[Amps] RF in the Audio

[Jim Brown]

On 9/25/2011 8:47 PM, Jim Thomson wrote:
>
>
> ## The 3000-6000 ohm recommendation is a good one.  Just beware that to pull
> that off, ur gonna  need a lot of type 31 material.

Not necessarily.  See my measured data.

> Where the problem starts is
> when folks have a 1500 ohm Z  with their type 31 material..then wonder why it blows
> up in their face.

YES!

>    Sure  type 31 will work.. but it's all resistive.  That's  wonderful..but
> it's still  I squared x R.   IE: 2 x chokes, each is 1000 ohms..and one is type 43  and the
> other is type 31..... the type 31  will go kablamo every time.   To get the type 31 to work
> right,  it's Z  HAS to be sky high...so the  current ends up really low....low enough  so
> the type 31 material doesn't heat up.

That's another reason for 5,000 ohms, and why more is better.

> While K9YC's site has a lot of very good information there is also a bit of
> disagreement with what others have published and I dont see that changing
> much in my lifetime. He hasnt bothered to reply to my request for a test of
> a balun feeding OWL fed dipole covering 160-10M and at 1500W.

As I said before, I don't have an application for OWL. All my antennas 
(about 15 at least count) are fed by coax (including some hard line). If 
you care, try the tests yourself.  When setting up the tests, consider 
how much common mode voltage that the transmission line can place across 
the choke by virtue of its length, AND how much common mode voltage the 
ANTENNA places across the choke by virtue of its imbalance. I've done a 
bit of that analysis in NEC. It's on my website as part of a Power Point 
presentation on coax chokes.

> Lets try this
> with a 4:1 and 9:1 as those are the common ones in use as well as
> deliberately varying feed line lengths to present worse case scenarios on
> different bands.

The case of impedance transformation using an array of common mode 
chokes places a LOT of voltage across each choke. At relatively high 
power level, 5,000 ohms is definitely no where near enough.

>
> ### The OWL  length will have to be tweaked precisely to cover a bunch of bands.
> A buddy  smoked and blew up every  4:1 balun made..including palstar, and also the
> amidon 4:1 unit.  [ both were rated for 5/10 kw].   The only commercially made balun
> that didn't blow up was the 4:1  made by AS.[ 5/10/20 kw]    Previous to all of this, he had tried various
> chokes on the INPUT side of the hb T  type tuner.   They all overheated and blew up, including
> the 5 kw palstar, made with bifilar wire on a stack of torroid's.   The torroids cracked.

This is consistent with what I noted above. W8JI has an app note about 
this on his website.

> One test I rarely see mentioned is to test your coax first.
> Leave in place and terminate the far end in 50 or 75 Ohms and then tune the
> bands recording any crud frequencies. Then add a bead balun and reterminate.
> Record any differences in signal levels. If you have a quiet receiver then
> any pickup with the antenna is likely in "antenna mode" as I like to call
> it. Many get confused with all the technical terms used. Any additional crud
> picked up in "interference mode" will be small and easily eliminated at the
> shack end with another bunch of beads and hopefully a decent RF ground.
>
> ###  Ok, this is a new one on me.   This is a damn good idea.   Are you saying
> to terminate the coax at top of tower with a 50 ohm plug ?  I would have thought
> that would result in zero  local noise pickup. Any small, shielded dummy load should work.

Don't forget about the transfer impedance of the shield, which is a 
measure of the perfection of the uniformity of the shield. In short, the 
higher the transfer impedance, the more differential voltage that 
appears on the coax as a result of current on the shield. The lower 
limit of the transfer impedance is the shield resistance at the 
frequency of interest.

BUT -- the role of common mode coupling is different -- it is a coupling 
path between the shield and the antenna itself.

73, Jim Brown K9YC