> By using 18 awg, I can space the bifilar turns to get some separation
> turn to turn to help increase voltage breakdown between turns.. and I
> won't use the full toroid..So I won't do a crossover winding, but
> still be able to get to the SO-239's on either side of the 4x4x2
> NEMA6 box without too much distance or extra capacitance.
Instead of mounting the toriod stack flat in a 4x4x2 box, why not
mount it "on end" in a deeper (e.g. 4x4x4 or 4x4x6) box? It might
take a little creativity to build a mounting baffle but standing
the toroid stack "on end" would allow it to be mounted very close
to the output side of the box and minimize the a amount of the
windings in proximity to any metal mounting plate or boom.
"On edge" mounting might also fit well with a rectangular box
for baluns with a larger number of cores - or series combination
of cores for wide band applications.
73,
... Joe, W4TV
On 5/31/2010 7:53 PM, Kevin Normoyle wrote:
> jim VE7RF mentioned the common use of crossover windings in toroid
> baluns, which is the area of some things I'm wondering about before I
> wind....
>
> Questions: Wouldn't it be better to use less of the balun (to avoid
> capacitance between the start and end turns) and not do the crossover
> to get to the other side?...I'm thinking double stacking type 31
> allows me to get away with only using 2/3 of the balun, which is
> good...
>
> Doesn't the crossover introduce physical closeness of two turns that
> have a larger capacitance between them, then the adjacent
> turns...i.e. it's making things worse for introducing capacitance
> that lowers the resonance?
>
> I thought the standard is to leave a 1/3rd gap on the toroid between
> start and end turns for this reason. So it would seem the crossover
> method violates this (halfway) ??? or am I misunderstanding?
>
> Another question: Everything I've read seems to say there's strong
> capacitive coupling between the turns and the toroid itself, (in
> addition to turn-to-turn capacitance).... which is why using
> enamelled or stranded wire turns are sometimes worse, since they wrap
> tighter to the toroid and get increased capacitance...I'm wondering
> if some of the benefits people see with coax turns is decreased
> coupling to the toroid...
>
> So insulated wire, or wrapped toroids, gets the turns up off the
> toroid, which is good. same thing when you see people doing these
> loose-turns around the balun.
>
>
> See here's what I'm planning with my toroids. I have some teflon
> insulated 18 awg stranded, which I think is fine for 1500 watts if
> reasonable SWR.
>
> I was going to double stack and teflon tape type 31, and wrap 10
> turns of that wire (bifilar).
>
> By using 18 awg, I can space the bifilar turns to get some separation
> turn to turn to help increase voltage breakdown between turns.. and I
> won't use the full toroid..So I won't do a crossover winding, but
> still be able to get to the SO-239's on either side of the 4x4x2
> NEMA6 box without too much distance or extra capacitance.
>
> I also can't understand when people say they try to get a bifilar
> pair to be 50 ohms based on wire spacing. From what I read, it seems
> like no matter what you do, you're going to get 75-100 ohms for the
> impedance of any wire pair. Has anyone actually measured 50 ohms for
> a real bifilar pair?
>
> I was also wondering if a triple core stack with just 8 turns might
> be even better. I can space the bifilar turns more, and less of the
> balun is used, so it's more of a straight run from so-239 to so-239
> on either side of the box.
>
> The inductance vs wire length tradeoff, I think, is better with the
> stack of two or three? I have to check that again, but since wire
> length is closely related to capacitance which causes the resonance
> issues, it seems like that's what we want: the optimal wire length to
> inductance tradeoff.
>
> I know Jim K9YC has measured THNN #14 awg on type 31, but I don't
> want to wrap pvc insulated wire turns that close for 1500 watts (his
> turn count is close the max you can get due to the inner core
> circumference limit?)
>
>
> -kevin ad6z
>
>
>
> _______________________________________________
>
>
>
> _______________________________________________ TowerTalk mailing
> list TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|