The AM Forum
June 18, 2024, 05:01:15 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
Author Topic: bandswitching inductors topology question  (Read 1751 times)
0 Members and 1 Guest are viewing this topic.
Offline Offline

Posts: 427

« on: December 31, 2021, 11:55:58 AM »

In bandswitched devices, selecting the tap on an inductor is common. Most often, one sees the un-needed part of the inductor shorted out. See my first picture, the top part. The bottom part shows the alternative.

The basic question: Why short it out?  Does it not to some degree make a shorted winding that sucks energy up? Would leaving it open on the end create some problem?

My actual application is the second picture, the Pi network of the RCA broadcast transmitter I am converting.  Due to available parts, I have a big coil of about 50uH, that can have taps on it. Secondly, a big roller inductor with a maximum value of 15uH. As noted on the pic, I need more than 15uH for 80m and 160m, so the two are in series. My intent was to set the taps so that the 40m setting of the roller would not change when the 80m or 160m segments of the fixed inductor are added, never using all of the roller's range. Why do I need to short the unused part of either inductor? The roller came to me set up to short the unused part. The fixed coil is a matter of how I strap the tap.  In my actual installation, it will be cleaner and easier to leave the unused end of the coil floating. There's less strap hanging around int eh way, less to possibly touch something, etc.  So why is the shorting way so common, what does it gain us? What will I lose leaving it floating?


* IMG_6209.JPG (520.97 KB, 2016x1512 - viewed 206 times.)

* IMG_6211.JPG (577.24 KB, 2016x1512 - viewed 197 times.)

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...

Offline Offline

Posts: 671

« Reply #1 on: December 31, 2021, 02:33:19 PM »


I think the best answer is go measure. Set up the pi match as desired, obtain your 20 dB return loss or better and measure the insertion loss. I suspect, that the open tap arrangement might not be the best. Primarily due to parasitic C around the inductor forming a resonate tuned circuit. If it is series form, ouch! By the way, you can get the insertion loss from the return loss. That is a more convenient method then actually setting up for an IL measure. At this same time, you can evaluate your completed pi match loaded Q and BW. Just to be sure that is correct. Lossy arrangements will foul that up!

Roller inductors are nice... but unless they are premium units, are lossy.

In the end, its all about UNLOADED Q. You want the unloaded/loaded Q ratio to be as high as possible.

Offline Offline

Posts: 1797

CW is just a narrower version of AM

« Reply #2 on: December 31, 2021, 04:36:18 PM »

This goes all the way back to crystal sets and loose couplers. The solution for them was complete disconnection of unused sections. This can be done with an exotic rotary switch design. This was to prevent "End Effects" in the receiver tuned circuits. In transmitters, coil sections shorted or left open sections can be Q robbing at best, and parasitic power killers and Zorch, at worst.

But in a multiband system, just using a tapped switch is simply too handy and it basically works good enough. Grandpa had it right with those big transmitter plug in coil assemblies - no compromise.

These are the good old days of AM
Offline Offline

Posts: 2658

Making AM GREAT Again!

« Reply #3 on: January 01, 2022, 07:56:23 AM »

You short out the unused (lower frequency) taps so they don't act like an auto transformer.

Imagine the Vpeak at the end of the 80 or 160 meter coil when operating on 10 meters

Pages: [1]   Go Up
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.051 seconds with 18 queries.