John Langridge KB5NJD / WG2XIQ (Originally written in August 2012)
Up until this point, I have detailed the construction of a loading coil with variometer, effective CW transmitters for 630 meters, and modifications to an MFJ antenna analyzer in order to make reasonably accurate measurements at these frequencies. Truthfully, that was the easy part. Implementing an effective antenna system is a completely different story. I have spent considerable time thinking about a starting point for this article because every installation will be different depending on the radiating element and the radial system. Since this is the case, I will explain the implementation of the loading and matching system from the perspective of my station, which is a worse case scenario in terms of complexity. For the average ham, implementation will be considerably simpler than what I describe but hopefully these descriptions will answer a lot of questions on the subject.
A word of warning! The antenna-tuning unit I am describing with a legal EIPR of 1W can still generate in excess of 12000 volts on the radiating element under certain conditions. At currents in excess of 1.5 amps, this can kill you instantly. Please use your head, don’t work alone, and ask questions instead of taking chances and making guesses.
To begin, I should describe the antenna I am interfacing with. The radiating element is a 66-foot inverter L that has been used for a number of years on 80 and 160m. This length of the antenna exhibits a natural quarter wave resonance on 80 meters and vacuum relays are used to switch in a loading/matching network for 160 meters. The radial system consists of 130 – 100-foot radials and coupled with a good receiver and excellent receive antenna system, this transmit antenna has been good for close to 300 countries on 80 meter and well over 100 countries on 160 meters. A number of 8-foot grounds are implemented at the feed point for the purpose of lightning protection. The main goal of this implementation was to keep the existing low band station completely separate from the 630-meter station. Since a separate receiver and transmitter were to be used for the 630-meter station, this would not present a problem if RF routing with quality high voltage RF contactors was utilized AND it was understood that only one station would be on the air at any given time. These were reasonable assumptions.
The first step in keeping the system completely separate was to introduce logic and switching at the base of the antenna to control where RF would be routed from the coax. Up until this point, all RF was routed to a network at the base of the antenna, which was controlled within the shack to allow the selection of 80 or 160 meters via high voltage vacuum relay. That would not change. The addition of a low voltage relay, rated at 1000 volts RF peak, would allow signal from the coax to be routed to either the 80/160 network on the new loading/matching network for 630 meters.
Low voltage relay mounted on the fiberglass pole routes RF from the 9913 coax between the 80/160 network on the left and the ATU for 630 meters that is out of frame (copper wire exiting frame at lower right).
A switch in the shack controls RF routing and the relay is affected by the selection of that switch. I get questions from time to time about leaving relays like this out in the open. The only holes in this relay are drain holes on the bottom of the relay. There are no problems leaving them out in the open so long as those drain holes are located in the down position. I’ve had relay like this covered in snow and remain resilient so don’t worry too much about them. High voltage vacuum relays are managed in a similar manner as you will see and they are even more resilient than a typical open frame relay.
But there is a catch – no, it’s not that simple. As I have related in at least two previous club presentations, I use separate transmit and receive antennas on the low bands ALWAYS in order to improve signal to noise ratio through directivity ( and consequently F/B). In order to achieve that, I have to detune this transmit antenna when I am receiving to prevent the coupling of noise into the high performance receive system and having a negative impact on directivity. To not detune this antenna would be like have a tribander on a tower that was devoid of any and all directivity and front to back. This meant more relays in the shack so that when I was on 630 meters and listening, the transmit antenna could be setup so that it was actually tuned for either 80 or 160 meters, well outside of the 630 meter band. When the foot switch is pressed, the relays almost instantaneously change state and RF would be routed to the 630-meter ATU. To do this right, I needed a second relay to switch high voltage from the active network to the radiating element at the same time as RF was being routed to a given network at the feed point via the low voltage relay. It’s a simple solution but an expensive one, as Gigavac, who makes the best high voltage relays in my opinion, sells the relay I need for around $200 shipped. It’s a lot for a single component that could easily be destroyed by a lightning strike, but this is the simplest solution and thus it is implemented as such. More about this in a bit.
At this point, the system is working like normal and I have access to 80/160 meters with the addition of a position for the 630-meter loading and tuning unit. Everything is cool.
The 630 meter ATU was built into a blue Rubbermaid tote capable of holding almost 50 gallons and available from Walmart for about $20. The container was ultimately painted brown to blend in with the surroundings – why advertise?
The ATU consists of four parts: 1) a tapped loading coil/variometer to manage the capacitive reactance presented, 2) a motor to control the orientation of the variometer for remote tuning, 3) a method of antenna system current measurement for power calculation and determination of system quality and 4) a means of matching the resistive component exhibited by the coil to 50 ohm feed line.
The first logical item to implement in the container was the large, tapped loading coil, complete with variometer described some months back. This simply involved centering the coil at one end of the box, punching a hole in the box for the variometer’s rotary shaft to exit the box for the control motor located outside of the box (an Archer antenna rotator mounted on its side), and mounting the coil on 6 inch ceramic insulators so that it was not sitting on the floor of the cabinet and presenting a high voltage hazard with respect to ground. Three ceramic insulators were located on the base of the coil to provide stability followed by two additional insulators attaching the coil to the sides of the box.
The center-right insulator is RF input to the coil and the left most insulator on the front of the box is high voltage out, intended to be routed back to the antenna high voltage relay. In this picture, the high voltage lead is not yet connected. Ground is fed to the ATU via the copper/tin flashing material and connects to the radial system and ground system at the antenna base.
It was determined that the motor would best be located outside of the box. Doing so would allow a longer shaft to be used, keeping the motor further away from high voltage. An archer rotator was mounted on a 2 X 4 and attached to two vertical pole set into earth with concrete. As this was a fixed location, the final location of the box was set by the location of the motor mount. The decision to utilize the rotator rather than an custom motor drive was simple: 1), the rotator was sitting in my garage, 2) the rotator offered limit control and predictable motion over the 90 degrees that I needed to cover, and 3) the mount was simple to devise using available hardware. I would have put forth much more effort with a custom motor with limit switches and a control box. I believe that I can better implement the control box in a form similar to a Ham IV antenna rotator and that is a long term goal to allow fine tuning of the position of the variometer, which impacts the match on any given frequency. For now I get to settle for the archer box and have become quite proficient with its control when matching the system.
The next component of the ATU assembly is the current meter. System current measurements are some of the most important measurements you can make. This information can reflect system behavior or environmental changes that might be occurring and can be used with what is known as the “direct method” of output power determination if the radiation resistance of the antenna is known (“I squared R method”). More about this calculation later or in follow up articles. The takeaway message here is that you need to be able to monitor the current through the antenna system and this is accomplished using a current transformer to sample current in a low loss manner from the main high voltage lead going to the antenna. The decision was made to give myself the option to monitor the system current in real time from the operating position in the shack as well as have the ability to switch to a local meter at the ATU for troubleshooting. This flexibility would be selected using a knife switch within the ATU box. Shortly after making the calculations for a 1:30 turn ratio current transformer that I would wind from scratch, my old friend and former coworker Bill Guyger, AD5OL, presented a solution I could not resist. Bill had removed a calibrated Bauer RF current transformer from a commercial broadcast ATU that was part of a multi-tower AM array and offered it to me for nothing. I told him I would buy him lunch and thanked him profusely as this current transformer was going to give me a lot better certainty compared to anything I could roll on my own. I quickly built a mount for the transformer and routed RF from the knife switch on the output of the big coil to the high voltage feed through output lead on the inside of the ATU box and I was set. Sampled signal is routed back to the shack on coax via a BNC feed through located on the ATU box. Its important to keep grounds isolated on the sampling line at least until the signal is back in the meter box in the shack. But that’s a whole other article.
The final component of the 630-meter ATU is the matching of the resistive component in the impedance. Measurements had indicated that my natural feed point resistance was around 28 – 30 ohms and about 2 ohms of reactance in a best-case scenario. Most of that resistance is loss but I still had to achieve 50 ohms.. I can live with this reactance value and my experience with the MFJ analyzer is that when it comes to reactance, whatever it reads, its actually closer to half that. The 28-30 ohms of resistance can be better since I using 50-ohm coax feeder. Even with the low loss associated with the feed line at these frequencies, I want to avoid any problems related to mismatch that could lead to excess high voltage on the coax, or worse, causing the MOSFET in the PA desk of transmitter to blow (remember, they sound like a shotgun going off when they blow up!).
