Rebuilding a Mosley triband yagi.
A while back I acquired most of a Mosley MP-33 triband which had been modified to be taken on a DXpedition 20ish years ago. The parts I got were the director, reflector and boom, so my task was to build a driven element.
My approach was to try and replicate the original design as accurately as possible based on the manual, but of course also doing this in a thrifty way, using as many scrap/recycled parts as possible. The driven element is a dipole with a trap at either end which uses the exterior of the capacitor tube as the driven element for 15m, and the tip as the driven for 20m. The trap tube I found had the same OD as the original, but is a little thicker, and the driven element tubes are a little smaller OD to the original, which has an unknown effect on the capacitance: I did not get a consistent measurement of the original trap, and the capacitance is changed by the smallest of environmental movements, like me trying to take a measurement. I found a 3d model for a replacement coil winder for the trap, and printed them in PLA+, a material which I have used extensively in antennas, and found that it holds up pretty well even in Arizona heat. The winders are like spools or bobbins, with a threaded groove so that uninsulated wire can be used for the traps. Though the manual suggested #10 (which I believe to be 10AWG), the pitch on the winder is 1.5mm at a diameter of ~20mm. The Vernier suggests that the original traps used silver plated copper wire at 16AWG, but I chose bare copper 18AWG since the 1mm wire should be plenty for 100w, and has better clearance on my imperfect winders, and it was also what I had to hand.
Copying the winding turn numbers, I assembled the antenna and put it up on the mast. I had no idea how different the readings would be on the antenna on the mast vs tilted over at the edge of the roof: turns out there was almost no change. I took a sweep on the antenna, and its 10m resonance was at 34MHz, with another dip somewhere near 20m. Confused I tried to find some input from friends and forums with little use. The driven element without the trap section (removing it would take about 150mm off the length of the element) was beautifully resonant on 10m.
HOW TO TUNE A TRAP:
Now, this is what I found to work, I don’t know how things would differ if I were modifying the director/reflector traps, but the method works for the driven elements, so I could use them to make a driven element for the WARC bands (Mosley sells one of these).
Firstly, all measurements need to be made with the trap isolated from objects: I used string from the ceiling. After re-winding coils adding a couple at a time, I concluded it was far less wasteful to wind the full length of the spool and tap the coil as needed.
The measurement that seems to work well is to measure the coil on its own first, and aim for a dip right before the target frequency. I did this using my RigExpert AA-35 in R/X mode, the dip in X meaning that the reading passes from inductive to capacitive just below the target frequency, which is the bottom of the band. This reading should be easy to make with a NanoVNA or an L/C meter.
When you reassemble the trap, the dip can move a little since you are adding objects in proximity as well as the capacitor to the L/C circuit: I had to make an adjustment or two. Measuring the complete trap (for a single band) is taken from the outside of the capacitor which-in this case-is the radiating element for 20m and the radiating element for 15m. I found that measuring both traps together was not showing the dips as you would expect, so it was not a useful test.
When I put the newly adjusted trap sections back on the antenna (dangling over the edge of the roof), the complete driven element showed the three dips just below the target frequencies, so I ended up shortening the 10m section by 50mm which brought 10/15m into resonance, and I also had to shorten the tips by 50mm to get 20m in a good place. One thing that is difficult with traps is that the coils do add a loading to the antenna, so adding a wind does lower the resonant frequency of the trap, as well as shortening the physical length of the radiator. I had to go back and forth from the antenna to the shed several times making very small adjustments before I decided to call it good and fixing everything in place with sheet metal screws.
In conclusion, it is very possible to build the driven element for a trapped Yagi (or a trapped dipole). Using the Mosley trap design is not as easy to navigate as a traditional coax trap or capacitor and coil, since these approaches have a lot of documentation to help. However, starting with the medieval approach of sculpting something that looks the same and adjusting it to work with a newfangled approach of measuring and adjusting is a completely viable approach. Is my antenna as good as it was in factory form? I have absolutely no idea, but what I can tell you is that I just spoke to a fellow Scot in Valencia on 10m, I had him at 5/7 on the yagi, and couldn’t hear him at all on the Cushcraft R5. Tests last night showed that a signal I could hear at S5 on the R5 with a noise floor of S3 was also S5 on the Yagi, but with no noise (as in RF gain dialed back a pinch to remove it).
If you need help with a similar project, feel free to comment here or email me. I am no expert, but I do have some notes and the antenna does work well.
Filed under: blog - @ February 23, 2025 7:13 pm