I tend to agree with 'Doc, this box might be nothing more than a remote electrical switch box...rather than mechanical. It could simply be switching the feed line port from one radiator to another, while switching the other two elements to ground in the process.
The link that Wavrider posted publishes the manual, but it is a pay-for deal to print out. It might be worth it for the owner if he doesn't have one. The text the author published, says these feed line jumpers should be a certain length, and he give us a formula he devised. He also says the coax is RG58, 50 ohm coax. So, I believe if this antenna really uses phasing...then 75 ohm coax would probably be used. The link also says the article is about building a modified 10 meter antenna.
Plus, I didn't look back, but I think Birddog and/or the manual has already told us the jumper lengths are shorter than a typical 11 meter 1/4 wavelength of coax. That said however, I would think a specific length might still be necessary for reasons other than phasing.
Just guessing.
Eddie,
you missed Birddog say the coax were 9ft, with a velocity factor of .66
that makes 12 ft a halfwave repeater and 6 ft a 1/4 wave inverter, 9 ft lies exactly 1/2 way between a 1/2 wave and a quarter wave which is a 3/8 wave, the manual states clearly not to change these lengths, its longer than a 1/4 section, not shorter.
unless you use different coax with a different velocity factor in which case you have too, to keep it 3/8 wave long. the 4th shorted matching stub is also 3/8 wave long at 9 ft also.
it doesn't ground the other two coax circuits, it leaves them open circuit so they become parasitic elements for maximum gain if i read it right.
there's more going on here than meets the eye., i'm not that familiar with 3/8 wave stubs (never had call to use one), and i haven't looked closely at element lengths either, or the schematic in detail as i haven't had time.
but i do know 50 ohm cable can be used for impedance transformation on resistive loads,gets a bit more complex when reactance is added,
for instance going between 25 ohms and 100 ohms would require a 1/4 wave of 50 ohm cable, the formula is square rt of the first impedance x second impedance or sq rt of 25 x 100 = 50 ohms,
so 1/4 wave transformers can be made from 50 ohms cable. i would imagine 3/8 wave can also transform impedance, only one that doesn't is 1/2 wave multiples as its an impedance repeater.
one we are more familiar with sq rt of 50x100 = 70.7ohms, what's used in cophasing, 75 ohm coax is chosen as its the closest off the shelf, but its not perfect so swr will rise slightly. we take a 50 ohm antenna use a 1/4 wave inverter of 75 ohm coax to get it too 100 ohms then we parallel 2 x 100 ohm points so transceiver sees 50 ohms, as 2 x paralleled 100ohms = 1/2 of one of the 100 ohm impedances, happy transceiver.
another situation we use the q section or odd 1/4 wave multiple is too match a 100 ohm feed quad, 100x50 = 70.7 ohms cable required, again 75 ohm off the shelf coax is used,
so we can raise it to a 100 ohms from 50 ohms or drop it from a hundred ohms to 50 by paralleling two antennas that have been raised to a 100 ohms each, both use a q section of 75 ohms, one requires 1 the other 2 q sections.
as CK says both reflectors are inactive and both act as reflectors and i think he may be right about the purpose of the two unused stubs tuning the reflector.possibly by the capacitance/inductance in them instead of impedance.
we saw a similar shorted stub on gainmaster, not sure this is just about phase at all, maybe more to do with impedance transformation in omni mode and reflector tuning in beam mode.
but once those 3 elements go omni, you have 3 x 72 ohm parallel elements which = 72/3 = 24 ohms, (although element closeness could change that like it does on a owa yagi), not good for a transceiver, I think the 4th 3/8 wave, shorted stub is to bring that too 50 ohms . bear in mind no matter whether its in beam mode or omni mode it is fed by a single 50 ohm feedline.
wish i had an analyser to see what a shorted 3/8 wave section measures at 27.185mhz, which this antenna is no doubt centred on as was for US market,i'll bet it makes an impedance/reactance when paralleled with 24 ohms gives 50 or thereabouts ohms in omni mode.
bare in mind an open 3/8 stub of coax is nothing more than a line of series inductance and parallel capacitance, i'm willing to bet the total reactance from this makes the 2 reflectors appear 5% longer than they are physically. which is what would be required to make them reflectors and what CK was getting at. The impedance match in beam mode may be helped by element spacing.
this was debated before and several mistakes were made, one i doubt very much the original manual states 29" as claimed by Radioman, i've blown it up and it still looks like 29', sure its a misprint as going by droop and knowing the original spreaders were 38.5" long, there is no way they could be 29", so must be 9ft or 3/8 wave.
others in this thread and last one also got wrong the one thing Radioman got right, 9ft of coax is NOT a 1/4 wave inverter or q section as its known, you have to include velocity factor, its a 3/8 wave section, just like the shorted stub.
http://www.worldwidedx.com/cb-antennas/151200-ms119-super-scanner-weird-swr-2.html
so all elements feedpoints are 135 degree phase difference to feed point at box. Not sure that makes any difference.