Duplexer Construction – Part 1 (The Components)

Fully stripped filter cabinets

The bare cabinets. Just out of view is a BIG pile of hardware!

Alright, it has been exciting to launch the repeater, and I’ve been focused on that the past couple of weeks – but let’s get back to how it got here!  When we last talked about the duplexer, I was deciding whether to build in the 6-cavity cabinet, or the 4-cavity cabinet.  You know by now that I decided on the 4-cavity cabinet, but why?

Advantages for the 6-cavity design are that I could include a bandpass filter for receive, which may or may not be necessary for protection of the receiver, due to nearby broadcast transmitters.  Advantages for the 4-cavity design would be smaller size and easier transportation (the whole thing fits in my wife’s Honda CR-V), and the inside partitions were nearly complete.  I was leaning toward the 4-cavity cabinet, but decided that the very first thing I needed to do was to build a prototype cavity in the 6-cavity box first, to see how it performed.  That way, if the duplexer is built in the 4-cavity box, I’ll still have the prototype to experiment with.

Inside the filter

Inside the bandpass filter. Click on the image for a super Hi-Res version.

Take another look at the 6-cavity cabinet (before it was fully desecrated!)  Notice that the left-most resonator has a longer bottom section.  This is because it needs to tune to a lower frequency, since it masks (or shunts) the low side of the passband.  For TV channel 3, this means about 60 MHz.  Assuming that the same design cabinet would be used for TV channel 2, which covers 54 to 60 MHz, I expected that the longest resonator could probably be tuned below 54 MHz.  The other resonators were all a few inches shorter and probably couldn’t be tuned that low.  LOTS of assumptions.  I needed answers!  “How low would it go” was the most important question that the prototype would answer.

Silver-plated telescoping resonator stubs.

Silver-plated, telescoping resonator stubs. NOT your average hardware store item!

Taking apart the diplexer and filter cabinets yielded a very nice pile of high quality metals; most of the cabinet was plated aluminum (not sure with what), and the resonators were silver-plated copper pipe with silver-plated finger stock in the slip joints.  Fancy stuff!  Building high-quality, stable, telescoping resonator stubs is the hardest part of a homebrew duplexer, so it is nice to have that work done for us!  Overall, it appeared that there was enough material that could be cut and reassembled to make duplexer cavities, but there wasn’t much room for mistakes or do-overs.  I did NOT want to buy any additional metal!  It was important to get it right the first time, important to do the research.

Generally speaking, there are three major components to a resonant cavity:

  1. The enclosure
  2. The resonator stub
  3. The coupling loop(s)

The Enclosure

The enclosures most often used for resonant cavities are usually round pipes, but they can also be square like the ones we’re working with.  Ours happen to be about 8″ square, and about 4 ft. 3 in. tall.  Eight inches square is perfectly fine, but only 4 ft. 3 in. tall?  That’s Aluminum partscutting it awfully close for a quarter wave at 6 meters.  In space (rather than in wire or metal), a quarter wave at 52.67 MHz is approximately 246 / 52.67 = 4.67 ft., a few inches LONGER than our enclosure.  We’ll have to hope that our prototype cavity has a low enough velocity factor to make it fit!

Notice that if we build the prototype at the left end of this enclosure, we already have a top (the 10mm thick plate leaning against the side), bottom, front (removed for the picture), back, and one side; we just need to build a bigger partition to close it off on the inside.  No problem, we’ll just borrow some partition material from elsewhere in the cabinet, and cut it to fill out the partition to reach all the way from the top to the bottom.  That will give us a fully enclosed box that measures about 8″ x 8″ x 51″.  If it works OK, we can build more cavities using a similar method.

The Resonator Stub

A resonant cavity is just a 1/4 wave section of coax – very LARGE diameter coax – open at one end and shorted at the other.  The length of the center conductor determines the resonant frequency.  You can’t accurately measure the center conductor and cut it to length ahead of time; you need some way to fine tune it by adjusting the length after it is in Telescoping resonator stubthe enclosure.  Usually some sort of telescoping section is used to accomplish this, and the resonator stubs from the old combiner and filter cabinets do just that.  We have TEN of them to work with, but we only need 4 (or maybe 5 or 6 if we add bandpass cavities), so maybe we can somehow combine them to make them longer, if necessary.

Something worth noting about these resonator stubs is the diameter, about 2-3/4″ for the top section, and about 4″ for the bottom.  That is somewhat larger than ideal for an 8″ square cavity.  The more air space you have between the cavity walls and the center stub, the better performance you get.  Commercial designs for duplexer cavities usually have thinner stubs, which results in a longer cavity, larger cavity volume, and higher “Q” (sharper tuning).  However, you can’t change any ONE thing in a cavity filter without affecting some OTHER thing.  In this case, that’s OK, because having a fatter stub means that it will be shorter, and our cabinet isn’t very tall.  A skinnier stub would probably be too long to fit, so having a little lower “Q” is a reasonable tradeoff for being able to make it fit!

The Coupling Loop

The diplexer and filter cabinets did not have coupling loops that were useful for our application.  We will have to fabricate our own.  Coupling loops are really just half of a transformer; they magnetically couple the RF into the center resonator stub (the other

The prototype coupling loop.

The prototype coupling loop, made with thick strap scavenged from the TV diplexer.

half of the transformer).  The basic design is shown in the photo; one end of the loop connects to the coax center conductor, the other end connects to a capacitor (for a reject cavity), or to ground (for a bandpass cavity).  Using a connector for both ends of the loop gives you the flexibility to go either way.  Different types of capacitors may be used, and can be mounted either inside or outside the cavity (we’re using a connector, so the capacitor will be outside).

As far as mounting the coupling loop to the top plate, the most common design uses a disc that fits into a hole, and is screwed in place.  To get the desired filter performance, you adjust the amount of coupling by loosening the screws and rotating the disc.  This alters the alignment of the magnetic field inside the cavity, and therefore alters the filter performance.  The top plate of our cabinet is 10mm thick aluminum, so we’ll make some discs out of material that was recovered from disassembling the cabinets, and make some holes in the top plate to receive the discs.

OK, we have a basic plan to cover the basic elements of cavity design.  I’m really itching to get a prototype assembled, and answer the BIG questions: How long does the resonator stub need to be for 53.67 (TX) and 52.67 MHz (RX), and will it fit in our cabinet?  And, how well will it work (how high will the “Q” be?)  In our next installment, we’ll see more particulars of how the prototype cavity was put together, and see how it performed.

Advertisements

About kd2sl

I'm a lifelong geek, interested in anything electronic, but especially ham radio, radio and TV broadcasting, and computers. Employed as a television engineer for several Syracuse, NY TV stations.
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s