[TowerTalk] Tuners

[Tom Rauch]

Some of this might be interesting as a general topic. I've edited out 
any names because it was an e-mail exchange.

> Isn't it so that like all T tuners one has an nearly infinite number
> of settings which give a match?  Some are very lossy and some 
are

Loss gradually and smoothly increases as you use less 
capacitance, and more inductance, than necessary. That's 
because voltage and current increases smoothly as you decrease 
capacitance

> not.  This is why I like the Johnson matchbox. Only one setting is
> best match and it's low loss to boot.  Not having to crank the 
roller
> inductor a thousand times to change bands is a real plus.  Turn 
the
> bandswitch, set the two knobs to the markings and you're done. 

Actually you suffer more loss than needed at some settings in the 
Matchbox. Minimum loss for a given impedance ratio (possible for 
the components used) is available only at one or two impedances 
in the Johnson Matchbox because it has fixed link settings.

Every matching network configuration requires a certain minimum 
"loaded Q" to match two impedances, and when you exceed that 
minimum loaded "Q" power handling and loss is compromised.

The Johnson Matchbox has what amounts to a fixed Q, that Q 
limits the matching range. With a T network, you make Q any 
amount necessary to match an impedance, and keep circulating 
currents and voltages (and loss) at a minimum.

The simple rule in a CLC T network is use as much capacitance as 
you can to match the two impedances.

An L network always insures you do that by forcing you to use 
minimum Q, but it also offers the lowest matching range of any 
network for a give value of components. That's true even if you can 
change the configuration of the L network.    
 
> Another problem I have with T-match units is that they don't have 
the
> first harmonic suppression of the other designs.  Not really a 
problem
> these days with amps etc meeting FCC specs (in theory).

Any simple network can help or hurt harmonic suppression, 
depending on the layout of components, length of cable, 
impedance presented to the PA at the harmonic frequency, etc.

A Matchbox isn't necessarily a lowpass configuration, especially 
when matching low impedance inductive reactance loads.

The best way to deal with harmonics, is to deal with them in a 
know impedance load....like the 50 ohm line. Harmonics are a non-
issue now in most cases anyway.  

> What I don't like about the Johnson Matchbox is that you have to
> change the configuration (in the unbalanced feed case) when 
going to
> 15, 12, and 10 meters.  If you don't you WILL melt the solder
> connections on the bandswitch due to high circulating current.

That illustrates the problem of using a fixed link when the turns in 
the secondary change. Matchboxes are near as good as people 
think, unless the link to secondary turns ratio is optimized for the 
desired load impedance on a given band. 
 
> It doesn't work on 30m or 160M.  Not bad for a 40 year old design
> though.  The total cost in 1954 according the the manual was 
$95.

What is that in today's dollars? 

At today's manufacturing cost, you would wind up with a $3000 or 
more tuner that had less impedance range and less predictable 
current balance in most real-world applications than a $500 T 
network with a current balun.

We often look back with fond memories on how good things were, 
and don't realize how good we have it. Cars and radios included, 
although the women were younger then!
73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/towertalk
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com