Martin Francis - NDB K9AY mast (Revised 13th July, 2004)

[ Main DX Pages ]
[ Modifications | Mast (Version 1) | Balun | Vactrol | Relay Box | Controller
Mast Problems | Mast (Planned Version 2) | Mast (Actual Version 2) | Cable Problems ]

This antenna was developed using details published here, but with certain modifications. And there's a RIGHT order to do the work in:

  1. Plan the mast location;
  2. Measure up for coax and controller wire;
  3. Erect the antenna;
  4. Sink your earthing posts (suggestion from Michael Silvers - pack the earth holes with moisture absorbent Kitty Litter to give a good earth even in the hight of summer!)
  5. Wind your torroidal transformer (the 'Balun');
  6. Measure the optimal values for terminating resistance with a portable set;
  7. Build your Relay box (with a vactrol capable of spanning required range);
  8. Dig a trench and bury your cables;
  9. Build the controller box;
  10. Mark the null points on the controller box for each pair of loop directions.

Modifications [ Top ]

The original article by Gary Breed on which this antenna was based used a very clever idea for relay switching via a control voltage up the antenna cable - a positive DC for the first relay, a negative one for the second, and AC for both at once. Only problem was, all that AC power up the lead caused me serious interference problems so instead I applied the modification later proposed by Gary Breed (see ARRL More Wire Antenna Classics Vol 2, or QST May 1998) - the use of a separate pair of DC control wires to control the switching - feeding DC up to the mast rather than AC. The results - much better performance in the NW direction. I'd recommend buying a multi-way screened cable for your control functions - if money is no object try a CAT-5 networking type cable with a screen, this way when you want to start adding switchable remote preamps and so on, you've got cable ways to spare without having to dig everything up again - plus, spare wires should your cable ever be partially damaged!

Mast (Version 1) [ Top ]

Take bearings first then drive in your four corner posts (mine were galvanised steel, 3 feet long, with holes drilled to take the eye bolts - use these to tighten everything up once the antenna is up and all the guys fastened down). Make a hole for the mast - I lined my hole with bricks to provide a good key. I recommend the NW/SE system as this results in signals switched to N,E,S and W. My mast was constructed of 3 ten foot plastic tubes, 1.75 inch diameter, with two sets of guys at the first and second levels. Plastic masts don't attract lightning so much as metal ones, and also don't mess up the antenna patterns. Pretty scary to set up - get a STRONG friend (I was very glad to have my father-in-law helping) and a couple of sturdy self-supporting ladders and pray for a calm day! Fasten the antenna wires (effectively the top guys) first, and let the nylon ropes dangle (better still, tape them to the sides of the mast temporarily before you start, so the wind doesn't make the whole thing into a maypole!) until you've tied the top antenna wires safely.

Balun [ Top ]

I wanted to use the system for LF work so wound a torroidal transformer with a primary inductance of 1.9mH (minimum usable frequency of 150KHz) - this worked out at 140 turns on my torroid for the primary. Now here's a great tip I came up with for you if you've ever TRIED winding large coils on a torroid - measure the length of wire required for a turn first, then measure out the total length required, plus a bit, and wind it onto one of those plastic wall mounting plugs and use this as a winding bobbin - you can even hook the wire into the slit if your hand needs a rest! Also, I wasn't using 50ohm coax but instead had the required 300 feet of 75 ohm TV coax to get from the bottom of the garden to my house(OK for this LF stuff), so I wound for an impedance ratio of 6:1 instead of 9:1 (2.5:1 turns ratio instead of 3:1) - this meant that the secondary needed to be 47 turns.

Vactrol [ Top ]

I urge you to test the antenna nulls before you build your controller box and find the optimum termination resistances before you even start to build your vactrol - with the wrong terminating resistance the system's virtually useless. You'll want to find a couple of stations optimally placed to enable you to ascertain your null points. I used a portable radio next to the mast with the balun feeding a short length of coax into the set, a couple of crocodile clip leads to make the connections, and a 1Kohm variable resistor (also with crocodile clip leads) to find the optimum values for termination resistance. I found that the value was different for each of the two loops. I then designed and tested an 8-LDR vacrtol array to ensure that it covered the required range (as low as 200 ohms for one of my two loops) - if your vactrol won't go low enough for your antenna system, you're pretty much wasting your time.

Relay Box Housing [ Top ]

Make sure you buy a big enough box - my first attempt was VERY cramped. By all means seal the ports with hot glue, but don't be tempted (as I was) to fix the entire housing with hot glue - seemed a great idea until I wanted to make a change (see note on Modifications)... I placed the box in the oven at 100 degrees Celcius to remove the glue and discovered that the ABS plastic box melted at about 90!.
Another tip is to fix an LED on the outside of the box to verify that power is reaching the vactrol actuator and the two relays, even if someone has been digging in the garden - you'll thank me for that!

Controller Box [ Top ]

I recommend using a 4 way switch with several poles - this way you can use one of the pole sets to control a series of LEDs indicating the direction. I broke off the end stops for my rotary switch, so it can be turned continuously round without stopping. Wire the switch so that the LEDs indicating direction rotate the same way as the switch movement (i.e. not the way shown in the original circuit as this results in an anti-clockwise movement for a clockwise change which is quite confusing). Mark the positions for optimum nulls for both pairs of loop directions. You could use preset resistors and the third set of poles on the switch to do this automatically.

Results? [ Top ]

I use my system for chasing navigation beacons in the 190-520KHz band - see highlights of my reception log here, or the whole thing here.

Update: Mast problems [ Top ]

The mast described does have a tendency to bend - in severe winds it can even 'banana over' completely, requiring manual correction.

Mast (Planned Version 2) [ Top ]

In November 2003 (after taking the lovely bittersweet sunrise photo above), I decided to rebuild the mast. My first plan was to use plastic again, this time constructing a lattice mast as shown in this design:

Mast (Actual Version 2) [ Top ]

The cost of all those plastic sections (espacially the 55 T sections) was prohibitive though, so I was delighted when I discovered our next door neighbour had an old TV mast in sections at the bottom of his garden. I know the design specs for the K9AY don't recommend metal masts as this causes loss of directivity - I decided to use a combination: plastic at the top (the top twenty feet of my old mast) with steel frame cased in concrete at the bottom. My father-in-law John Tornambe (shown in the photos) was very kind not only to let me construct the mast at the bottom of his garden, but also to mix the concrete for the base and help me erect the mast once again. I took the oportunity increase the height and base of the antenna, taking it from about 80 to 140 fett diameter per loop. When digging a hole for your mast, in Canada at least a four-foot deep hole is required to ensure that the base is below the frost-depth in winter, or inch-by-inch your mast will gradually be pushed out of the ground after a couple of freeze/thaw cycles.

Cable Problems [ Top ]

One problem I have had is that the cable has been cut a couple of times where it passes under the vegetable plot shown in the the photo. This has now been corrected by reterminating both controller and signal leads and burying then in galvanised steel conduit between the vegetable plot and now right up to the antenna itself.