Topband: Share of a common mode current experience

[Jim Brown]

On Mon, 11 Jun 2007 00:13:25 -0700, Jim Sullivan wrote:

>I checked on the cost of the toroids and the 2.4" #31 units seem to be >
>$15.00 each.

This is a VERY high price. 

which of the fair-rite distributors gives the best prices for 
>    quantity 10-15

Try Lodestone Pacific and Kreger. See the distributor's section of Fair-Rite 
website for contact info. I've done biz with Kreger several times, and 
recommend them. 

Also, Fair-Rite is VERY open to selling direct for large quantities, and the 
prices get a lot better. I've coordinated two group purchases for local ham 
clubs. When you buy direct, there are price breaks at 250, 500, and 1,000 
pieces, and they are quite substantial. I suggest that you get together a 
group purchase for at least 500 pieces, and preferably 1,000 pieces. When you 
do, the per/unit cost will be under $3, with shipping. 

To buy direct from Fair-Rite, you must spend at least $500, and must buy their 
minimum quantities of a given item. The #31 toroids are packed in boxes of 
100, and I believe that minimum is at least one box. To get the 250 piece 
price, you must buy 3 boxes. 

>As a result I have 4 questions.
> 
>1) Increasing the turns reduces the number of toroids required for the
>target 5K ohms
>   What is the down side of using fewer toroids?  (dissipation?)

I'm working on an update/expansion of the "cookbook" section of my piece on 
baluns. Unfortunately, work is getting in the way of my collecting enough data 
to put it in the form that will will most useful. 

>2) Can I use teflon coax like RG142B/U (.195" diameter) to allow ~15 turns
>with
>    fewer toroids to achieve the same 5K ohms?

Yes. The data for small wire is a first approximation of what this small coax 
will do. 
> 
>3) Does the diameter of the turns passing thru the Toroid stack effect the
>predicted Rs.

R comes from the short length of the coax passing through the ferrite cores, 
not the coil. The coax outside the cores adds inductance and stray 
capacitance, both of which lower the resonant frequency and raise the Q. An 
ideal choke has lots of R and a very low Q (which makes it more broadband). 

So it is important to minimize the contribution of the coax that is outside 
the cores. That contribution will be least if the diameter is smallest and the 
turns are widely spaced. Increasing the diameter of the turns lowers the 
resonant frequency of the choke a lot if you allow the turns to be closely 
spaced outside the toroid stack, but not a lot if you do not. The MOST 
important thing about winding these chokes is to keep the turns widely spaced 
outside the stack. I recommend the smallest diameter that does not compromise 
the recommended bend radius of the coax. 

The following links are to some "internal communications" I prepared for other 
guys doing research on this topic. They include photos of the chokes I 
measured, some details of the measurement setup, and some data. I suggest you 
read them in the order they are listed below, which is how I wrote them. 

These data are the best I have right now for chokes wound with coax. Note that 
they are scalar, so they are the magnitude of the impedance. Circuit analysis 
101 tells us that this is a parallel resonant circuit of very low Q, with the 
magnitude of R equal to the peak of the Z curve. The computation of L and C 
values is left to the student. :) As we move away from resonance, we need to 
view the choke as its series equivalent R and X (different values from the 
parallel equivalent, and different values at each frequency), and the X 
component will combine with the X of the feedline. At resonance, Rs = Rp, and 
as we move away from resonance, Rs gets smaller. Again, the computations are 
left to the student. BUT -- the graphs in the tutorial of measured data are 
good examples of how the circuit behaves, and can help you get your head 
around it.   

http://audiosystemsgroup.com/ChokeMeasurementsK9YC-K6TD042207.pdf
http://audiosystemsgroup.com/ScalarChokeMeasurements1.pdf

When Kevin (K6TD, who worked with me on the first series of measurements 
documented above) put up a 40M beam last weekend, he wound 4 turns through a 
stack of 7 #31's. He only did 4 turns because that's the best he could do 
without taking the connector off the coax. Two more turns would have been 
better. 

73,

Jim Brown K9YC