diagrams of how to get this wrong, and how to get it right.

With that I routed my feed line around near the electrical service entrance burying it couple inches underground. At the service entrance I repurposed an old PrimeStar box I found in a pile of junk when we bought the house, along with an old PrimeStar Ku band dish thats now hanging up in the garage awaiting a project...


The copper strip I found at Lowe's in the plumbing section, this had the right sized holes for bulkhead N-type connectors, its actually lightly copper plated steel, I gave it a light coating of Jet-Lube SS-30 Pure Copper Anti-Seize to prevent rust and so the connectors would make a good electrical connection. The right angle N-type connectors are a little hard to find at a good price, HRO had them for about $6 each. The green wire runs about 3 feet over to the electrical service entrance grounding rods - I plan to upgrade this to something better such as a copper strip or heavy gauge wire later.


Feed lines run up the wall into a junction box I got from Lowe's, it allows for enough of a bend radius with LMR 400 to then go through the wall. The shack is on the upper level. A little work left to tidy things up nice.

On the inside I used a standard 4 port Keystone wall plate designed to take the Keystone snap-in connectors such as the RJ-45 bottom right - I plan to use that with CAT5 cable for K9AY loop control cable. The N-type bulk head connectors were just right with the o-ring and washer removed to allow the connector on the other side to screw up tight. Looks nice eh?!

Bonus pic, while I was walking around with the camera phone, a bumble bee had flown through the garage to the window and buzzed up the spider web. Bumble bee made a lucky escape and flew away back out the garage door as I mounted a rescue effort. You can see a large spider made an appearance in anticipation of some lunch, in-fact at least two of these guys live here, there was another one to the right just out of view!

Friday, March 22, 2019

43ft Vertical to Inverted L

Last weekend I decided to upgrade the vertical to an inverted L, specifically a 1/4 wave on 80m.

Measuring the 43 ft vertical purchased from ZeroFive, it has 14 x 32" sections and the base section is 36" for a total of 484" or 40.3 ft - where is the other 3 feet? No idea! Running some rough calculations I added a 22 ft horizontal wire to the top, and after a fight with the maple tree snagging the wire a few times I got it up.

I was a little surprised at the changes, SWR:
  • 80m, 3.5 MHz is near 1:1 and rises towards the top end of the band.
  • 60m, SWR high, remote ATU fails to find a match.
  • 40m through 10m SWR high, remote ATU matches it in no problem.
Since I have no real idea as to what load the antenna is presenting to the ATU, its hard to know much more, to solve that I have a RigExpert AA-55 antenna analyzer ordered.

With that, I decided to bang my head on antenna modeling to help get a better understanding on what to expect from this inverted L I have created. Being brand new to modeling software, I found 4NEC2 "too hard". I took at a look at MMANA-GAL, which includes an inverted L file. After a quick edit of the geometry to match what I have, the results were interesting. Granted I might have missed an important detail with this, it could easily be totally wrong.

Starting with 80m and working our way up, blue is horizontal, red is vertical polarization:

80m, what I expected, the small horizontal component is from the horizontal wire at the top. An anecdotal observation is the small horizontal component has made receiving local signals via NVIS stronger.

40m, calculating the length on 40 turns out to be a 3/8 wave, so that works.

For reference, here is the contents of the MMANA-GAL file:

Inverted L: v40 ft + h22 ft
0.0, 0.0, 0.0, 0.0, 0.0, 12.28, 0.015, -1
0.0, 0.0, 12.28, 6.7, 0.0, 12.28, 0.001, -1
1, 0
w1b, 0.0, 1.0
2, 0
w1b, 0, 0.0, 400.0, 0.0
w1b, 1, 10.0, 0.0
400, 40, 2.0, 2
2, 0.0, 4, 50.0, 120, 60, 0.0
Feeding direct coax 50 Ohm.
Bandwidth (SWR<2) 210 kHz.
Load 2 ia a ground loss.

Tuesday, March 19, 2019

WSPR - 40m, 1 Watt, 16441 km

Distance wise, this is the 2nd best spot I've reported yet at 16441 km / 10216 mi, from ZS1LCD in Cape Town South Africa who was running 1 watt on 40m [7MHz]. ZS1OA remains king at 16511 km / 10259 mi on 30m [10MHz], and ZS3D is a regular in 3rd place at 16233 km / 10087 mi on 40m.

