Electra-Print Logo - Audio Tube Amplifier Components 
Bookmark and Share

Electra Print Audio

Tech Blog

Home About Amps for Sale Schematics Tech Blog eBay Sales Contact/Ordering

Single EL-34 Amp coments

Testing Transformers - Elliano  (current page indicator)

SE Transformer Design Considerations

PVA Pre-Amplifier

Testing Transformers - Elliano


Jack Elliano


Ringing from square wave measurements has grown into a very negative response within the incorrectly informed audiophile group. The influential ones spreading this have probably measured an amplifier that has a flat frequency response from 100hz to 6000hz. This would show no ringing at all.

In order to slow down and hopefully stop a continuous barrage of incorrect technical statements used for influence and profit, it becomes necessary to confront them with facts. An example is: the measurement of speaker interconnects of its inherent capacity and inductance involved as featured in advertising. Realistically there is a problem with music that is played in the megahertz range. One statement heard to confront this is, "These advertised measurements are like being concerned about the aerodynamics of a golf cart." This is the intent of this article. TRANSFORMERS THAT RING WHEN SQUARE WAVES ARE APPLIED

A square wave is composed of a leading edge with a rise time in the megahertz range (depending on quality of generator and frequency used) then to a given duration of a DC voltage across the top, then switched down (same speed as the rise time) in the megahertz range then to cross zero and repeat in the opposite polarity. Just to inject something here, we did have a transformer that will pass a DC voltage, it was shorted out.

This bipolar waveform, when measured with a Fourier series analysis, shows the square wave is composed of its fundamental frequency and an infinite number of odd harmonic frequencies. The even order content would have to be zero for a perfect square wave.

The use of a square wave to analyze an amplifier is shown in full in Tremaine's Audio Cyclopedia, 2nd edition, Audio frequency measurements chapter, pg. 1521, as a means to test the transient response of an amplifier. If the amplifier has ringing at the top of the leading and falling edge, this shows an extremely good high frequency response from the transformer in use. Well, there it is folks!!

If one remembers the ignition system on the old engines, they used a set of points, generating a square wave of 12volts and applying this to a capacitor and spark coil (a very high ratio transformer for those who don't know). They rang to 15,000 volts or so. This is what a square wave will do to this combination of coils and capacitors.


A basic audio transformer with, lets say, a 10k primary and a 600 ohm secondary. We do know that the primary has many more turns than the secondary. Lets say that the total length of the primary wire is 1000 feet long and the secondary is 100 feet long. For this transformer to be perfectly coupled (little or no ringing), it would need to have the 100 foot secondary wire fully coupled next to the total of all the 1000 foot primary. How can this be done?? Well it can't! End of statement. If one is to build a transformer with different primary to secondary ratios, like most of them are, then some portion of the primary is not coupled to the secondary, therefore ringing will occur due to what is called "leakage reactance". This is the measured primary inductance left on the primary when the secondary/s is shorted. This reactance will react with the capacity within the coil to resonate. when a square wave is used and it is composed of the frequency (usually in or near the megahertz range, ringing can occur. When a transformer is in operation and loaded, the open winding or static inductance measurement is lower. Much higher though than the shorted "leakage reactance" indicated above. The ability to ring with a square wave is in play usually within a 10khz square wave.

An example is a class C operated RF amplifier, the output power, which is a pulse, going to the tuned tank circuit (coil and capacitor), tuned to resonance, will show a ringing waveform. The flywheel effect of the high Q resonant tank circuit, stores, combines and transfers this energy to the antenna as one frequency. By the way, this tank circuit is a transformer with a primary about 20,000+ ohms to a 50 ohm secondary, with no core. That's another story.

The one audio transformer that will show little or no ringing is the bifilar wound (two equal lengths of wire wound together equally in turns); this is a 1:1 ratio type with total coupling of all wire involved. Try to design and build an amplifier with only 1:1 transformers. Smaller variations of this bifilar wound transformer are known as pulse transformers, usually used in radar and other pulse operated circuits. A pulse is a square wave of one polarity. Get the picture.

To sum it all up, a transformer is probably the most complicated device ever invented. Its many intricate designs and uses, then to have the few accepted, and common measurements made by those who never even built a transformer, evaluate them for all, is simply absurd.

Have you ever seen a wideband oscilloscope preamp circuit use a transformer? Those of you that have circuit experience will answer NO!!

To cap it off, those influential few who apply square waves to test and evaluate transformers could probably say "we played the clarinet by blowing into it backwards and it sounds very bad so in our opinion don't buy one."

Last updated: Monday September 19th, 2016 at 12:27:27 PM

Site MapLinksPhotos
Copyright 2007-2022 Electra-Print Audio   All Rights Reserved