[Amps] minimum length of 1:1 transmission line balun ?

[Larry Benko]

Carel,

This is a very easy problem to solve with a minimal amount of test 
equipment.  Your 1/10 wavelength (VF = 1 ion outside of coax) is a 
pretty reasonable number but this length can be reduced considerably by 
winding the coax in a coil (1 or 2 turns @ 2m) or running it through a 
low loss (u=40 or 125) ferrite.  You are looking for choking impedance 
in the 250+ ohm range.

To do a test, terminate the far end of the 50 ohm coax in a 50 ohm SMD 
resistor very carefully.  Drive the coax with a low power (do not blow 
up the resistor) and measure the line SWR.  It should be extremely low 
if you did a good job with the resistor.  Now touch the shield  
(resistor end) of the coax to ground.  The SWR Should not change.  Now 
touch the center of the coax to ground.  The SWR will rise and if the 
choking impedance of the line to common mode current is high enough the 
SWR will rise very little.  Try coiling the coax an see the SWR drop.  
You determine the maximum SWR but something like 1.3:1 is fine.  In your 
application you will be operating as if the center of the 50 ohm 
resistor is grounded which requires less choking impedance than if the 
center is grounded.  If you use a network analyzer to make the tests 
just sweep the frequency and you can see what length is needed for a 
given frequency.  I have made these tests many times and learned a lot 
making the measurements.  If you decide to use some type of ferrite you 
later will need to determine that the ferrite does not heat up at full 
power.

73,
Larry, W0QE



On 10/22/2013 9:49 AM, Carel, pc5m wrote:
> Hi, busy constructing  a QRO 2 mtrs power amp with balanced FET's.  It will
> a.o. be utilizing a rudimentary (coax) 1:1 balun at in- and output  to
> transform 50Ohm unbalanced to 2 times 25 Ohm balanced (single run of coax,
> outer braid connected only to ground at 50Ohm end).  I tried to figure out
> what the length of this piece of coax should be ? Doing some RF simulations
> with Agilent ADS seems to indicate that this length can be very short, one
> tenth of wavelength, although I got into the issue of determine the even-
> and odd-mode impedance of a coax cable ..would odd-mode impedance of coax
> equal to his characteristic impedance and even-mode impedance just be the
> reactance of a the coax outer braid ?
>
>   
>
> Anyone can point me in the right direction for the design criteria regarding
> the length of  a (coax) 1:1 balun ?  I have found some app notes / designs
> but a bit reluctant to blindly copy something I do not fully understand ;-).
> Of course the shorter those baluns can be, the happier I would be.
>
>   
>
> 73's Carel, pc5m
>
>   
>
>   
>
>   
>
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>