After running WSPR for about three months, I took a break from it about a week ago having built up a good feel for what I can expect with the current antenna configuration. Last night I had it running again after adding a 20 ft horizontal wire to the top of the 43 ft vertical turning it into a 1/4 wave inverted L on 80m.

With the exception of 60m, Remote ATU has no problem finding 1:1 match 80 thru 10m, 40m was the surprise since the antenna is close to a 1/2 wave. Hard to know with out an antenna analyzer to see what is really going on. I'll soon have a better idea, ordered a RigExpert AA-55 analyzer, and another 40 x 32 ft radials for the vertical / inverted L to bring the total to 60.

Monday, March 18, 2019

Extra Class

This past Saturday I made the trip to Renton where the Mike and Key club run VE testing sessions.

It took about 4 months of study and the practice exams on QRZ to reach an average score of 80%. To pass you need a score of 74%, out of 50 questions you can get 13 wrong.

The good news is I passed!

I think the incentive based system here is a good way to manage access to HF bands, in addition to creating an incentive to further your skill and knowledge, regardless of how "irrelevant" some think the subject matter and questions are.

Wednesday, February 20, 2019

43ft Vertical Is Up

Last weekend I took down the Carolina Windom 80, and up went the 43 ft vertical [Zero-Five] in the back yard with 16 x 30 ft radials, and LDG RT-600 tuner at the base. This comes after much internet searching, reading, and many emails with a friend of mine who knows more than I do about verticals and DX. I've yet to finish the radials, need to add more, at least 30 to 40 and bury them.

Left - Close up of base, radial plate - I'll get some paving stones and gravel to make it nice. Right - The lense in my iPhone makes it look "bent", the tree on the right does not lean, nor does the end of the deck or anything else back there that looks like it's falling over. 

Transmitting WSPR, initial results are promising. Near sunset here in the west, my beacons are spotted from the mid west to the east coast - New York, Connecticut and down to Florida. Overnight on 80m they have been spotted in VK5 and VK7, and Costa Rica so far.

Receiving WSPR with the rig band hopping 160, 80, 60, and 40m on receive at night, I spot a few stations from Australia, Japan, New Zealand, and South Africa. A nice surprise was Indonesia and Papua New Guinea on 60m! Note the "official" WSPR 60m frequency is 5.2872 MHz - this is not one of the permitted 60m channels in the US, so we can only receive here and not transmit!

I plan on using WSPR as a measuring stick to compare changes - since I can leave it running on receive 24/7, and enable transmit when I'm home. Monthly archives in CSV format can be download, I can pull this into Excel/Libre Calc to make some pretty charts.

For a low noise receive antenna after much reading about Pennants, Waller Flags among other things. I decided on the the K9AY loop, direction is remotely switchable, and the well made control boxes can be purchased directly from Gary K9AY. Using known quality makes a good starting/reference point for later experimentation with the loops or other low noise receive antennas.

The K9AY loops need about 30 ft of space. Even with a 1/4 acre lot I'm limited on places to put it out of the way. Emailing with Gary who was very helpful, I learned the loop size can be 60 - 70% of original size and be effective on 80m. Gary is also modifying the filter and preamp for me to allow operation up-to 7MHz / 40m as I'm interested in 60 and 40m DX also. Gary also recommended an INRAD model RX7300, this adds a receive-only antenna jack to the Icom IC-7300. Photos here by K0PIR.

I have ordered the control boxes from Gary, and the rest of the hardware:

  • 13 AWG, 19 strand 40% copper-clad steel wire, polyethylene jacket, (part 531) from The Wireman.
  • 150 ft of LMR-240, INRAD RX7300 and ON4UN's Low Band DXing book from HRO.
  • Diawa antenna switch, RCA manual fit plugs, RCA to PL-259 adapters (easy RX switching between the loop and vertical), 150 ft of cheap CAT5e cable (for the K9AY control cable) from Amazon. 

Next I plan to add a 160 and 80 Meter Matching Network for your 43-foot Vertical [AD5X] for more efficiency on 80m. Ideas for later include adding a movable passive director element [VE1ZAC] for directivity and gain on transmit! And ultimately maybe extend mine to a full 1/4 wave on 80m by getting a wire off the end up into the top of some 65+ ft pine trees I have in the NW corner of my back yard!

Sunday, February 3, 2019

Low Band DX - The Idea Bites!

Tuning around the 80m band a couple mornings ago I hear some big signals out of California on 80m - 3.796MHz - chasing DX stations around, that - unsurprisingly - I could not hear at all with my OCFD at around 40ft.

