Autoformers: The Benefits of Matching Amplifier to Speaker
There has been a great deal of conversation about autotransformers (autoformers) in this forum. Many think they are like the output transformers we use in tube amplifiers. They are not for some very important reasons. First, they are not wound the same way as they have no high voltage insulation. Autoformers are wound with heavy low resistance wire and all the winding is used all the time. In addition, part of the signal current is direct, and part is transformed. Now let us consider a few more points:
THE WINDING: When we make a traditional output transformer we must insulate the primary from the secondary for over 1000 volts. This insulation takes up space and winding space is most dear to the designer as we want as much copper in there as possible. We then must section the windings and interleave them. An interleave of 5 is good and some think 7, 9, or even 11 is better but that raises the capacitance of the transformer and is hard on the tubes at high frequencies. An autoformer has no DC voltage in the windings and thus can be bifilar wound (taking 2 or 3 or more wires at once). This increases the coupling and extends the high frequency response by a factor of 2 or more. My output transformers are good to 65 KHz and the autoformer is good to 140 KHz.
THE CORE: As to the core, an EI core is preferable over a toroid as the toroid will have saturation problems if connected to an amplifier that has DC offset. An offset as low as 20 mV can swing the core in one direction toward saturation. An EI core has a very small air gap that will allow it to ignore rather large offsets.
IN THE AMPLIFIER: Here is where the difference between a conventional output transformer and an autoformer occurs, and why Wiggins at Electro Voice created the CIrclotron circuit. In a conventional tube amplifier, for most of the signal, only one half of the output transformer is active. One half of the primary is used for the positive going signal and the other half used for the negative going signal. It is very difficult to make the two halves of a push pull transformer identical above 20 KHz where the feedback really cares about phase shift. