Multicoupling

[Pump3]

Hey guys,
Im looking to learn some about Multicoupling....
Are there any good articles out there describing how it works, how you would tune and cable a set, I have 3 sets multicouplers I want to use but I didnt get the wiring harness's with them and I have an idea how it works I just want to be positive...
Is it possible to run a repeater off a 1 set of multicouplers and nothing more? A few years back I was in a shack that used to share 1 (both tx and rx) antenna between Police and Public Works, but Public Works moved to a different site, so when they left it looked like they had 8 cans, 4 for tx, 4 for rx for the one repeater with a dummy load on the end of the "chain"

Im talking about Sinclair C2037's for the most part not that it makes much difference, I would love to see a block diagram of a multicoupled system and how the wiring harness is done...

I know usually with a lot of multicoupled systems they employ 2 antennas one for Rx and the other for TX, but im looking at possibly doing it with 1 antenna for 2 or 3 repeaters.... also what kind of losses would I be looking at? One figure said it could be as much as 50-75% yikes....thats a lot when you only want to run like three 30watt repeaters

[Pump3]

Found this excellent article from Sinclair, ill post it here for future reference:
http://www.dstar.ca/docs/C-Series-Multicouplers.pdf

but I'd still love to hear any comments anyone may have!

[Bill_G]

Those are some pretty cool cavities, aren't they? Sinclair really understands rf. And yes, with something like their system, you can use just one antenna per band for both transmit and receive for many radios at once. These are perfect for a three channel UHF LTR at a convention center or a mall driving radiax to get maximum indoor coverage, or for a city five channel VHF system on top of city hall where you do not want the rooftop to look like a ham fest. These really do cut down on the antenna clutter on a tower, and reduce your feedline costs.

But! (there's always a but ...)

You better not plan on ever making any changes to your freqs once you have settled on this system. The cable lengths between cavities is critical and must be expertly made. They are not easily field modified by the average technician.

You must have at least 800kc minimum separation between channels, and preferably more to reduce your losses. Other transmitter combiners can be as little as 250kc (or less in hybrid combiners) between channels if you use a separate rcvr antenna at a different elevation.

These are more expensive than typical rcvr multicoupler and transmitter combining systems. You trade a single antenna and feedline for a more expensive combining system. You will have to crunch the numbers to see which is best.

If tower loading is an issue, or if visual blight on the roof is a consideration, if your licensed freqs meet the minimum split requirements, and you plan on never making any changes, these are great. If you want something more flexible, then consider other systems.

[escomm]

And ho boy are them puppies expensive!!!

[Bill_G]

But, they are purdy.

[Bill_G]

Back to the OP - you asked how to calculate the critical cable lengths between the cavities - In general the short interconnecting cables are 1/4 wave or 3/4 wave of the pass freq, but some experimentation is required.

[Pump3]

Where exactly do you start measuring from? One thing I read said to include the coupling loop inside the can as well....
Thanks

These guys have some excellent articles too ill post for reference http://www.emrcorp.com/techinfo.php

[Bill_G]

It's tip to tip of the cable you are building. So, you have to take the physical length of the connector into consideration. That's why people keep handfuls of right angle adapters around when building these - so they can add length easily if necessary.

[psapengineer]

Rx Multicoupling is a great thing at UHF and 800 where the Tx and Rx frequencies are in splits that are significantly apart. It's much harder to apply in the VHF band. Instead, at VHF, you often have to resort to individual cavities per leg, much like a Tx combiner.

[grandnational]

The interconnecting cables are measured from the nested stop of the center pin inside the connector end-to-end for 1/4 wavelength or 3/4 wavelength times the velocity factor of the cable. In addition to that, some fudging is often required to get it "just right". No cable will be 100% repeatable. Everything is manufactured/spec'd to a tolerance, of course, not exactitude. Each connection will introduce loss, phase error, etc. as will the cavities themselves since they aren't a constant 50 ohms - far from it. Tuning individual cavities and connecting them together can give you very different results than you might first expect. Tune them as a network for best results, and take your time to observe the interactions.