The guys in Cali, one 800 miles south, the other 400 miles south, then started talking between them selves about ZL contacts, and someone I have known for a long time from ZL was mentioned, who has a nice 80m DX station setup. Something I have admired - a 4-square, which is pretty much the gold standard for 80m DX - nice if you have the space! I fired off an email to Nick to say hello, and that I heard these guys talking. Nick replied and said it would be great to chat on 80 some time. Right, challenge accepted!

What I have:
  • 1/4 acre lot with a 70ft Pine tree in the north west corner. 
  • A high noise floor on 80m, S5 to S9 on a horizontal wire antenna.
  • A general class license that limits me to 3.8-4MHz on SSB/AM voice modes, the DX seems to lurk in the upper end of 3.7MHz.

I researched the inverted L, with plenty of long radials these seem to work well on the low bands for DX.

Using a Delta loop came to mind, each side is about 80ft, I have 80ft between the trees at the back and a shorter tree at the front, but lack of height is a problem, I need about 70ft of height at the top of the triangle. Research also suggested that Delta Loops don't work that well for DX compared to other things.

I then remembered an article by W6NBC - High-Efficiency 40m Vertical Without Radials. These are essentially a 1/2 wave vertical dipole, one side is full length, the other leg is shortened, overall the vertical is around 2/3 of full length. In the hope of avoiding running miles of radials I discussed scaling this to 80m with a couple people who have experience with DX verticals, and none of them seemed very keen on it. However, I'm very interested in trying one of these out to compare.

1/4 wave verticals with a good radial system keep coming up as the best option. Some searching turned up a series of articles published in QEX by N6LF - Series of QEX articles on ground system experiments, I found this very informative and cleared up misconceptions I had about radials and ground planes. I have disliked antenna systems that require dozens of radials, its safe to say my thinking has been changed!

Tricky part, how to get something up into the top of that tree? While discussing my insane idea a couple nights ago on a local repeater, one of the guys suggested they know someone with a drone that can carry 25lb loads, and could get something up there no problem! I'm thinking of putting my 43ft vertical from Zero-Five to use. Attaching a ~20ft wire to the top, and pulling that up into the top of the tree to make it up-to a 1/4 wave on 80m.

How it might look from the street. 43 ft vertical will be phase 1. Phase 2 will be to add the 20 ft extension wire and support line from the top of the tree. Not sure of the effect of a "bent" top, but given the tree locations, and where I can fit the vertical in with reasonably symmetrical radials, choices are limited. Approximate but not to perfect scale. 

The proposed radial layout. Radials shown in yellow. Chainlink fence (galvanized) shown in cyan, this extends for some distance east/west down the street. Green is a wooden fence.

The noise floor is the next problem, noise will be much worse on a vertical!

Things to think about:

The last issue, getting full privileges on 80m so I can transmit in the 3.7MHz segment is solved by getting my extra class license.

Wednesday, January 23, 2019

Auckland VHF Group Repeater Site

I recently ran across some photos of the primary Auckland VHF Group repeater site someone had sent me, I don't recall who. However, repeater systems are a subject that interests me, so I tend to absorb details and turn inside-out with excitement when someone offers to take me along for a site visit.

The Auckland VHF Group is located in Auckland New Zealand. Along with the Waikato VHF Group, and the Wellington VHF Group, these clubs have strong membership and skill levels. A subset of members have day jobs in RF engineering and radio communication fields. These three clubs built some great repeater systems in their respective areas.

They would also outreach into other areas and help clubs get well engineered repeater systems up and running. Most of this work was done in the mid '80s, during what was arguably the heyday of ham radio clubs when memberships peaked.

Getting back to the Auckland VHF Group, they built a repeater site named Klondyke after the road it's on. Located at Port Waikato on the west coast, it's about 40 miles south west of Auckland city.

This site served as the Auckland end of the 70cm [430MHz] national link when first built which linked to Wellington [Belmont 439.875-] about 300 miles south via one intermediate site located on Mt Taranaki [Egmont 434.900+]. Over the years it has since expanded to dozens of linked repeaters, exclusively RF linked on 70cm - NZART map here - its approximate at best. Google Map that I created here, the locations are with-in tens of feet in most cases.

