For the first time since I studied transmission line theory (1996-1998) and scribbled my way through Smith diagram application exercises, and for the first time in my ham career I made a stub. And it sure feels good ticking one off the ‘things to do in this life’ list! Especially since they work and solve my problem.

The problem mentioned is simple: two rigs, two amps and only one set of band pass filters. I got away with this as long as radio 2 was barefoot. But now I need a second set of BPF on the other rig as well. The two combinations that result in high RF levels on the RX are TX40-RX80 and TX20-RX40. So I need a stub on the 80m antenna that blocks 7MHz signals, and one on the 40m antenna that blocks 14MHz. The latter is simple since 40m is a monoband antenna. For the 80/160 GP I’ll need to add some switching later on. But maybe the stubs will become obsolete as soon as the BPF is in place. I’ll need to wait for that while it’s engineer/seller is playing pile up in Micronesia. Yes, I ordered the 5B4AGN band pass filters. The price was the same as a Dunestar 600 set would cost me (add S/H, convert to Euro and add 21% VAT). I have one of those and I’m quite pleased as it has done well for over six years already. The OM-Power set, while not doubt very good, is a bit too expensive. Mike SJ2W recommended the 5B4AGN filters and I trust his technical advices almost blindly. The filter gets shipped as soon as Bob is back from V63.

In the mean time, and maybe needed even with the filters: stubs to the rescue. K1TTT has some info. And W2VJN’s book “Managing Interstation Interference “ seems to be the reference for this kind of problems. Now I know I have that book here somewhere. I bought it in 2006 when I was building the SO2R station. But I didn’t find it at first. Maybe some local ham friend took it home to read? Although I can’t think of anyone around here doing SO2R. After some digging I found it fallen behind a shelf.

The book just confirmed the drawing at K1TTT’s site and is a textbook example of transmission line technology. I could see myself again in the school’s RF lab drawing Q lines and SWR circles on A3 sized Smith chart diagrams. From short to open is half a circle is a quarter wave, from short to short is the full circle is a half wavelength. A quarter wave on 80 to short 7MHz and a quarter wave on 40 to short 14MHz. I have some loose ends of coax of various types so I looked up the velocity factor for these and off to the shop.

Hmm, it seems my stock of coax is not so big as anticipated. Either short lengths for jumpers or really long runs which are a shame to cut. I found a length of RG-213 which is most interesting since it has the lowest velocity factor making the needed length of cable much smaller. I recognized the crappy soldered N-plug: my very first piece of coax when I got my VHF / ON1 license in 1999. I learned many things in school but soldering plugs to coax I learned ‘on the job’. Post 1999. I made a short on the other end and connected the AEA analyser. BTW my analyser really gets worn out now. There’s the known problem (link) and now there’s something wrong with the DC receptacle too. I used the analyser to find the lowest impedance around 7000-7100 kHz. Turns out the run was a tad too short. Not wanting to cut a really long cable, I added a jumper with a barrel connector. That shifted the frequencies of open/short down. So in the desired 40m range I had 2.2 ohms resistive. And that made the dummy load look like 49Ω purely resistive on the lower end of 80m. No imaginary friends here!

I wish I could show you nice pictures of the stub’s frequency response made with Rohde & Schwarz network or spectrum analysers with tracking generators, but unlike SJ2W I have no access to such nice toys. Not anymore. Pour la petit histoire: in my pre-ham life I have worked for R&S Belgium in their service shop, repairing and calibrating RF T&M goodies. Little did I know then! Anyway while a bit cumbersome to handle with the defects and intermittent keyboard, the AEA did its magic and helped me make another stub to pass 7 MHz and short 14 MHz. Since I ran out of 50Ω coax, I took a length of 75Ω “made in Belgium” cable TV distribution coax. Velocity factor 0.88. The initial length turned out to be seventy two centimeters too long. I calculated this by making the difference between quarter wave length for the desired frequency minus actual measured quarter wave length. Of course quarter wave derived from [300/(f/2)] where f was the frequency where the shorted stub presented a short to the analyser. So I cut away 68 cm and cut another 2 cm away to make the short. Repeat measuring procedure et voilà: spot on!

I was all racked, stacked and wired for serious SO2R in the UBA DX CW contest. These stubs have effectively solved the problem while not introducing other problems – at least not that I have noticed so far. I hope the stubs won’t be needed anymore when the second set of band pass filters is in place. Or I will need to provide yet another interface to switch the stub in / out on 80. And the to do list is as long enough as it is already…

One reply on “Stubs”

Good job Franki! I bet you will love the 5B4AGN filter, you will think it is pure engineering porn when opening it up!

And I am also honored to be mentioned even twice in your post 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.