My original hope had been that I could simply use the additional unused turns on the loading coil for a ground connection to achieve a 50-ohm match. This is pretty common and how I match virtually all of the vertical antenna systems I design or implement. Not only does this offer an easy means of matching the system to the coax but also it provides constant static discharge, which can be a problem with large antennas. Unfortunately, this was not going to be so simple. Systematically connecting all of the lower taps on the coil to ground did not achieve a better match, in fact, it often made matters worse. It was pretty obvious what was going on here: I was going to have to use a capacitor shunted to ground. Fortunately for me I had bought several high voltage mica transmitting caps at hamcom from a guy that seemed to have an unlimited supply. I sorta felt like he was sent there for me because using one of his caps, a .003 MFD between the input connection at the coil and ground brought me to 50 ohms without the addition of any reactance. Ok, that’s sorta what happened. I did find that the best match was achieved by connecting the cap to the feeder tap by way of a smaller coil that I had wound for another project. I could have just as easily re-tapped the large loading coil in between taps and connected the capacitor to ground achieving the same solution. In reality, the resistive match is achieved with a series LC network shunted to ground and was not as simple as a capacitor to ground at the coil. Your mileage may vary. Remember that every system is likely going to be different in some way, which presents my dilemma in explaining this.
RF enters to cabinet from the bottom center of the picture and routes to the proper tap on the coil. Resistive matching is accomplished by taking RF from the same tap to the mica transmitting cap in the lower left of the picture and routing to the coil to the left (see below pic). Current measurement is accomplished with the Bauer current transformer where high voltage from the variometer (via the knife switch at the top) is routed as a single turn primary to the output feed through on the left side of the box. Signal is routed to the coax and back to the shack via the black and white/red and black pair in the upper left. Ground is presented around the top of the box. This gives a convenient connection point and a nice location for arcs to go rather than through me if I an working around the box while it is active. Note: Mothballs keep the critters out!
Series LC circuit to match the roughly 30 ohms of feed point impedance (mostly loss resistance!) to the 50-ohm coax.
Once the RF leaves the ATU, I wish it were as easy as to say it was on its own. Unfortunately, that is not the case. Recall from a few pages back that I had to be able to detune the antenna when receiving to avoid impacting the receive antenna system. Also recall that I needed an expensive high voltage relay to select which network was going to be providing RF to the radiating element.
Meet the Gigavac G2-Ham vacuum relay:
High voltage relay installation mounted on plexiglass on top of the 80/160m network.
This relay will switch 43 kilowatts at around 15000 volts so it’s the perfect solution to my problem. From the picture, the green wire connects the 80/160-meter network to one pole of the relay. The large black cable on the left comes from the 630-meter ATU to the other pole of the relay. While this cable looks to be sitting on the ground in the picture, it has since been supported with a PVC standoff about 2 feet off of the ground. The brown wire headed up through the holes in the plexiglass is the antenna radiator and it connects to the common on the relay. The gray power cable leads away from the based of the relay to a PVC standoff and the on to ground where it is connected in parallel with the coax relay connected to the fiberglass pole at the base of the antenna. Using this arrangement has presented no problems at all and RF is decoupled from the high voltage relay control line with a .1 uF ceramic capacitor.
So those are the components. I’ve built a complete 630-meter CW station from scratch except for the receiver, which was heavily modified. Next month I will discuss the mathematics behind determining how about 50 watts coupled into this system will result in only 1 watts radiated in addition to the next phase of the project: A transverter with a 28.472 MHz IF and 28 MHz local oscillator to allow for the transmission of FSK-based digital modes (like OPERA and WSPR!) without any fears when using my class-D/E MOSFET amplifier. Many Europeans are already enjoying the new band and have packed the band with a multitude of both CW and soundcard generated digital modes. I hope to bridge any gap that exists with station capabilities through the additional of computer-generated modes.
Limited system testing has yielded good results and nothing has caught on fire yet. I anticipate at this point that the effort I put forth in scaling the system for higher power has paid off in a system that can withstand the perils of high power, low frequency operation.
Finally, here is a poorly drawn block diagram of the antenna system. Since we are talking about 630 meters, I have only detailed that particular part of the system and only added the 80/160-meter network for clarity. Note that both relays are switched in parallel, thus isolating each network when receiving. When the transmit switch is actuated, the relays switch before RF is applied.
Feel free to ask questions…
73 and see you in the pile up!