Tag: Project

I enjoy experimenting with antennas. There is something deeply satisfying about building an antenna and then getting to communicate with others around the world on the thing you built. I also enjoy playing around with commercial antennas, but it seems that the cost for those is always significantly higher than the home brew variety and you don’t get the same satisfaction of using your own.

While at Dayton Hamvention in 2023 with my good friend Gersohn KO4UIK we saw a whole host of antennas on offer. Most were quite expensive. I also noticed that many vendors were using 3D printed parts to make some of the components of their design. Being an avid user of 3D printing technology both as a consumer and as a creator of my own models, I was inspired. What if I could make my own design and make it for a fraction of the cost of commercially available options?

Now the question was, what would I build?  I already had a few vertical antennas using whips that I was quite fond of, so I was initially drawn to that.  However, I had always been intrigued with dipole antennas.  I had used the electrically similar end fed half waves extensively at home and in the field due to their ease of deployment and their ability to work on multiple harmonic bands.  Their downside, if it can be called that, is their impedance matching network, which allows for the multi-band trick at the expense of a small amount of efficiency.  Center-fed dipoles had always been interesting due to their efficiency but their lack of versatility always kept me away.  So I thought, what if I were to make a dipole using two collapsible whips?  I could modify the length of them to make them resonant on any band where I had sufficient length from the whips, eliminate the need for a radial network, and get all that efficiency lost to impedance matching back in the process. And while I’m at it … why not both?

Enter the N2EC VersaHex.  Here was the idea: have a 3D printed hub in the center of the antenna assembly to provide the structure and electrical connections.  For a vertical antenna a single whip would connect at the top and the ground connections would connect at the bottom where a radial system, ground spike, or tripod connection would allow it to be mounted.  For the dipole configuration the two sides of the hex adjacent to the top would each have a whip mounted to them.  This would create a v-shaped dipole and gravity would cause it to sort of approximate a conventional dipole a bit.  As for the components, I decided to utilize the vibrant ecosystem of parts from Chinese marketplaces like AliExpress and Banggood where a lot of the metal bits I would be needing were significantly less expensive.  This meant standardizing on the M10 thread size for the whips (which could be had for as little as $12 US) and getting a lot of associated hardware.  It also meant I could get inexpensive loading coils, ground stakes, and radial systems that would easily connect with my design. With the basic framework of the project in mind, it was off to CAD.

The design came somewhat quickly and I had ordered a bunch of the bolts, washers, whips, ring connectors, and wires to connect it all together. In my first prototype my wall thickness was too large for the BNC connectors to be secured, so I had to come up with a way to make that part thinner. I decided to do that with a sufficient radius to allow the part and then adding in a chamfer to get it to connect smoothly with the main walls. This worked well. I also created a snap fit lid for easy assembly and maintenance, but the early prototypes had it coming off when the whips flexed the plastic. I deepened the flanges and that seemed to help greatly. I printed everything in PETG filament, which would provide greater strength, flexibility, and heat resistance while out in the field. I also found that printing this design solid with 100% infill was very helpful as the cantilevered whips do place a lot of strain on the 3D printed hub.

After a few prototypes, I had something I really liked. Since everything just connected up, this design was incredibly versatile. If you had limited space, deploy it as a vertical. Have more room? Do the dipole. And since everything is modular, you can connect lots of accessories. I found a lighting stand on Amazon for about $50 that allowed me to get the dipole up about 10 feet while still being stable. This would end up being one of my favorite ways to deploy the antenna. At home I set it up and did some QRP tests on the bands and got promising results, starting first on the dipole mode. It really seemed to get out. So now it was time for a real test.

My local club, the Mount Vernon Amateur Radio Club, has an annual special event we operate to celebrate George Washington’s Birthday.  We operate from his estate, Mount Vernon, and celebrate over the course of the weekend with the world.  I usually am one of the ops for the event, so I decided this would be a great opportunity to test my new N2EC VersaHex in the field.  Right away it was clear I was on to something special.  I got some great pile-ups on the bands and lots of unsolicited reports from the folks who worked us.  They asked if we were running an amplifier (we weren’t – just 100 Watts barefoot) and then asked “What antenna are you using?”.  When I replied it was a homebrew system, they asked if I was going to share the design.  That was a lot of fun.

There were some other benefits of the system, too. Since we were doing a multi-op, we usually interfered with each other … but since this was a horizontally polarized antenna, and the other station was using a vertical, we didn’t bother each other at all. No band pass filters required!

Since then I’ve used the antenna system for portable operations and at a Cub Scout event where we got a bunch of the kids on the air to show them what amateur radio is all about. I’m pleased to say that even with challenging band conditions, every kid who wanted to have a turn at the mic had a successful QSO that day, and many of them were on the N2EC VersaHex.

Electrically, the N2EC VersaHex it is quite simple. The dipole feed-line enters at 4 on the diagram at the left BNC connector. The center conductor of BNC 4 connects to connection 1. The 16 AWG wire is soldered to the center pin of the connector and then is connected via a ring connector to connection 1 using a M10 bolt, 6 M10 washers (3 on each side of the plastic case) and a M10 coupling nut. The ground of BNC 4 is connected in a similar fashion to connection 2 but instead of soldering to the BNC, a ring terminal is used on both sides of the wire.

The vertical feed-line enters at 5 with the center pin of the BNC connector going to connection 3 (with the same bolt, washer, ring terminal, and nut arrangement of the dipole).  The ground connection of BNC 5 is connected to connection 6.  Things are pretty tight in here, so I find that pre-folding the ring connectors as seen in the diagram can help with installation.  Also, DO NOT SKIP THE WASHERS!!!  These distribute the forces from the whips on the plastic.  In testing users who didn’t use the washers had the plastic hub fail catastrophically.  Use the washers!  This may also be a good time for me to say that this is a hobby project, so if you decide to build one of these it is at your own risk.  I’m sharing this so other folks can have some fun, but I can’t provide support to everybody on the internet, and if something breaks, I can’t help you fix it.  As they say, no warranties are expressed or implied.  The model files are available via this link. The Bill of Materials I used (along with some optional bits) can be found here.

Alright … so how does this thing perform?  Pretty well!  I’ve messed around with some modifications using those aforementioned accessories and have had some good success with it.  The one thing that didn’t work well for me was using two loading coils in dipole mode to get 40 meters and 30 meters.  I could not get decent SWR on those bands with the loading coils.  I even tried with some aluminum extenders to not load at the center.  That helped, but not enough.  The  loading coils worked well for vertical mode, so if you want to use it for the lower bands, do it in vertical mode.  Also 10 meters in vertical mode was below 2:1, but only just … not sure why, but there again I’d use dipole mode.  It is good to have some versatility! Here are some sweeps from my RigExpert Stick Pro in the various modes.

20 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

17 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

15 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

12 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

10 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

6 Meter Dipole Mode (higher line is elevated 10 feet on lighting stand)

40 Meter Vertical Mode (Ground Mounted)

30 Meter Vertical Mode (Ground Mounted)

20 Meter Vertical Mode (Ground Mounted)

17 Meter Vertical Mode (Ground Mounted)

15 Meter Vertical Mode (Ground Mounted)

12 Meter Vertical Mode (Ground Mounted)

10 Meter Vertical Mode (Ground Mounted)

6 Meter Vertical Mode (Ground Mounted)

This has been a very satisfying project. I really enjoy being able to use my own antenna out in the field, and the fact that it can be realized for well under $100 on a per-unit basis while outperforming a lot of my more expensive antennas just adds to that satisfaction. What does the future hold? Since this is such a modular system, I’ve already started to build components to turn this into a multi-element Yagi system utilizing some aluminum extrusion bars and commonly available hardware. I’ve sourced all the parts for that project, but have not gotten around to actually testing it. When I do, I’ll keep you updated. If you do end up building one of these drop me a line and let me know how it went.

Until next time, best 73.

When I am operating CW from my home station I am usually using a K1EL K45 CW Modem as the keyer.  It has aptly been described as a “Swiss Army Knife of CW Keyers” as it is an incredibly capable device.  I have built and used some other keyers in the past, including the Hamgadgets Ultra Pico Keyer that works very well and I was quite pleased with.  The reason I decided to go for the K45 was that as someone who teaches multiple CW classes with the Long Island CW Club, I really need a visual indication of what speed I’m sending at as well as an easy way to be able to modify that speed in single word per minute increments.  The Pico Keyer will send its speed in CW after adjusting a potentiometer, but making those updates while teaching can be challenging.  I had been using my Yaesu FT-897 as a sidetone source for the classes (and still do), but that meant that I had to have a separate setup for teaching and operating other rigs.  The K45 has a LCD readout that allows you to see your speed (along with a whole lot of information) so it seemed like a good choice.

Of course, the K45 allows me to do iambic keying as you would expect. It also can act as a Winkeyer to allow for keying from the logging software of your choice for use in contests which was a plus as I was starting to get more interest in working contests.  Additionally, it provides a pleasant sidetone, which is helpful as my Hermes Lite 2, which I have been using a lot lately, does not generate sidetone when keying CW.  The K45 does a whole lot more than that though — and I don’t really even use the half of it.  It can key from a keyboard and has several keying memories for automation.  The keyboard also allows you to go deep into the menu system without needing a computer.  It can also do RTTY, HSCW, and QRSS.  It also has the ability to decode CW by sending your receive audio into the unit (I’ve never used that function, but it is there).  It also has a USB interface for use with the Winkeyer, the remote display application (which shows the LCD information on your computer), and to allow for firmware updates.  It is also very well designed and has ESD input protection and RFI suppression on all the connectors … which is a great thing, but is something which created an interesting problem for me.

I have a lot of different keys and paddles, and I love playing with different ones to keep my operations fun and interesting.  I have things setup so I can have multiple paddles feeding into the K45 via a splitter, then I have the output going into another splitter so that I can patch-in manual keys and bugs that feeds into a stereo jack switch that allows me to switch between multiple rigs while keeping them electrically isolated.  The end result is that I can easily switch between my favorite keys and paddles at a moment’s notice and route them to whatever rig I feel like using.  This all worked very well until I tried to connect my 9A5N Solid State Paddle to the K45. 

The 9A5N Paddle is a unique paddle that has no moving parts.  Instead of a standard contact closure when you touch the paddle it uses a series of strain gauge sensors to drive a microcontroller to provide the keying.  This provides some advantages.  You can set the force required to activate the key from 10 to 50 grams of force to your preference.  There are no contacts to get dirty and stick.  And the paddle is automatically “tight” as there is no gap at all to worry about.  It is a very nice bit of engineering.  But it appears to have a problem.

When I connected the 9A5N to my K45 it would cause the K45 to reboot itself and behave strangely.  It was very frustrating, as I wanted to be able to have it connected, but it just would not work.  I reached out to Steve K1EL over at K1EL Systems (the manufacturer of the K45) about what might be happening and right away he told me he was familiar with the issue.  It appears that the 9A5N paddle has some large transients that are generated as the microcontroller closes the switching transistors that allow it to work.  Those transients are large enough to overwhelm the input protection on the K45, which causes it to protect itself and exhibit the behavior I was seeing.  Steve was quite responsive and was willing to fully refund my purchase for the K45, or even remove the ESD protection for me if that was desired.  I had no desire to return it, and I’m a fan of ESD protection as I hope to use the K45 for many years, so I decided to just let it be and the 9A5N paddle went back on the shelf.

A little while later Steve reached back out to me and mentioned that another customer had come up with a solution to fix the issue by taking 33 ohm resistors in line with the tip and ring connections between the 9A5N and the K45.  The resistors were limiting those current spikes, and appeared to allow the two devices to work together as desired.  This seemed like a great idea, so I thanked him for following up and added it to my queue of projects.  Life got in the way as it always does and it sat on the list for a few months until just recently when I decided to build the device, 3D print an enclosure, and give it a try.  The result was what I’m calling the N2EC Paddle Tamer.

As mentioned, the circuit could not be simpler.  The box has two stereo 1/8 inch (3.5 mm) jacks on either side connected to a bit of perfboard with the tip and ring connections connected through separate 33 ohm resistors that I had in my supply.  The ground lead is just directly connected through between the sleeves of each connector.

I had some perfboard from Elegoo lying around that I had picked up for another project from Amazon while back, the resistors were also on hand, as were a couple of 1/8 inch (3.5 mm) stereo jack cables I had supplied for another project, so I didn’t need any parts.  All I needed was a box to put it all in, so I went into Fusion 360 and designed an enclosure to house the board and provide holes for the cables to enter and exit.  A quick print in PETG with my Bambu Labs X1 Carbon later, I had the box and was ready for assembly. After soldering the connections as per the schematic and tying a loop in the cables  to provide strain relief for the cable and connections, I closed up the snap-fit case I designed for it, complete with my call sign and the name of the device and gave it a test.

Happily, it worked perfectly.  Now the 9A5N Solid State Paddle was able to send crisply and accurately via the K45 and was able to coexist peacefully with the the rest of my keys and paddles.  It was a very simple project, but one that really has had an outsized impact on my CW operation.  The 9A5N is particularly good for QRQ (high-speed) operation, so I’m glad to have it available to me for all my rigs in the shack for those times when I feel like pushing my speed.  It is always fun when a project works well from the start and a few spare parts can make something work again.  Lots of fun.