With that out of the way, the point of this post is more the 2m repeater at the site. 146.625- a.k.a. "Auckland 6625", my knowledge of the equipment used is somewhat vague, but this is from what I recall of conversations I had around 10 years ago with Colin ZL1ACM who was one of the principal engineers of the site. The repeater hardware is [or was] Tait T300 gear, with two voted receivers, two exciters, two sets of duplexers, and two power amplifiers set to about 40W - This balanced the TX and RX range with typical 25W to 50W 2m mobile rigs, no one likes a repeater that talks further than it hears!.. Nothing too amazing there compared to some of the ham radio repeater systems built in the US..

Now, onto the fun part. The antenna system they built I think is an amazing feat of engineering for a ham radio club!

These are broad side collinear arrays, but rather than have resonant reflectors, they have a grid type panel behind the driven elements. The antennas were constructed from steel, the driven element stand offs [The center of a 1/2 wave dipole can be grounded with no ill effects] and dipoles are angle iron, everything - reflector panel, stand offs, and driven elements - are welded and hot dip galvanized as one piece.

There is a north bay, and a south bay [hence the two sets of receive and transmit chains], each panel has three sets of two half wave dipoles fed in phase, there are four of these panels stacked on each side. This adds up to around 18dBi of gain per side?! The beam width is 70 degrees from memory.

Above is a shot of the tower and building, HF wire antenna can be made out, this is used to transmit the monthly NZART official broadcast on 3.900MHz, as well as out on the 2m repeater and 70cm national link system.

Here is a close up with some details marked out, each of the four bays on each side is fed in at the middle set of dipoles, you can make out the crossed phasing lines.

You can also see someone climbing the tower, they moved the 70cm link yagi to Mt Taranaki [Egmont 434.900+] further up the tower as there were some pine trees near by growing up and obstructing the path

These photos are from 2009.

One of the really interesting things about this 2m repeater and its astronomical antenna gain, it is very very very sensitive to band conditions once you got around 100 plus miles away. If there was the slightest hint of a band lift, this repeater would shoot up the S meter in strength. Once it pegged at S9+30dB, it was time to start looking for band openings. Some times it would consistently swing in signal level from seconds to minutes, other times it would bottom out and disappear all together..

This repeater could be reliably worked up-to 150 miles in each direction, with the potential for 300 mile separation between stations at the opposite ends of its coverage. Much of this is over less than ideal paths, not bad considering the repeater site is only 1309ft ASL. When band lifts are present the range would easily extend beyond this.

Now some say this antenna design was borrowed from the Mt Kaukau VHF low band [45MHz and 55MHz] TV broadcast antennas, click the image below for a high res version, and look at the top, horizontally polarized for TV, but looks familiar doesn't it? The Belmont 147.100+ repeater built by the Wellington VHF Group also used antennas based on this design.

Saturday, January 19, 2019

19" Radials on the Diamond X50

After reading Runt Radials I ordered some 19" radials from Comet, these also fit the Diamond antennas.

My Diamond X50 came with the 7" radials, and I've had problems with the coax getting "hot" [common mode currents] with RF when using it on 2m.

Here's what the 19" radials look like

Diamond X50 mounted on 2nd level deck. I'm located in the south Puget Sound at 400ft ASL, and have a clear view north towards Seattle (35 miles north), this is a secondary antenna connected to my Kenwood TH-D72 hand held. With 500 mW I can open the 145.190MHz N7GDE repeater located 100 miles north of my location.

While I have not done any form of testing that's scientific, the anecdotal results are slightly better performance on both 2m and 70cm, with the added bonus of the feedline being better decoupled from the antenna on 2m.

Sunday, January 13, 2019

CHIRP, Kenwood TH-D72 and WWARA DB Extract

How to use the WWARA coordinated repeater DB extract, Excel or LibreOffice Calc, and CHIRP to program a Kenwood TH-D72 - or any other CHRIP supported rig. Noting this to remind my self how to do it in 17 easy steps!

