[TowerTalk] Carolina Winsome and vertical radiator question

[TexasRF--- via TowerTalk]

Out of curiosity, what power level was used for the simulation?
 
Gerald, K5GW
 
 
 
In a message dated 1/26/2015 7:20:15 A.M. Pacific Standard Time,  
steve at karinya.net writes:

Kelly,

Maybe this will help understand what is  happening:

Take a look at the SPICE schematic here - It models the  feedpoint of a 
low(ish) 80m OCFD fed one third the way from one end  through a 4:1 
voltage  balun:
http://www.karinya.net/g3txq/temp/ocfd/80m_ocfd_spice.png

Points  A and B are the feedpoint connections to the Long dipole leg and 
the Short  dipole leg, respectively.

If the dipole had been fed at the centre, the  two leg impedances to 
ground would have been resistive and equal to around  25 Ohms. But by 
shifting the feedpoint to one side of centre, the Long  side is now 
significantly longer than a quarter-wave - its radiation  resistance has 
increased to 100 ohms and its individual impedance to  ground has become 
highly inductive [100+j319 Ohms]; whereas the Short side  is now 
significantly shorter than a quarter-wave, its radiation resistance  has 
dropped to 12 Ohms and its impedance to ground has become highly  
capacitive [12-j319 Ohms].

As far as a differential signal applied  across the A-B feedpoint is 
concerned, just as we would expect the  impedance appears to be a 
resonant 112 Ohms because the reactances of the  two legs cancel. But 
those high reactances are key to understanding the  properties of an OCFD!

The schematic shows a 100vpk source being  applied differentially through 
a 4:1 voltage balun to the dipole  feedpoint. A differential voltage of 
200vpk appears across A-B, and a  current of 1.785Apk flows in the dipole 
legs. Nothing new  there!

But now look at the effect the 1.785A  has flowing through  the 
individual leg impedances to ground: it causes the feedpoint to  "float" 
to a very high voltage with respect to ground; Point A goes to  
597vpk/72.6degrees and Point B goes to 570vpk/92.2degrees. Point C - the  
centre-tap of the balun where the braid is connected - floats to  
575vpk/82.2degrees.

So - applying just 100v across the input of the  balun forces the braid 
balun connection to float up to 575v above ground  !!! The explanation is 
*not* that the feedpoint offset has caused the  Short and Long leg 
impedances to be very different from one another;  rather, it's that the 
individual leg impedances have become highly  reactive.

Resistor R3 has been included to represent the impedance  looking back 
along the outside surface of the braid to Ground. It has been  set to a 
very high value so that the fundamental operation at the  feedpoint can 
be demonstrated without being affected by a large current.  But you can 
see that setting that braid path impedance to something  realistic (a few 
Ohms to a few hundred Ohms, and complex) will likely  result in very 
significant current flowing because of the high 575v at the  balun.

You can swap the balun connections around to make it 4:1 UnUn,  but not 
much changes - you still get very similar voltages at the braid  connection.

It's actually a bit misleading to say that a voltage balun  or an unun 
"forces" the CM braid current to flow; it doesn't  - the  driver for 
braid current is the high voltage generated because of the  dipole leg 
reactances; the voltage balun or unun simply "allow" (fail to  impede) 
the CM current. Only a true current balun with high CM impedance  can 
substantially reduce the braid current.

Hope that  helps.

Steve G3TXQ




On 25/01/2015 15:50, Kelly Taylor  wrote:
>
> So, the question is: while a 4:1 balun is the correct  choice for an OCFD,
> would it necessarily replicate the matching unit  in a CW? If it's 
designed
> to prevent CM current on the coax, maybe  not.
>
> I don't know the answer, which is why I'm asking.
>  73, kelly
>  ve4xt
>
>
>

_______________________________________________



_______________________________________________
TowerTalk  mailing  list
TowerTalk at contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk