CFA Tuner
I have been a radio amateur about 35 years and I have always suffered from bad locations, where it was difficult or impossible to use decent antennas. That is why I have been interested on building equipment and small, simple and preferably invisible antennas. About a year ago I found the antenneX website. I printed all of the CFA articles and read them through several times. Already at that time it came to my mind that the CFA could be done with 3 wires.
Metric conversions:
Millimeter to inches = 0.03937
Centimeter to inches = 0.3937)
Meter to inches = 39.3701
Meter to feet = 3.28084
I was figuring a vertical antenna with horizontal polarity for 80 meters, but I started with the original design of 45 cm height and 1 x 1 meter ground plate. I made the CFA with 1.5 mm thick aluminum sheet with 6 cm center hole for wires, cutting the ground plane 1 x 1 meter and then cutting the corners off to make it octagonal for better portability and handling. D-plate was 40 cm in diameter with 6 cm center hole. The E-cylinder was made by turning the aluminum sheet to a 20 cm diameter tube. I riveted the ends of the plate together to make it hold its shape. The length was 25 cm. The distance of the D-plate from ground plate was 10 cm and respectively 10 cm from D-plate to E-cylinder. The spacers were made of 10mm PVC bar with slots to hold the E-cylinder and holes for screws to fix to ground and D-plate.
After putting the CFA together I started to make the tuner part of the antenna. I had an old AEA AT300 tuner which I decided to use for this project. I had two inductances, a capacitor, a balun transformer , an SWR meter and a switch for other antennas. I had to add one capacitor and rearrange the connection of the components inside the tuner box. First I used a separate coil for the parallel circuit to drive the E-cylinder. It was difficult to adjust because the LC-circuit was inside the cylinder. I changed it under the ground plate and now I had two tuning capacitors in the tuner box. I found the same warming toroids and high voltage arching at the capacitors as experienced by others who tested this circuit before me. (Figure 1)
Then, as I had two separate switchable coils, I made the tuner with two adjustable LC–circuits and changed the parallel circuit to a similar circuit as the D-plate feed. I had both capacitors connected to ground and it was possible to tune to all HF bands from 1.8 MHz to 30 MHz. The power rating of my circuit was low. I could drive it with about 40 watts . (Figure 2)
It was at the beginning of July 1999 when I had several contacts with Finnish amateurs with my CFA and I sent a message to antenneX Publisher, Jack L. Stone, of my successful testing of a CFA antenna. He asked if I would like to join the GARDS group to develop small antennas. I agreed and continued my tests with the CFA.
I changed the dimensions of the CFA and found that by increasing the height of the E-cylinder I was able to increase the power used , but then I noticed that my feeder cable started to radiate. I had to increase the ground plate dimensions also. I left the dimensions as they were and continued to evaluate the tuner circuit.
Feeding Tuners
In early August of the following month, one of the GARDS team members proposed an idea of feeding the power trough an LC circuit only to the D-plate and having a separate parallel circuit connected between ground and E-cylinder. (Figure 3) It worked better than the original circuit, but there were two adjustable circuits and the tuning of one capacitor affected the other as before. I wanted the tuner to be as simple as possible.
During September I made some QSOs to Germany and Holland on 20 meters with my CFA. At that time, I then became distracted with another new small antenna, the DLA. It was November before I would return to the tests with the CFA. I made a tuner with serial connected coils. (Figure 4) I had one LC circuit to D-plate and from there one coil to E-cylinder. The circuit resonated on 80 meters. I was excited by this progress and increased the power to the antenna just to find where the SWR would go up. I couldn’t use the AT300 tuner coils any more because they were short-circuited.
Now I continued to operate on 80 and 40 meters with my DLA. It wasn’t until January 2000 when I continued the CFA tuner tests. I had read the antenneX article by Joel C. Hungerford, KB1EGI and continued by rebuilding my tuner with T200-2 toroids. I thought I was using the circuit in Figure 4 and spent a whole day when trying to get the circuit operate. At last when using a T300-2 toroid (for source of Toroids, see note below) it made the circuit operate. I was pleased for that and let the project sleep about one week. Then after a week I rechecked my circuit, I noticed that I was using a parallel feed instead of a series feed in Figure 5. Now I was feeding D-plate through an LC circuit and E-cylinder trough an inductor and these inductors were parallel coupled to the input feeder cable. The circuit operated, but I was not happy with it.
I saw an advertisement on a British amateur journal. It was Maurice Hately, GM3HAT who advertised his various antennas including a 3-wire antenna. Recalling the beginning of my CFA testing I now decided to continue with 3-wire antennas. My tuner operated with the 5-meter 3-wire antenna I had built. I happened to posses a roll of 3-wire flat cable used in TV industry. Its width was about 20 mm and it seemed to be just what I needed for the project. In a 3-wire antenna we have the same parts as in the CFA, a ground wire, a D-wire and an E-wire.
Balun Feed
At the end of January 2000, Joel Hungerford mentioned using a balun feed. I made a balun coil with 5.5 uH inductances and used 2 pieces of 250 pF capacitors to tune the antenna as in Figure 6.
