Tuning The WB5WPA 6M Duplexer

     General - The WB5WPA Six Meter Duplexer stubs are tuned individually then
     connected together with 1/4 electrical wavelengths (36") of RG-58 coaxial
     cable. If the length of the hardline has been calculated to be 9.8% short
     of the electrical 1/4 wavelength (assuming 88% vel factor) at the frequency
     the notch is supposed to attenuate, tuning then consists of a coarse adjust
     on the 'coaxial' or gimmick capacitor followed by a fine freq adjustment
     with the trim tab and then adjusting the shunt reactance for minimum
     attenuation at the pass frequency.

     If more attenuation is required, another section can be simply added to the
     transmit leg or receive leg with another piece of 1/4 wave RG-58 as
     required - but first determine if xmtr noise (at rcv freq) or transmit
     signal (at xmit freq) is causing desense and needs to be knocked down more.

     To find the notch, some sort of signal source and detector are needed.
     Examples of detectors are multimode rigs with S-meters and diode detectors
     (a diode across a 50 Ohm resistor into a DVM). Signal sources include
     scanners, Service monitors and signal generators. The use of a scanner as a
     signal source *may* require some explanation. Most scanner use a 10.7
     MHz (or 10.8 MHz, as some of the Bearcats did) first IF (Intermediate
     Frequency). Program a frequency into the scanner 10.7 (or 10.8 - as
     applicable) MHz lower than the frequency of the notch. This means 41.51
     for a notch of 52.21 (52.21-10.7=41.51). This *should* cause a moderate
     signal to 'appear' at the antenna jack of the scanner using the multi-mode
     rig as a 'sensor'. This way is the hardest - you'll have to sweep the band
     to find where the notch is in order to decide if the notch is high or low
     and the series 'gimmick' cap needs to be made shorter (raise freq) or
     longer (lower freq).

     Tuning - Each of the band reject "stubs" is tuned separately - thereby
     eliminating the necessity to measure attenuation values in excess of 70 dB.
     The attenuation of an individual each stub at the 'peak' or nose of the
     attenuation notch is around 19 dB - don't worry, connected together with
     1/4 lamda electrical lines they will show over 70 dB!

     Tuning the Series Capacitor - The length of the RG-8/RG-213/214 'gimmick
     cap' determines the 'rough tuning' of an individual stub. The series
     'gimmick' is tuned for a frequency 30 to 40 kHZ *higher* than the target
     notch frequency - the Notch Freq Trim Tab makes it easy to tune the notch
     *exactly* on freq. Hook the equipment to the 'GOZINTA' and 'COMSOTA' and
     find the notch. If it needs to be moved up, desolder the center conductor
     of the RG-8/213/214 and work the center conductor and dielectric out of
     the braid a little bit - resolder and check freq again. The 'heat' from
     soldering/de-soldering will change the notch freq some - let the assembly
     cool some before making final notch freq adjustments with the Notch Freq
     Trim Tab.

     Shunt reactances - The capacitor or coil are usually added after tuning.
     They *do not* affect notch freq so it really doesn't matter (if they do,
     the thing is built wrong). I went through many iterations of this thing
     perfecting it that it is best to tune the notch first and add the other
     later. If adjustable inductors/capacitors are used, adjust them for minimum
     insertion loss .5 MHz on the *pass* frequency. If using lab equipment less
     than .4 dB insertion loss of each 'stub' should be had after adjustment.

     Notch Frequency Trim Tab - changes the frequency about 50 KHz. This is used
     to 'fine tune' the notch's frequency. Pushing this little square of metal
     closer to the to the Heliax *center* conductor moves the notch freq
     *lower*.