Implementation of Remote polarization switching
Last updated 26 May 2000
This is based on an article appearing in both the RSGB Communications Handbook 7th Edition, so I take absolutely no credit for the idea. The original article suggested using PCB RF Relays, and creating a PCB with 50Ω striplines. My own attempts at making PCBs have been so bad that I didn't even start to go that way, although the use of PCB RF Relays dead-bug style was used for a couple of attempts.
I've attempted to work out a scheme like this using + rather than X configuration, but I seem to end up with a number of extra relays severely adding to losses and SWR degradation.
Coax lengths need to include any velocity factors in calculations. I used an MFJ-259 to check the velocity factor of a fairly long (about 30'/10m) piece of coax. I found the MFJ-259 was not very useful for cutting 1/4 wavelength sections at 144MHz, as there seems to be another 1.5" or so of length inside the unit. Instead I cut the 1/4 wave stub to about 10% or so longer than expected from the velocity factor calculation and then started cutting, checking the SWR at each end of the band in question to see if I was getting close (or too far!). Be prepared for several abortive attempts. Patience (and some spare coax) is a virtue here. Remember that adding connectors can add to length. Getting an SWR much less than about 1.2:1 on 144MHz or 1.4:1 on 435MHz I found to be a waste of effort.
I have done implementations of the above successfully in the following ways:
The last method really is the way to go when cost is not an issue. Remember that not only do you have to purchase the relays, but also all those N-type connectors.
Update 26 May 2000...
Regarding (a) the relative position of the driven elements on the boom and (b) feeder lengths between the switch and the antennas. Jim Walls K6CCC mentioned that this issue hadn't been raised in this note. The design assumes that (a) the driven elements, although at 90° to each other, are at the same position on the boom and (b) the feeder length to each yagi is the same.
In real life, it's often not possible to have the DEs at the same position on the boom (indeed some manufacturers have deliberately placed them at ¼λ for CP), so you'd want to compensate for this by adding an equivalent amount of coax to the feeder of the antenna that's furthest forward, taking into account feeder velocity factor.
For example, I had two 7 ele ZL specials on the same boom, but the -45° was 3" in front of the +45° yagi. With a feeder VF of 67%, I added an additional 2" to the coax feeder to the -45° yagi to get them back in phase.
Mail Howard, G6LVB