The reference of "nested stop" of the center pin may need some explaining. Looking into your typical N-Male connector, the pin will have a tapered tip with a flat area inset from this apex. The apex nests inside the N-Female "fingers" & the flat surface just mates with their edges when the connector is properly torqued. So when connected, the apex is electrically irrelevant to the measurement for our purposes. This will vary by connector, however Type-N, BNC, TNC & Type-C connectors are VERY similar in their design. UHF connectors, on the other hand, are just junk. (Not very bashful about that) Avoid them above 300Mc & certainly in duplexer/filter applications. Physically, they're inherently problematic in design, as they are electrically. If measuring the stop is difficult, measure from pin tip to pin tip & this will get you close enough.

Use quality cable. No RG-58C/U or RG-8A/U here. Always double-shielded, and PTFE (Teflon) insulation/dielectric whenever possible. RG-142B/U or RG-400/U make good choices, as does RG-223/U. 223 isn't Teflon, but still good. 223 & 142 are solid center, 400 is stranded. Teflon is a luxury when you may be soldering the centerpin several times when testing, without corrupting the dielectric by the thermal cycling.

Here's a good article about cable:

http://www.picwire.com/technical/paper2.html

They have other good articles about grounding & such also worth reading. Cavities/Filters/Isolators, RF networks period, are very fun to work with. You don't need expensive network analyzers & full RF lab to do it. All the fancy tools are nice, sadly hobbyists seldom can afford them. There's a lot you can do experimenting with a bit of knowledge to jumpstart you.

On ham/low budgets, a full TX-RX isn't usually in the cards. It's easier to combine each separately using two antennas. Having all that transmitter energy in the mix while trying to receive in multiple places, avoid interactions, have decent noise figure, etc. is complex. I'd start with home-brewing your receiver system. Then collect some circulators & start designing your transmit combiner. In the meantime, perhaps individual TX antennas. The cable rules for your TX combiner are basically the same as your RX combiner system. Some fiddling is required when you go beyond 2 channels where T connectors are introduced. Most T connectors give you an impedance bump that you'll have to correct against. One channel begins the chain, goes to a T, the outboard end of the T going to the next channel, the inline end going to the next T & the next two channels, etc. Better than this is a properly constructed "boss" junctioning the many connections together to a single connector.

Again, thinking hammy/ham budget. Put it together Xmitter->Isolator->Pass Can->T Junction. Be sure your antenna will handle the total power & assume roughly 2dB loss through your combiner on a good day. The receiver system goes together as Antenna->Block Filter->Pre-Amp->Splitter->Receivers. The block filter gives you the desired band-pass window, feeding a good pre-amp with a high intercept (such as an Angle Linear model), which feeds your splitter & to your receivers. Calculate the system losses of the filter & splitter, selecting your pre-amp appropriately. There can be "too much" gain. Most splitters have at least a 3dB loss. Aim for roughly 8-10dB of gain at your receivers after losses.

Good luck! And most of all, enjoy.

[Astro Spectra]

I've done a bit of multi-coupling work on VHF and UHF. Always use two stacked antennas. It is so much easier to keep TX and RX separate. You get typically get 20 to 30 dB of isolation just from the vertical separation.

Some of the commercial 4 dipole stacks can be ordered with two feeders, one to the top two diploes and one to the bottom two. Even with typical stack spacing you still get more than 20 dB isolation. These antennas are good to work with as you know the TX/RX isolation before you go on to the site. Mounting two individual antennas on the tower means that you normally need to take test equipment to the site to measure the separation which usually means a later trip after you done the combiner design and build.

Note that there are two combining TX situations. If the transmit frequencies are close you need to look at a hybrid type combiner and you’ll lose half the power into the load. If you’re combing two channels that are at least 250 kHz apart on VHF you can use a pair of bandpass cavities per transmitter combined into the ‘spider’ or 'boss' (usually three, five, or more N connectors mounted on a block). This is usually a simple T if there are just two transmitters. You will need a circulator on the output of each transmitter to avoid IMD.