A small garden antenna project using remote antenna selection by relay switching.
Like most amateurs in the UK, my garden is rather more compact than I’d like and a 150′ tower with lots of aerial arrays won’t quite fit! My inverted-L for 80m rises just behind the tree in the picture to the left and just fits a quarter wave in.
For 160m, I have a loading coil in at the base of the 80m aerial and originally switched between the two bands with a toggle switch at the bottom of the garden. Of course you always want to change bands when it’s dark, wet and cold and after years of muddy feet and peering into the gloom I decided a better solution was needed.
Aerial Relay Selector Unit
My good friend Tony Emerson G0DLX, had an aerial selector unit that he had made some time ago and not used.
The unit had 4 relays which gave 5 aerial options!
The relays are controlled by a 3 wire connection; 2 signal wires and a common. Control is achieved by passing either +12v or -12v down one or other wire, selecting the appropriate relay.
The selection unit has 5 push buttons to route the control voltage appropriately down to the Aerial Relay Selector Unit and tucks neatly on the side of my shack bench.
G0DLX Circuit Diagram of the Aerial Relay Selector Unit
If you follow the signal path from the input at TX, you will see that the default state is to pass the signal through each relay, daisy-chain style, to output at A1.
Control signal voltages are input at points 1 and 3 against the common point between them. 1nF ceramic decoupling caps are used between each signal line and ground and each signal line also has a choke in series (not shown). +ve and -ve signals are separated then by use of diodes D1,2,3 and D4 which route the control voltages to the appropriate relay.
The transmission path, through the relays between A1 and A5, is then broken and routed to which ever relay has been activated.
The relays used were Varley DPDT 24V relays which switched at below 12v.
The relays were mounted on matching sockets; components were mounted on a tag board and tin plate screening was employed.
Installation of Mk 1 and its ultimate problems
In August 2008, the unit was installed and gave great service with me adding a 20m Delta Loop aerial to the unit too.
During the Autumn of 2010 I started getting erratic SWR problems. I had checked the unit, twice a year to check everything looked OK and generally it had done but when I opened the unit I was dismayed to find water pouring out and significant corrosion damage. Finding this in Autumn wasn’t good as the weather wasn’t suitable for doing any work to it. The Winter of 2009/10 had been more snowy and colder than average but Winter 2010/11 was exceptionally cold!
In March 2011, I was able to recover the unit and bring it inside for closer inspection. I found the whole interior of the unit looked rusty..even the plastic! 3 out of the 4 relays had fused coils and it appeared that damp trapped within the plastic covers had caused arcing and damage.
How the unit had become so wet cannot be 100% determined however the most likely cause is considered to be condensation. The unit had gaskets on all seals with petroleum jelly or silicon; it had no means of drainage other than a tiny hole in the bottom of the box. Subsequently, it was found that the drain hole wasn’t large enough to let water drain through it because of surface tension!
Aerial Relay Selector Unit – Mark 2
Combating condensation was of prime concern; in conversation with Tony G0DLX and Richard G4EIE, use of a fan was discussed and my US amateur friend, Joe Wonoski N1KHB, suggested running a car bulb to keep the interior warm. I was concerned that the original design had many metallic surfaces that aided condensation.
The new design uses a fibre board with the copper surface to the rear which has been treated with aero electronics standard anti-corrosion treatment. The new design, stayed with the original control circuit approach as this meant no changes were required indoors.
Only 3 aerials needed to be switched and this could be achieved with 2 relays. Tony had some more “beefy” relays, without damp retaining covers, that looked more up to the job and he suggested using the relays to ground each aerial when not in use. This approach couldn’t be used for the 80m and 160m options as the same quarter wave on 80m was used for both aerials and would require more relays so I decided to opt for a simpler design.
To operate 3 options, only one control connection is required as it has 3 states: +12v, 0v and -12v. The other control connection is left redundant as a spare. The control input uses the white 2 way connector; the common feeds both relays (see green/yellow wire); the control signal is sent to the original tag board using a red/brown wire. The tag board was cleaned up and actually has a dual set of components on there ready to work two additional relays but this extra set is also redundant and kept as a spare.
Consideration was given to remotely monitoring the temperature in the relay box through the redundant control circuit – see below.
The transmission path input is at the far right end of the 8 way connector block. It is fed to the relay on the right which defaults to passing the signal through to the relay on the left. The default (unpowered) state is to output on A1. With the left relay energised output is routed to A2 and with the right relay energised, to A3.
Voltages and Currents
The relays are 75-ohm each and consume around 150mA at 12v, roughly 2w. Although the relays were found to switch at 6v, these tests indicate that a better power supply should be used to ensure a good connection. This will be important to avoid arcing causing both more heat and potential carbon deposits on the contacts!
I decided to run power from the main shack PSU and so wanted to drop the voltage a little. Rummaging in the junk box, I came up with 5×68-ohm resistors in parallel which dropped the voltage to 11.7v and would stand well over half a watt dissipation in the resistors.
There is also a voltage drop to the bottom of the garden which brings the voltage down to 11.3v.
Avoiding Corrosion / Condensation
Assuming condensation to be the main problem, I opted for a free flow of air from bottom to top. The base has a window with a nylon gauze covering and the top has a chimney with nylon gauze at the top and a cowl! The connection at the top needs to be water tight and so a hose connector with washer was used.
Hosepipe is connected to the hose connector and kept rigid with some PVC water pipe pushed over the top. Black water pipe was used, primarily because I had some spare but also because it would help keep the chimney warmer and so draw an up draught of air through the unit.
At the top of the chimney, nylon gauze was stretched across to prevent insect access and a PVC Tee-piece used as a cowl to prevent rain access.
The aerial coaxes, ground line and relay control cables all enter the unit at an angle and silicon paste + nylon gauze is used to create a seal. Twine was used to maintain the angles for the cables to ensure water run off is away from the unit.
Installation of Mk 2
The completed unit was installed in May 2011 and is shown below. The loading coil for 160m is shown to the left and the plastic milk bottles are used as a counter weight for the pulley on the 80/160m inverted-L.
Monitoring relay box temperature remotely
A couple of years ago I got a cheap and cheerful ATECH Weather Station which included a wired temperature sensor.
I chopped the remote temperature sensor cable which was about 8′ long and connected it up to the spare control channel to the relay box and was somewhat amazed to find that it worked fine!
We had frost the first night followed by a beautiful sunny daytime and it registered -1.7C min to +22.9C max suggesting that conditions for condensation did exist in the Mk 1 unit.
Durability in weather extremes
A year after the Mk 2 unit was installed, I found a dry day to inspect it for any signs of water or other damage. The winter of 2011-12 has been a long wet one in Yorkshire and to my delight the Mk 2 unit looked almost like new. No signs of corrosion whatsoever!
Last updated: May 2012