Baluns - To twist or not to twist the trifilar winding

[Richard Ferryman]

I am in the process of making a 1:1 balun for QRP use and this has made me
think! I have seen many balun designs and construction articles. A fair
proportion of theses advocate twisting the wires before winding on a toroid
core. I have never seen a clear explanation as to why they are twisted.
Does it improve the balance? Does it affect the coupling? Is it better to
twist or not to twist - that is the question.
Dick G4BBH

[Roy Lewallen]

You don't want the differential-mode field (primarily between the
conductors) to couple to the core. Twisting the conductors together
helps prevent this, so that's what I recommend. If you don't twist them,
the two conductors of the transmission line should lie together as
closely as possible.

The third conductor (you mention trifilar) makes the balun a "voltage"
balun, which degrades its performance in antenna transmission line
applications. This is explained in the article at
http://eznec.com/Amateur/Articles/Baluns.pdf.

Roy Lewallen, W7EL

[Richard Clark]

On Sun, 11 Jan 2009 19:31:33 -0000, "Richard Ferryman"
<richard...@btinternet.com> wrote:

> I have never seen a clear explanation as to why they are twisted.
>Does it improve the balance? Does it affect the coupling? Is it better to
>twist or not to twist - that is the question.
>Dick G4BBH

Hi Dick,

Twisting is to bring the two wires into very close proximity. This
reduces the characteristic Z of the transmission line they form. The
reason for wanting to have a low characteristic Z is related to the
impedance transformation (this Z being the geometric mean of the
source Z and load Z). This is generally an issue with one of those Zs
being very low and then you may want to use flat (almost foil) wires
(all wires being insulated, of course).

73's
Richard Clark, KB7QHC

[Bob Bob]

To tell you the truth Dick this is the first time I even thought of this...

My first reaction would be to reduce radiation TX and RX like one does
with CAT5 ethernet cable (aka twisted pair). In a transformer though?

You have asked the right group though, there have been heated balun
discussions here before. One of the contributors wrote a PDF article on
it that may be useful... Sorry I dont have the URL on hand..

I'll wait for an answer....

Cheers Bob VK2YQA

[Ian Jackson]

In message <x5idnU8BzrqOBffU...@posted.easystreetonline>,
Roy Lewallen <w7...@eznec.com> writes

>Richard Ferryman wrote:
>> I am in the process of making a 1:1 balun for QRP use and this has
>>made me think! I have seen many balun designs and construction
>>articles. A fair proportion of theses advocate twisting the wires
>>before winding on a toroid core. I have never seen a clear
>>explanation as to why they are twisted. Does it improve the balance?

>>affect the coupling? Is it better to twist or not to twist - that is
>>the question.
>> Dick G4BBH
>
>You don't want the differential-mode field (primarily between the
>conductors) to couple to the core. Twisting the conductors together
>helps prevent this, so that's what I recommend. If you don't twist
>them, the two conductors of the transmission line should lie together
>as closely as possible.
>
>The third conductor (you mention trifilar) makes the balun a "voltage"
>balun, which degrades its performance in antenna transmission line
>applications. This is explained in the article at
>http://eznec.com/Amateur/Articles/Baluns.pdf.
>
>Roy Lewallen, W7EL

FWIW, in the 1960s and 70s wideband CATV amplifiers, splitters and taps
(typically 50 to 600MHz), the bifilar and trifilar windings of the small
ferrite transformers were almost invariably NOT twisted.

I had a fair amount of experience in making some of these devices, and
can confirm that twisting tended to reduce the performance at the HF
end. With bifilar and trifilar windings, adjacent turns were carefully
interleaved. At least one manufacturer had bifilar and trifilar wire
specially made, with the strands lying side-by-side, and glued together.

If the individual strands of wire are already in intimate contact,
twisting them does not lower the characteristic impedance. Of course, if
the windings are NOT in intimate contact, the characteristic impedance
will be higher, and less-well defined. But, as the length of the
windings will be short compared with the wavelengths of the RF signals,
does this matter?

Twisting has the effect of allowing the increased possibility of part of
one winding randomly lying immediately adjacent to the next turn of the
same winding, so it tends to increase the self-capacitance of the
winding. This is almost certainly what reduces the HF performance.
--
Ian