TopBand: Beverage matching

Tom Rauch
Fri, 26 Feb 1999 11:17:42 -0500

It's actually tough to make a standard transformer that does NOT 
work well from far below 160 meters to 10 meters and beyond. As 
Earl pointed out, this stuff isn't that critical.

The minimum number of turns used, other than finding the "sweet 
ratio", relates only to the self-impedance of each winding. The self-
impedance has to be perhaps five or more times the load or source
impedance it terminates to prevent the windings impedance from 
loading the circuit. 

What most people fail to realize is that impedance relates almost 
exclusively to the core length the wire parallels INSIDE the core 
window. The impedance change is linear with exposed length but 
follows the square of turns.

For example, a 73 material core, of almost ANY reasonable 
outside dimension, cross section, or shape, has about 120 ohms Z
per inch for a single pass through the hole. That Z goes to 480 
ohms for two passes. With a broadband low power core, all you 
really care about is winding Z when selecting a core. The Z varies 
with frequency, and does NOT relate to dc permeability by itself. 

For reference some common materials in terms of Z per inch at 2 

75 = 150 ohms
73 = 120 ohms
43 = 40 ohms
44 = 25 ohms
61 = 10 ohms

Materials at the lower Z end handle high power with less heating 
and produce transformers with less bandwidth, the higher Z 
materials are best for low power applications broadband 

A 73 material core of any dimension requires about  1/3 the turns 
of a 43 material core of the same dimensions. The 75 ohm cable 
Beverage transformers I use have 73 material, and only need two 
turn primaries and 5 turn secondaries and function from below the 
BC band to 30 MHz plus with no special windings. If I used 43 
material, they would need at least six turn primaries and 15 turn 
secondaries for the same LF cutoff, but would not reach the high 
end because of stray capacitance in the windings. 

For low flux density application like 160 meter through ten meter 
receiving applications, 73 is almost ideal. Whether the winding is 
transmission line type (bifilar, quadfiliar, trifiliar) or simple isolated 
turns makes NO difference at all in required turns when the correct 
core is used. The tight coupling extends the upper frequency limit, 
NOT the lower.       

> In retrospect, I think another benefit of this structure is to reject
> out-of-band signals. I think (it's been a few years, I'd have to find my
> notes to be certain it was part of the plan) that's why I did not use an
> (i.e., crud from the broadcast band or static bursts, which look like
> 100 kHz pulses if memory serves me)
>  is to use separate windings - primary and secondary. In an
> autotransformer,
> there is no isolation, so anything
> picked up by the beverage goes straight into the receiver. With separate
> primary and secondary windings, the inefficiency of the transformer at
> lower frequencies would provide at least some attenuation

The static we hear is IN BAND noise.

> of a BC band signal and keep it out of the receiver. The same reasoning
> applies to a lightning impulse. And the primary winding goes straight to
> ground at the Beverage.

Bypassing out-of-band signals does NOT reduce in band noise 
unless the receiver is a total mess. The noise you hear on the 
receiver, just like signals, is noise that is really at the operating 
frequency. The only exception would be if the receiver is 
overloaded, and in that case the ONLY cure is a bandpass filter or 
new better quality receiver.

An isolated primary is a GOOD idea if you make sure the cable 
shield is NOT grounded to the antenna's ground. An isolated 
winding does NOT "cut-off" or attenuate BC signals or reduce noise 
coupled from the antenna itself. It simply stops the cable shield 
from being part of the antenna system.

73, Tom W8JI

FAQ on WWW:     
Administrative requests: