Tuning The WB5WPA 6M Duplexer ( modified by KF6YB ) General - The WB5WPA Six Meter Duplexer stubs are tuned individually then connected together with 1/4 electrical wavelengths 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 notch depth 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. Good reading on the theory and tuning of a duplexer can be found at: http://www.seits.org/duplexer/duplexer.htm MUST READ! => http://www.repeater-builder.com/rbtip/notchduptuning.html ----------------------------------------------------------------------------- Tuning -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 Service monitors and signal generators. I strongly suggest the use of either a relyable network analyzer or a signal generator and spectrum analyzer with a tracking generator for these procedures. There is so much work involved with building them that the measurements should be solid and repeatable or tuning will be a nightmare. Use the 1/4 wavelength interconnecting RG-58 pieces between the source and the analyzer to approximate the best match. At this point, find the notch using your test set. If the notch freq is higher than the intended frequency by more than 500Khz, it may be better to start again with a new longer piece of Heliax. You can see that it may be smart build the first stub an inch longer than needed just to be safe. Once you get a completed working stub you can build the remaining stubs the same length as the good one. If your notch freq is too low, then cut off the shorted end bit by bit re-shorting it and finding the notch. yes, this is a tedious task but it wont take long. Stop a few Khz below the intended notch frequency and fine tune with the gimick capacitor. 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! The length of the RG-8/RG-213/214 'gimmick cap' also affects the tuning of an individual stub. 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 or tube a little bit - resolder and check freq again. I have also had great luck by pushing the braid down into the center conductor in small increments using a small poker. This will raise the notch frequency and it is easy to do the final tuning this way. 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 .8 dB insertion loss of each 'stub' should be had after adjustment. Notch Frequency Trim Tab - changes the frequency about 50 KHz lower. 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*. Be aware that the distance between the thruline copper strip and the center of the Helaix also affects the notch frequency, be sure to try to keep the distance constant to avoid confusing measurements.