Auxiliary Antenna Switch

Woe to those with too many antennas! No, not really, but it does create a switching nightmare that isn't present in most stations. I have finally reached the capacity of the switching hardware in my station.

The core of my switching system is the 2×8 Hamplus antenna switch that is at the base of the closest tower. Putting it there means there are only two coax runs into the house, one for each operating position. Eliminating that multitude of coax runs into the house simplifies cable management and keeps the holes in the wall quite small. 

The cost of the remote switch is ~18 control lines from the shack to the switch (3 × Cat5), weatherproofing (and insect-proofing) and occasionally having to stomp through the snow drifts for mid-winter maintenance. Few contesters of my acquaintance do it the way they have. They prefer staying indoors to work on their switching systems.

I've been able to stay within the 8 antenna capacity of the switch until now mostly by luck. I freed one port when I moved the VHF antenna (6 meters) to a separate run of Heliax into the house. The Acom A1500 amplifier has 3 antenna ports, and I do the switching there. But the 6 meter antenna can't be accessed by the other operating position unless I add an antenna switch inside the shack. There is little point since the second station's L7 amp doesn't have 6 meters.

When I add a 2 meter antenna, I may switch them at the top of the tower with a remote switch. I have no great need to access both antennas at the same time. How I switch the coax within the shack will depend on what I settle on for 2 meter equipment. For now it is enough that I have removed VHF antennas from occupying ports on the 2×8 switch. 

Before installing the additional transmission line, I could plug either the 6 meter or 160 meter antenna into the 2×8 switch. I did it once in the spring and once in the fall. It's good that I no longer have to do this since 6 meters has become a year round band with the rising solar flux. I am also investigating methods to make the 160 meter vertical a year round antenna rather than having to roll out and roll up the radials before and after the winter season.

The addition of the 80 meter inverted vee spells the end of seasonal port swapping as a viable strategy. The 80 meter vertical yagi had to be disconnected. That's not a problem right now since it developed a fault this winter and isn't working properly. But I'll fix it soon and I need an antenna port to plug it into.

The 8 ports are currently for: 160, 80, 40 (2), 20, 15, 10 and the TH6. The time had come to build auxiliary antenna switches for 80 and 40 meters. There is rarely a reason that more than one antenna for these bands would be in use at the same time. By freeing up a port from the two allocated to 40 meters (3-element yagi and XM240) I can use it for another multi-band antenna or perhaps antennas for the WARC bands, when I have them. 

One antenna switch port per contest band makes it possible to place single band high power BPF at that point in the switching system. No matter which antenna is selected for each band, the BPF will be the correct one. That can greatly simplify use of BPF since no switching is required. I am happy with my 6-band switched low power BPF and there is no plan to spend big on high power BPF.

Hardware

I used the smallest sheet metal aluminum enclosure that would comfortably accommodate all the parts and leave finger room to manipulate the connectors. Small size keeps the lead lengths short, and that minimizes impedance "bumps".

There are 4 different styles of UHF chassis sockets. I used what I had, and one I had to purchase at a local flea market. Extra holes were drilled to increase the number of sheet metal screws. That is to improve the seal against water and insects. There is no electrical reason to do that at HF; the wavelengths involved are too long to "leak". I was sloppy about it but there are no bonus points for attractiveness!

The +12 VDC control line terminals are #4 screws through plastic flanges that isolate the screw from the chassis. It's adequate and avoids the trouble of mounting a multi-conductor connector of some kind to the enclosure.

The 4 antenna ports point down for additional weather protection. The angle bracket on the other half of the enclosure is for screwing the switch to a wood panel alongside the 2×8 remote antenna switch.

There are no surprises inside the switch. AWG 18 wires carry RF. The SPST relays are positioned to minimize the lead length. Solder lugs ease wiring of the relay coils and suppressor diodes (1N4007). The relays are mechanically supported by the RF wires. The TE System high current relays are used in antenna switches throughout my station. They easily handle a kilowatt provided the SWR is not too high and hot switching is avoided.

Performance measurements

This is the kind of switch that hardly rates testing beyond a continuity test with the relay turned off and on. RF at 40 and 80 meters is very forgiving of sloppy layout and wiring. Nevertheless, it is worthwhile to take a few measurements for peace of mind and to see how it does at higher frequencies. I will likely need one or more similar auxiliary switches for the high HF bands and VHF as I continue to add antennas.

Testing the switch does not require a 2-port VNA when you have an accurate single port antenna analyzer. Two ports to measure insertion loss is pointless for this circuit, and port isolation is likely to be quite small. There is no requirement for high isolation between the antenna ports of each switch since only one of those antennas is used at a time. High isolation is almost certain between the two switches at low HF frequencies. I considered and rejected putting a shield between them.

A sweep of the SWR from 2 to 30 MHz is about what I expected. At 30 MHz the impedance is approximately 52 + j3.5 Ω. This is about j2 Ω higher than the 50 Ω load alone. That is due to the inductive reactance of the internal wiring and relay. Both ports of both switches have identical impedance curves(within measurement error).

The impedance at 50 MHz is higher at about 53.5 + j5.5 Ω. The 1.15 SWR is low enough for most applications. I experimented with TLW to design an L-network that would compensate for the stray inductance. As for the 10 meter stack switch and its longer wires, a shunt capacitor is sufficient since the calculated series L is no more than 0.05 μH. A value of between 20 and 25 pf works well to correct the small mismatch on both 10 and 6 meters. Addition of the capacitor hardly disturbs the match at lower frequencies. 

I'll eventually want a switch to share one run of Heliax on 6 and 2 meters. Compensation for the stray inductance would be needed at VHF even if I make the switch more compact with less stray inductance since the path through the relays can't be shortened. The present switch for 40 and 80 meters needs no compensation for excellent performance.

Software

Port sharing requires additional conflict management in the recently completed station automation system. The Arduino software and switching software already has 8 auxiliary control lines for controlling devices such as this. Each antenna has pair of values specifying the port on the 2×8 antenna switch and auxiliary control line (or none). However the Arduino software does not yet prevent conflicts. That job is left for the hardware lockout in the 2×8 switch.

The UI (user interface) software running on the PC has been enhanced to check for port conflicts. If another radio is using the same port as that for a selected radio, the antenna is unavailable. In almost all cases this should only occur with multi-band antennas since, for each port with two or more antennas, those antennas are for the same contest band. 

Antenna selection for mono-band antennas is by auxiliary switch, like the one described in this article, or stack switch (currently only for the 10, 15 and 20 meter stacks). In a contest there would rarely be a port sharing conflict for radios on different bands.

Installation

The auxiliary switch has been installed and connected to the 80 meter port of the 2×8 antenna switch. The photo shows it mounted under the rain and snow shield before weatherproofing was completed. Both 80 meter antennas -- the inverted vee and the vertical yagi -- are connected to the switch. 

The control lines are not yet connected so only the inverted vee is currently accessible -- I made it the "default" for the time being. I used a permanent market to label the NC (default) port and the NO port, corresponding to the relay contacts. The letters are barely visible in the photograph.

The software and hardware additions to the automation system are in place and partially tested. Final testing has to wait until I patch the auxiliary switch Cat5 cable to one of the spare cables running to the tower base. That job will be completed in the coming days now that warmer weather has finally arrived. I will only take the opportunity to diagnose and repair the fault with the 80 meter yagi. The 40 meter antennas will be moved to the auxiliary switch later this month.