860 mhz quarter wave stub on handheld doing ok!

[nanZor]

I just built an 866 mhz quarter wave stub for the handheld and it is doing great. Although these stubs are used for notching, I didn't care about the notch and just wanted to apply some overall attenuation on 800 mhz everwhere before I reached for the attenuation button. I read the other threads about stubs and wanted to try it out.

1) Attached a bnc-tee to the handheld with a male-male bnc adapter, RS 800mhz duck to the other side, and then the stub to the center of the tee.

2) This stub is nothing more than 2.5 inches long **as measured from the middle of the tee**. I just cut up an RG-58 jumper I had laying around, and what's cool is that the cut is just past the point of where the injection molding of the jumper ends. Looks like I'm hanging a race-antenna off it. I've read the threads about using low-loss coax for a good notch, but in this case, I don't mind the higher loss rg-58 for general attenuation instead of a sharp notch.

Now the 164 itself is not attenuated, and things are peachy. What's really cool is that I now have somewhere to go if I wanted even more attenuation with the built-in 20db attenuator. The other bands don't even know the stub is there.

I'm still playing with it and I'm sure I've really put a notch on something somewhere, but I don't have any test equipment to see it. It's going to take a bit of listening to my other 900mhz systems and do comparisons.

Either way, the stub is so small that it doesn't make the 164 top heavy, and just hangs rigidly over the volume/squelch controls without extending too far beyond the chassis. Neat!

Update: I couldn't wait very long so I sat on my 866 mhz control channel and trimmed it back from 2.5 inches about 1/6th inch at a time. BAM! Got the stub right on freq and compared to the built-in attenuator, they are about the same in effectiveness! So on this freq, we're talking about 40db attenuation at my disposal.

That is the FIRST time I ever had an open-stub work for me.

 

[nanZor]

Update - very careful pruning to 2-3/16 inches for the stub really does a number on 800mhz for me with RG-58.

You have to take velocity factor into account when measuring, so if I were to make a general recommendation without knowing the VF for the cable, I'd start out at about 3 inches from the middle of the tee, and progressively cut back in 1/16th increments. (solid dielectric with around .66vf and foam dielectric with about .80 approximately.)

Use a sharp pair of cutters to make sure that when you cut, the shield and center conductor aren't shorted together.

If the system is so strong that you can't really determine if there is any attenuation taking place, enable the attenuator and then do the cuts. Of course turn off the attenuator if you later find you don't need it and the stub suffices. At least now you have an option for more.

 

[nanZor]

Note: The tee adapter I used has all female ports like the Radio Shack #278-111.

If you try to use the other tee adapter they have which has a single male port #278-112, by the time you put a barrel adapter on it, the stub will be too long for 800mhz - although you could use it with a stub for lower bands. Just be sure to make all the measurements from the center of the tee!

 

[nanZor]

Why 866 mhz stub didn't physically measure up!

Aside from this stub, I've been building other 1/4 wave open stubs and half-wave shorted stubs on lower bands with the same coax and have been pretty spot-on with the following formula to get the length from the center of the tee in inches:

246 * VF / F mHz * 12

Or for my 866 mhz stub with solid-dielectric coax:

246 * .66 / 866 * 12

That works out to 2.249 inches.

So why is it that when I cut it to this length it attenuated my 900 mhz systems, and eventually I had to make the stub just a bit longer to about 2.5 inches to bring it down to 866 mhz?

Then it hit me - at 866 mhz, much of the stub is comprised of the tee itself and mating adapter - which has a different velocity factor than the coax. So I have a little mix in the works requiring a cut and try method here on 866.