Duplexer Basics

Cavity filters can be used as narrow BandPass filters, a single tuned circuit with a very high Q (12000+) coupled into and out of by various methods, usually a loop coupling to the, in cavity magnetic field.
They can also be a narrow BandReject filter, usually done with one coupling loop and where the cavity, at resonance, presents a very low impedance. Thus is used to "short out" the signal on the coax line at a Tee junction.
Because our Ham Radio repeater split frequencies are so narrow, particularly on 6m and 2m, we need a different cavity type, the BandPass AND BandRegect type (BP-BR).

There are FIVE main types of BP-BR Cavity filters.

1) The single connector series loop and trimmer to GND type.

2) The two connector, couple a bit of energy between the coupling loops
    with either L or C.

3) The two connector, couple some energy magnetically between loops.

4) The two connector, parallel resonant circuit between connectors.

5) The single connector capacitor divider, with either L or C in the GND side.
    As in the 2" heliax duplexer for 6m.

AND, the aperture coupled Band Pass UHF cavities behind the image Ant_Tee_IMG_3734b.jpg.

AND, the Hybrid Ring, perhaps lost in time as it is for a small frequency range only. Hybrid rings from 1966


Type 1 Series coil and cap, trimmer to GND

Lepzig UHF duplexer.....
Note: No connectors, hardwired and Gold trimmer.

Type 2 Couple a bit of signal between loops, L or C

Alan VK2ZIW's 5Note phase swap of LHS loop, makes for larger L.

Type 3 Magnetic coupling

Alan VK2ZIW's UHF test cavityMagnetic coupling. Pass then Notch, 2x 195mm
Notch then Pass, 2 wires 2mm x 205mm Coil ID 14mm.

Type 4 Parallel resonant circuit metween two connectors

image from Coupling Loop Research, RepeaterBuilderTwo connectors parallel resonant circuit between.
Three adjustments, cavity plunger, rotate probe assy., adjust cap.

Wacom 2m duplexer, 4 canNote the caps between the connectors.

4x 5" cans in the 19" rack nice, I question if it is good enough for our 600KHz spacing?
Can't be Wacom WP639 or WP641, no down the side coaxial stub
Tx noise supp. at Rx freq. 80 dB at 0.6 MHz, 90 dB at 2 MHz
Rx isolation at Tx freq. 80 dB at 0.6 MHz, 90 dB at 2 MHz

Type 5, Capacitive coupling at the Hi Voltage end

See A 1972 article on 2" heliax as resonators for 6m also on UHF the Motorola T1500

This below is an area of discussion NEEDS TO BE ADDED on Repeater Builder.

The Antenna Tee Where the two cavity chains join to feed one antenna.

At the notch frequency, types 1 and 5 (above) present a Very Low impedance thus back at the antenna "T" where we need a High impedance as the energy at this frequency is to go straight past, we can use a quarter wavelength line to achieve this.

Type 4 presents a High impedance at the notch frequency so half wave lines are used to the Antenna "T". Between cavity lines, see Ref (h).

Types 2 and 3 and the Band Pass cavities present a High VSWR but neither High or Low impedance exactly, this is where it gets hard. A Vector Network Analyser is a great tool here but I doubt they had one back in the 1960s. Instead, General Radio Inc. made a constant impedance adjustable line. Adjusts 20cm and they would have used other known lengths of coax to calibrate it.

The LHS of my picture Ant_T_IMG_3734b.jpg has an extra half wave of hard-line to
get the lengths right (as best I could).

Between cavity lines

Type 1 (Series to GND) are easy, 1/4 wave. Others can cause interaction between the cavities such as Type 4.
A Vector Network Analyser should be used at this point. But, can we get away without one with our choice of cavity type. The BIG question.

So, now you know why some of the coax lengths are not quite quarter wave or half
wave. Particularly where the Rx and Tx chains join, the Antenna T.

Some examples of home-made (type 4) coupling loops for 6" RFS or AEA cavities.

IMG_3733b.jpg TNC connectors

Note the coil. I "filsched" this idea from Repeater Builder, "Coupling Loop Research", where they increased the loop inductance with a coil not in the "main" cavity magnetic field. I found this made the pass-to-notch spacing, much easier to adjust. Adjustment by rotating the assy. See also reference (e). But be careful of Ref (h) TX-RX instruction....

Tuning Basics

The cavity set pictured below.
I've made up the loops, set them up for 146.9MHz pass and 146.3 notch, make up all the 1/4 and 1/2 wave patch leads, connect it all up, this chain works, a treat. But, the other set, individually look good but together, the pass is a double hump and 6db down. What do I do?? I tried putting a bandpass cavity between the two BP-BR cavitiies, no effect. I tried a 5db pad from the VNA, no effect.

OK then, how do we fix it? PocketVNA 10db loss double hump, bad

Here we must remember, when we check a single cavity with our two port analyser, we terminate the output connector of the cavity with the input of the analyser. In my case a VSWR of 1.5:1 at 147MHz.
When we connect two cavities in series, we are terminating the first cavity with perhaps a 20:1 VSWR at the notch frequency.
How this effects various cavity coupling loop types is, largely why I'm on this project.

At the end of the day, we want to know what works and is easy.
And, to know WHAT DOESN'T WORK or is HARD. This is MOST important.

The secrets are here actual Vari Notch tuning by Jeff WN3A.

2018, my ex. 70cm 3/4 wave cavity set, ready for 2m conversion.

PocketVNA measurements

Magnetic coupling, Type 3, RHS picture above.

Magnetic coupling, Type 3, Lower picture above.
Both can be done, pass then notch, notch then pass.

Magnetic coupling, Type 3, Lower picture above.
This is with about 4" of connectors added inline from the calibration point.

More detail on www.unixservice.com.au/hamradio/repeaters/duplexers/2Mnotes.html


a) A 1972 article on 2" heliax as resonators for 6m
b) A 1965 IEEE article by Phelps Dodge
c) A general article by EMR Corp
d) A 2010 article by Sinclair
e) Where the loop cannot be rotated...VE2AZX
f) Motorola T1500 UHF series, with internal pictures.
g) TX-RX Coupling Loop Project on the Repeater Builder website but no followup on a full duplexer build.
h) TX-RX-instruction-manual-vari-notch-duplexers-with-6-inch-cavities.pdf 1/4 wave inter-connect cables, not quite correct.
i) andrew-about-rf-communications.pdf Section 4, a general article.
j) Hybrid rings from 1966
Vari-Notch patent
Chapter on the Hybrid Ring by W2EUP
Duplexers, Theory, and More by I1WQR
My page on the Hybrid Ring
GE M2 comparison Tx PLL to Tx multiplier systems (VHF)