On 24 jun, 17:42, Owen Duffy wrote:
On Saturday, 25 June 2011 01:01:00 UTC+10, Michael *wrote:
I've been reading several articles on how to build a Guanella 4:1
current balun. *Some articles show a single toroidal core being used
for all windings, and some articles show two separate toroidal cores
being used. *Which is better?

The single core does not work properly, seehttp://vk1od.net/balun/gsc/.

Owen

Hello Michael,

Owen's reference explains everything in detail.

If you want a very short answer:
The two-core balun has common mode suppression characteristics. That
means when your balanced load isn't balanced in real world, the balun
will reduce common mode current in both coaxial line as well as
balanced line.

The single-core balun is electrically more like an autotransformer
with the coaxial ground on the center-tap. This arrangement does not
provide any common mode suppression. So only in case of a perfect
balanced load, everything will be fine (with the single-core balun).

Now the longer part
If you still want to go for the single-core solution, you may add a
separate common mode choke. For common mode behaviour, the two cores
in the two-core balun are in parallel. So the effective common mode
impedance will be half that of the CM impedance of a single core.

Regarding the number of turns for the balun; when you need 8 turns on
each core for the two-core solution, you can use 4 turns for each
branch in case of a single core solution. So if you make the single
core balun with coaxial cable, you will have 8 turns of coaxial cable
around the core.

Slightly off-topic.
Be aware of the fact the actual input or output voltage determines the
heat loss in the ferrite cores. So if the VSWR at the balun is very
bad, and the balun is in the voltage maximum, core loss may increase
significantly.

With kind regards,


Wim
PA3DJS
www.tetech.nl