Topband: W8ji ATR-10 design 160M?

[Tom W8JI]

> 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.

As early as 1983, when I reworked the ATR8 that Prime inherited from Amp 
Supply's engineering, I re-wrote the manual. Prior to that virtually every 
antenna tuner manual or "how-to guide" suggested setting the capacitors at 
half scale, and peaking the inductance for maximum noise.

When I designed the ATR15, I used fixed inductor settings that limited Q and 
tuning range. This eliminated the issue of picking too much inductance.

Later, I rewrote MFJ manuals, eliminating the suggestion to set the 
capacitor at half value.

There are three problems caused by using too much inductance, or too little 
capacitance.

1.) Voltage is too high. This causes arcing of capacitors, and can result in 
an instantaneous failure problem.

2.) Voltage across the inductor is too high. This can cause too much 
circulating current, and damage the inductor. This is a **time heating 
issue** related to average power.

3.) The network has to be retuned more frequently with frequency changes.

All of this is far more complex than a simple dissipation issue, because the 
self-Q of an inductor varies wildly with inductance tap. Not does unloaded Q 
vary, the ability of a roller to dissipate heat also varies. If most current 
is concentrated in a few turns, the inductor will be able to dissipate much 
less power. Optimum power rating is often not where inductance is lowest, 
even if that is lowest inductor loss.

It almost takes a book just to describe the actions inside an inductor as 
the system is changed.

The result of this, just like with baluns and MOV surge suppression, is that 
virtually any article, discussion, test, opinion, or web page will give an 
incomplete view. The system is too complex to give ten line answers, let 
alone three page answers.

This is why Jim Garland had to write a three section article, and even doing 
that, it remains a narrow slice of everything.

Also, despite claims by the ARRL and others that more capacitance increases 
power ratings, adding more capacitance does not always increase power 
rating. This is because an inductor's ability to dissipate heat, and even 
unloaded or self-Q, can increase faster with more inductance than electrical 
stress on the component decreases.

Simple answers are good for sales, and everyone likes a "this is what you 
need or don't need" answer, but devices with complex loads or complex system 
interactions cannot be represented by a simple model, simple analysis, or 
one line answers.

The world of tuners is like a big hobo soup kettle. T network tuners are not 
called T network tuners, and Pi network tuners are often not really 
operating as Pi's. Many things are called something just to help marketing, 
even if they aren't really working as what they are called. An optimum 
design depends not only on the load and band, an optimum design depends on 
the physical construction of the box all the way down to the construction 
and behavior of components inside the box.

The only accurate, simple, answer is "it depends".

73 Tom