Tim wrote: > Hi Kevin, > > Yes, the original used RG-58.. couldn't see using it. Figured > it would give me desense! > > As far as the amount of notch, I was surprised that the total > was more than each one added,
Don't be - it's common for the apparent notch to be deeper than the measured addition of individual cavities - added together. Don't be fooled, the depth was there the whole time (no black magic here), it just wasn't all being realized in the test set-up. Proper interconnect cable lengths can do a bunch for notch depths. > but with 50 watts > coming out of the antenna port, I see -55dBm at the RX > port. - kinda made me think the original measurements > were ok. > What you see coming out the RX port depends on several factors: How good the match of the attached equipment is - at the frequency being tested, and the phase angle of the connecting cable. Other factors play a part too, but these two play the bigger part. > I cut them to 35" each. (visible coax showing between each > end of the PL-259). Not exactly to length. Might affect > performance a bit. I wouldn't have chosen UHF connectors for building the duplexer. IMHO, their construction leaves a lot to be desired when trying to optimize a duplexer for maximum performance, unless you take their inherent flaws into consideration when using them. Wacom built a good duplexer which uses UHF connectors, but they actually used those properties to their advantage. 35 inches, plus the amount inside the connector results in approximately 37-1/2 inches of interconnecting cable. From my calculations, I have extrapolated your optimum notch depth to appear on 54.44 MHz. I believe, with RG-142 coax, that a length longer than 38 inches will be needed, depending on your exact frequency. From your photo, I believe some of your problem to be the UHF connector and how you have it mounted. A solid ground all around the connector is important to achieve a low impedance (low inductive reactance) connection. I have always preferred the "Utah Cap" approach for building the Heliax duplexer. It allows you to realize most/all of the potential of the stub. The thicker the copper (within reason) the better. Also - it looks like you might be using only one connector per stub. If this is true, and you are using a UHF tee connector with the stub - it won't work as expected. The electrical length of the tee will throw the whole thing off - this might be your biggest problem with your design/construction. Kevin
