Topband: W8ji ATR-10 design 160M?

[Jim Garland]

As Tom notes, the widest possible matching range for a T-Network is with two
adjustable capacitors (and, presumably, a variable inductor, typically a
roller inductor). Constraining any one of these parameters reduces the
available matching range. Further, Tom correctly observes that many hams
damage their tuners by unwisely choosing a combination of C and L that
dissipates much of their transmitted RF in the tuner itself.

Several years ago, I wrote a three-part series of articles for QST about a
high-power homebrew T-match autotuner. Part I of the series (QST, April
2002, p.40) shows two graphs (Figures 2 and 3) that give the range of
adjustment values for Cin, Cout, and L for a T-match network, using the 160m
band as an illustration. The first graph (Fig.2) shows the values of Cin and
Cout (for different values of L) required to match resistive loads from
3-800 ohms. The second graph (Fig.3) is the more interesting one for this
discussion, however, because it shows the percentage power loss in the
T-network for different values of L. 

Here is the key point: With a T-network, the lowest power dissipation occurs
with the smallest value of L that will give a match. To illustrate the
importance of this rule of thumb, suppose one is matching a 6.25 ohm
feedline (8:1 VSWR) at 1.8 MHz. According to the graph, the largest
practical nductance that can  match this load is about 25 uH, while the
smallest practical inductance is about 2 uH. Although each inductance will
give a satisfactory 1:1 match (assuming Cin and Cout are properly chosen),
the 25 uH choice dissipates nearly 40% of the transmit power in the tuner,
which is about 600W at the legal limit. Goodbye tuner!  On the other hand,
the 2 uH choice dissipates only about 45W. (Note that these values are only
for resistive loads, but they illustrate the general principle.)

There is a tradeoff, unfortunately, in using the smallest possible value of
inductance to match a given feedline, and that is that large values of Cin
and Cout are required. For this example, using a  2 uH inductance requires
Cin and Cout to be greater than 1000 pF. With a 25 uH inductance, Cin and
Cout only need to be about 100 pF. The problem is that designers of
commercial and homebrew T-network tuners sometimes skimp on the capacitance
range available, instead opting for larger (and cheaper) inductors. Buyers
are initially pleased at the low cost and wide matching range of their
spiffy new tuner, only to discover later that the tuner destroys itself when
they turn on their amplifier. And, as Tom notes, the problem is compounded
if they tune their T-networks incorrectly by following the wrong rule of
thumb!

73,
Jim W8ZR

> -----Original Message-----

> Subject: Re: Topband: W8ji ATR-10 design 160M?
> 
> Respectfully, the X match is nothing but a  common T-match with a fixed
> capacitor in one leg. In a normal T-match, operating Q can be varied over
a
> wide range by adjusting L/C ratios.  By restricting range of one leg,
> operating Q range is limited. So it actually **is** a T match, it just has
> two adjustable branches instead of the more common three adjustable
> branches, restricting the operating Q range and matching range.
> 
> The ATR10 is more like the common old Johnson matchbox, with the
exceptions
> instead of a link it has a tap and it is single ended. The tap sets the
> operating Q, just as the link ratio sets the operating Q in a Matchbox.
> This restricts the matching range and operating Q range.
> 
> Tuners that fix the operating Q at a certain value, or limit the operating
Q
> range, will always restrict matching range. The operating Q restriction
> limits peak voltages or currents by preventing grossly improper
adjustments,
> but the very same thing that limits voltages or current by definition also
> limits matching range.
> 
> An L network limits matching range the most of any network for a give
range
> of component values, but also limits operating Q the most. It has only one
> operating Q available at any given impedance ratio. The Q varies with load
> impedance. You cannot have too much Q, or it won't match.
> 
> This is, unfortunately, the way the world works. Everything is a tradeoff
of
> matching range, cost, complexity, and power rating. No single network,
just
> like no single balun, is all things to all cases.
> 
> The widest matching range for a given cost is a T network with two
> adjustable capacitors and one shunt inductor. Unfortunately, people think
> (and articles repeat) the silly idea that the proper way to tune is to
start
> with capacitors at half and tune for maximum receive.
> 
> Most of any improvement centers around preventing people from doing
> something silly, by limiting what they can do. There isn't any magic
> circuit.
> 
> Tom
> 
> ----- Original Message -----
> From: "N4XM Paul D. Schrader" <n4xm at iglou.com>
> To: "Jim GM" <jim.gmforum at gmail.com>
> Cc: "topband" <topband at contesting.com>
> Sent: Wednesday, October 16, 2013 11:01 AM
> Subject: Re: Topband: W8ji ATR-10 design 160M?
> 
> 
> > Jim,
> >
> > I did not make the comments you said I made below.  And the XMATCH
Antenna
> > Tuner is NOT a "T" match.
> > It is a patented circuit.  This unique circuit is a United States
Patent.
> >
> > And see http://n4xm.myiglou.com
> >
> > Paul  N4XM
> 
> _________________
> Topband Reflector