Common failings of EFHW matching transformers

I have written many reviews of published EFHW matching transformers, and in most cases the reviews have reported estimated or measured losses that are appalling and not disclosed by the ‘designers’.

Why is it so?

I am asked, why is it so?

Up front, I do not know the answer definitively, but let me offer some thoughts based on the designer’s own articles and discussions by ‘online experts’.

Apparent reasons include:

  1. lack of understanding of ferrite and powdered iron core material behaviour;
  2. lack of understanding of coupled coils, and mutual inductance;
  3. use of inductor design tools that are inadequate at radio frequencies;
  4. lack of competency in basic linear circuit theory analysis for AC circuits;
  5. failure to make meaningful measurements of the built article;
  6. focus on input VSWR as a single metric indicating goodness;
  7. reliance on QSOs for evidence of performance;
  8. an attitude that antenna system radiation efficiency doesn’t matter, particularly for QRP (if the term antenna system radiation efficiency is even understood as a quantitative metric).

Lack of competency in basic linear circuit theory analysis for AC circuits

Item 4 in the list, lack of competency in basic linear circuit theory analysis for AC circuits is a most fundamental problem, fundamental in that it may explain the reason for most of the other failures, and a failure to appreciate this discussion.

Calculating impedance of an inductor at RF

Inductor design tools are sometimes cited by designers, but I can not recall ever seeing a tool used that estimated the complex impedance at radio frequencies.

At PD7MAA’s BN43-202 matching transformer for an EFHW I reviewed the design of his transformer with a 1t primary. Here is a calculation using one of the popular online tools.

Inductor impedance is estimated at 0+j98.1Ω, the zero real part means that it is lossless. In fact this tool is calculating the nominal inductance at around 10kHz, and based on that, the impedance at 7.1MHz, but the method is flawed.

I used a different design tool which estimated the impedance of the inductor at 28.5+j40.9Ω. A prototype inductor was subsequently measured and it was close to that value, validating the estimate.

It is this single item that reveals that more power is converted to heat in the transformer core than is radiated.

The rest

The rest of the contributing reasons are evidence of the amateur nature of amateur radio, and as time passes, the hobby evolves into a more amateurish activity with diminishing interest in science or engineering.

The Internet provides a great communications channel, providing opportunity for authors to publish designs, good or bad. Any filtering of the wheat from the chaff must be done by the reader.

The growing emphasis on popularity, eg the Like button or equivalent on so many platforms (including amateur radio forums such as QRZ), reinforces the notion that popularity determines fact… about as far from Scientific Method as you can get.

There is no doubt there are a lot of EFHW matching network designs that are very popular, but the fact is that many have low efficiency  and contribute to low antenna system radiation efficiency.