2.4:1 balun design failure

A lost soul searching for enlightenment on impedance transformation sought advice on a transformer at 2.4 : 1 BALUN.

Inevitably one of the forum experts counselled:

Assuming your quad is a single-band HF antenna, a conventional transformer using #2 powdered iron would be my choice for the balun function. The reactance of the secondary winding would need to be at least 600 ohms.

So, let's put the forum expert's advice to a practical test.

Fleshing out the proposed solution

I have at hand a T200-2 core, so lets calculate the secondary turns to satisfy the proposed solution.

Screenshot - 27_05_2015 , 07_05_11

Above is calculation from a popular online calculator. For 14MHz, the secondary should be at least 23.8t. We will use 24t.

The implied solution is a transformer with turns ratio (120/50)^0.5=1.55, we will use 1.5 or 3:2 (secondary to primary) and with a 24t secondary, the primary will be 16t. Impedance ratio implied by turns alone is 2.25 (a touch short of the requirement of 2.4, but practical).

Measurement

The transformer was constructed with 16t of the 24t secondary twisted bifilar with the primary to minimise flux leakage.

The transformer was loaded with a 120Ω resistor which was measured to have negligible reactance at 14MHz and swept from 1 to 30MHz. Based on the forum expert's advice of a minimum secondary winding reactance of 600Ω at the frequency of interest (14MHz in my example), we might well expect poor performance below 14MHz.

With the ideal transformation ratio of 2.25, we expect input impedance to be 120/2.25=53+j0Ω

Screenshot - 27_05_2015 , 12_03_33

Above is a sweep of the input impedance of the loaded transformer.

The transformer is not anywhere near ideal, the input impedance at 14MHz is 57+j143Ω, input VSWR is 9:1.

An appalling failure of design by any standard.

Reasons underlying this failure are discussed at Coupled coils – a challenge for hams!

Conclusions

Simple turns ratio based models and explanations are not an adequate general model for these transformers.

Footnote:

Regular readers of my blog will have noticed a series of articles casting doubt on the reliability of the AIM system. The measurements above are from the AIM system using AIM900A software (and there are some minor defects in the display), but they have been validated with a Rigexpert AA600 which gave very similar results.

Since the measurements made above, the developer has released AIM900B but as is customary, he does not admit to or identify the problems that were fixed and impact on accuracy of previous measurements. The Array Solutions AIMuhf is a truly low quality measurement system!