  • CHIRP - Free programming software
  • WWARA > CSV Database Extract [includes CHIRP formatted file] 
  • LibreOffice - Calc / Spreadsheet - used to edit the CSV file [CHIRP sucks at this]
The Process:
  1. Create backup of the rig to be programmed, in my case I use Kenwood's MCP-4A application.
  2. Download and unzip the Database Extract from WWARA.
  3. Open the CHIRP formatted file in CHRIP, and export it as a CSV file.
  4. Open the CHIRP formatted CSV file in Calc or Excel.
  5. Delete the rows you don't want. In my case I remove the 6m, 220MHz, everything above 70cm, and anything that is not an analogue FM repeater - for that I use the custom column sorting to group repeaters marked as digital together for deletion, and column sort Location numbers again when done to get whats left back in order.
  6. Optional, the Location column numbering can be fixed using the fill down function to sequentially number the memories we are keeping.
  7. I add in my custom APRS, IRLP nodes, and other things starting around memory location 980. The Location column can skip unused memory locations out, e.g. mine jumps from 266 to 980.
  8. Save it as a CSV file again and close Calc / Excel.
  9. Open CHIRP.
  10. Connect the radio to be programmed.
  11. Perform a read/download, and save it as backup.
  12. Open the edited CSV file.
  13. Eyeball it to make sure nothing is out of place, if so, close CHIRP and fix the CSV file.
  14. Clear/delete everything from the tab containing the information from the read/download, copy and paste from the CSV file tab. The exact steps to do this have changed at least once between CHIRP versions.
  15. Write the config back to the radio.
  16. Done!
  17. Not so fast, I then read the config in the Kenwood MCP-4A application again, and save it as a backup.

Saturday, January 12, 2019

Comet GP-15 and 6 Meters

The Comet GP-15 is a single piece tri-band antenna that covers 6m, 2m and 70cm.

The 6m band is "tuned in" via loaded [1/4 wave on 6m] radial that is adjustable with about 1MHz of bandwidth, the other two supplied radials are 1/4 wave on 2m, and double as 3/4 wave radials on 70cm, which is fine for these two bands. The GP-15 is phased as a single 5/8 on 6m, but the single radial lets it down..

Issues with this configuration, the single 6m radial results in narrow bandwidth, and more importantly, may not properly decouple the antenna from the feed-line! Take a look at:
  • Runt radials - This talks about the Diamond X50 and Comet GP-3 with 7" radials [1/4 wave on 70cm] which do nothing on 2m [1/3 of the frequency]. Same theory, since the two 19" radials [1/4 wave on 2m] on the GP-15 do nothing on 6m [1/3 of the frequency].
  • Modifications to the Diamond V2000 and similar antennas - This reveals the Diamond V2000 is less than ideal on 6m. Diamond call it a 1/2 wave, which is technically correct, but its more like 1/4 wave with one radial, that performs less than ideally.
  • EA4EOZ also did a V2000 teardown, interesting to see what's inside these things.
  • Experimenting With 6 Meter Ground Plane Antennas - If these guys have a problem with the single radial on 6m, then it's worth paying some attention to.
Since learning about the compromise radials on these antennas, I decided to try out a set of three full sized 1/4 radials for 6m, the comparison:

First with the supplied tuning radial centered around the FM repeater inputs used in Western Washington:

SWR plot on an Icom IC-7300, 50.0 - 54.0 MHz [500KHz steps]. With the supplied tuning radial, the bandwidth is 2MHz at most.

Quite narrow banded! SWR starts to get high at the bottom end of the band with the North American SSB call freq at 50.125. SWR starts to get high again at the upper end of the FM repeater inputs.

Next up, three full sized 1/4 wave radials, I trimmed these until the SWR just started to sneak up from 1:1 at 50.0 MHz:

SWR plot on an Icom IC-7300, 50.0 - 54.0 MHz [500KHz steps]. With three full sized 1/4 wave radials, bandwidth is 3MHz or more.

Very nice, the bandwidth is nearly doubled! The antenna is now usable down to 50.0, and up into the FM repeater inputs in my area. I'm thinking I could cover the top 2MHz of the band with a second set of radials cut shorter and placed in-between the longer ones.

The 6m 1/4 wave radials I made out of some wire, with lugs crimped, soldered, applied "Lanocote", some heat shrink to seal it in, and offer strain relief where the wire is crimped. I attached them between the locking nuts and antenna base where the radials screw in, applied more "Lanocote" for good measure. The antenna is on a short TV J-Mount on an roof apex, two of the wire radials are laid on the roof which slopes away - note my roof is composite shingle over plywood and not tin/metal as is more common in other countries - the other has a short section of thin nylon rope to hold it up at about the same angle.

I replaced the 6m "tuning" radial with a regular radial like the other two supplied radials - I ordered this directly from Comet, along with three more 19" radials for my X50, total cost with postage $36, and yes they use the same thread and locking nut sizes!

The EA4OEZ V2000 mod replaces the factory radials with an M6 bolt and aluminum tubing.

As for the 19" radials, it's not clear where to get these radials from, no one lists them as something you can buy, not knowing where to start, I emailed HRO, Diamond, and Comet asking if they sold these. Comet quickly replied.