The balun feed and a Quad Phaser network resemble each other quite a lot but the phasing of the coils are not the same. I checked the operation of my tuner by reading the antenneX articles of Harold Allen, W4MMC and his Quad Phaser designs. In the tuner I have a transformer with a resonated secondary winding, so it has to be 90 degrees phase difference between primary and secondary sides. The tuner worked and now I had a tuner with a bifilar coil, but still with two capacitors. I replaced the other capacitor with a 400pF capacitor from the old AT300 and noticed that I did not need the other capacitor. Now I had a tuner with only one capacitor as in Figure 7.
Success is Near!
I checked the operation of my tuner by re-reading the articles about the quadrature tuner by Harold Allen, W4MMC published earlier by antenneX. In the tuner I have a transformer with a resonated secondary winding, so it has to be 90 degrees phase difference between primary and secondary sides. I made several contacts to Norway and Germany and then to Siberia where I contacted RK9YXI, Eugeno. He gave me 56 report from Barnaul, near Novosibirsk about 3300 km east from Finland. I was quite near discovery of a small and simple tuner for crossed field antennas.!
Success Arrives!
I found a problem when testing the antenna with my friends OH2LJ and OH1VL—the feeder radiation was still there. I tested a ferrite balun over the feeder which helped but there were power losses in the ferrites and I didn’t want any losses in the system. I solved the problem by using an isolated link over the cold end of the bifilar coil and coupled the cold ends to the ground plate of the CFA. (Figure 8) The whole tuner is “hot”, so I have to use isolated box and knob with the capacitor. Now we have a simple tuner for all crossed field type antennas! It is easy to control the capacitor remotely or use digital automated control of the capacitance.
The tuning range of my last tuner above in Figure 8 is from 3.3 MHz to 6.2 MHz without gaps in the range! It is a one-band tuner. I have tested a smaller coil and shorter radiator and got the system tuned for frequencies between 6 through 12 MHz, but I have not had any QSOs on other frequency band than 80 meters with a 3-wire antenna. My friend OH3HEI built a 3-wire antenna using the normal European 3-wire power cord with twisted wires. The antenna had to be cut to about 3 meters to resonate on 80 meters.
I have tested a tapped version of the balun tuner, where the input feed was tapped to the primary coil 3 turn from the cold grounded end and the secondary (E-plate connection) tapped 15 turns from the cold end of the coil. The feeder radiation was still there, so I made the isolated feed with 3 turns link.
The tuning should always be done with low power. I tested this by driving 100 watts into the antenna and turning the capacitor. A strange sound was heard when the capacitor crossed the resonance point and later I short-circuited one toroid coil this way and found that my PVC wire isolation had melted. There must be high currents circulating in the coils during the tuning near the resonance .
After writing the accounts of my experiments above, I downloaded a copy the demo version of the Micro-Cap V program from their web page and tested it with my newly developed CFA tuner. The program is fully operative with minor limits.
I drew a picture (Figure 9 circuit) of my tuner with CFA antenna element capacitances and a 377 ohms resistor for power loss (radiation resistance). I found a place for this 377 ohms resistor by running several AC probe tests and changing the place of the resistor. I added a 50 ohms resistor at the input of the tuner to be able to show the input voltage drop at resonance.
In Figure 10 there are different RF voltages at the input and outputs of the tuner. The green curve is the input RF voltage. Blue curve is the output voltage from the primary coil connected to D-plate (377 ohms resistor in my simulation). Red curve is the output voltage of the secondary side tuned circuit connected to the E-cylinder (377 resistor in my simulation).
Figure 11 includes the phase curves of the output voltages. Blue curve is the phase of primary side voltage and red curve is the phase of the tuned secondary side voltage. From the voltage curve picture we can see how the input voltage, a green curve, drops a little when the resonance point is reached. The blue curve shows how the primary output voltage starts to rise when resonance is approached and then suddenly drops when the antenna starts to radiate. After resonance, the voltage rises slowly to about 1V level at about 20 MHz. The secondary tuned circuit voltage (red curve) rise is sharp at resonance and after that it ceases to very low value again.
In Figure 11 the output voltage phase curves show how they change with increasing frequency. The phase difference is less than 90 degrees before resonance and at the resonance the phases move fast up and down and, at one point, we have 90 degrees difference between them (actually at 2 points, but I couldn’t find them separately when tuning my CFA). Then after resonance the phase curves go down to something like –160 degrees without any sudden changes and the phase difference is less than 90 degrees all the way.
These simulated voltage curves show that we can expect relatively low voltages at the antenna. With 100 watts drive the assumed voltages are about 140V from D- and E- plates to ground and about 200V between the D- and E-plate.
The co-operation of the GARDS‘ team has given me plenty of new ideas and I think that I could not have gotten these results alone in the short time of a few months. We still have a lot of work to do to get the tuner to cover the whole HF band from 1.8 MHz to 30 MHz, but with every experiment, we get closer
Originally posted on the AntennaX Online Magazine by Heikki Antman, OH2BGC
Last Updated : 22nd May 2024