Notes on 2m Duplexer Cavities for HAM bands

2m 600KHz or 0.4% spacing,

This is where the BP-BR cavities come to the fore.

Compared to: 70cm with 5 to 10MHz or 1-2% freq spacing, and 6m with 1MHz 2% spacing.

Thus there are three methods used:
1) the series loop

loops3.jpg from

2) The couple some energy between loops method:

my photo IMG_3116.JPG

my photo IMG_3117.JPG

Note, none of the commercial units I've seen use the couple a bit of energy from one connector to the other approach.
Because, 3MHz is a much more common frequency spacing on the 150 to 180MHz band.

3) The least common method, the parallel resonant circuit between connectors


HIGH loops21.jpg from


LOW loops13.jpg from

Note the extra inductance out of the main magnetic field.


From this web page by Kevin K. Custer W3KKC, Jeff WN3A used a 6 inch cavity for his tests:
(S11) - (S22) return loss and notch depth.

Factory Low-Pass: 22/25 dB return loss, 38.3 dB notch depth
W3KKC Low-Pass: 21/26 dB return loss, 39.2 dB notch depth

Factory High-Pass: 25/22 dB return loss, 39.9 dB notch depth
W3KKC High-Pass: 26/22 dB return loss, 40.7 dB notch depth
So we hams can build them as good as commercial too. From their measurements, the performance is near 10db better than the serial loop and "N" Tee. (resistance in the T pin and socket I bett)

=========== Advantages / Disadvantages =============

The Series Loop

Here we get two notches (about 2MHz apart) with a pass between them. So we adjust the pass to be closer (600KHz) to a notch, either higher or lower, depending if this cavity is in the Rx or the Tx chain. See IMG_3250.JPG below.
Another advantage At the notch frequency, the impedance at the cavity is VERY low so we can use a 1/4 wave (or 3/4) line to it and present a HIGH impedance at the antenna "T". Makes our life easy with separating the Tx and Rx signals. No need for a circulator.

Disadvantage No "out of band" rejection.

The two loops and coupling method

Here we have only one pass and one notch and some "out of band" rejection.
Only one resonant circuit, the cavity.
The pass/notch sense can be swapped by using capacitance instead of inductance as seen in my above picture.
There is much innovation here in ham literature. But, very little in-depth discussion. Eg. the wire size used in the loops, does it matter? Yes, thicker wire or strip has lower loss but but but, lower self (or stray) inductance. So don't just jump in and thicken up the wire size when refurbishing.
Note also, one can swap the phase of the loops as below.
Note the much higher inductance used compared to IMG_3116.JPG above.
Conventional phase: Both loop active leads come from near the inner. Here LHS one is not.

IMG_3222a.JPG from me

IMG_3224a.JPG from me

Even more innovation, Magnetic coupling

What about placing the two loops inside the cavity close to one-another such that there is significant magnetic coupling?
Yes this works. At UHF in 4" square cans, I've got 5MHz spacing. Note the slots Motorola provided.
But to swap the pass/notch to notch/pass is very problematic due to the magnetic field not being uniform either up and down or in and out.
my photo IMG_3242.JPG>
Disadvantages: As a first can after the antenna "T", we have problems because the impedance at the can is not an easy very LOW impedance so we have to measure the impedance and calculate the correct coax length that will present a HIGH impedance at the antenna "T". I suspect our predecessors had zillions of coax male-to-mail and female-to-female joiners and/or a length adjustable coax line.

The parallel resonant circuit

Here again we have two notches and a pass...
IMG_3251.JPG from me
IMG_3250.JPG from me
35db notch is not bad but the RL could be better. First attempt.
This 6" cavity has no main tuning yet. Was 70MHz. I'll cut it to 148.2MHz and put an insulated sleeve in the inner, protruding to bring the freq. down.
A fiberglass rod as pusher.
The pictures loops21.jpg and loops13.jpg above show best practice.
Again see:
Advantage: Hi Z at notch. Therefore use 1/2 wave line to antenna "T".
Disadvantage: Poor "out of band" rejection.

================== My 2nd try and it works! ======================

RIMG0003.JPG today 2/6/17

IMG_3713.JPG today 5/6/17
1MHz markers, 35db notch, not bad. Lower trace Return Loss

================== Hints ======================
1/4 wave lines can be used to swap the pass and notch frequencies, the U in hard-line at the cavity.


Avoid "T" connectors as the internals can and do give problems.

Solder is a poor conductor. Avoid tinned wire or strip, bare copper is better, painted or enameled to keep out corrosion.

================== My 4" cavity set ======================

1/4 wave lines to 1st cans, series loop type. Next two cans, two loop types as per IMG_3222a.JPG and IMG_3224a.JPG.
Antenna "T" hidden, a plug on the LHS ends of the two teflon coaxes plugged into the RHS brass boxes.

================= Is it good enough, well NO ==============

4" cans are barely adequate. Go 6" or larger.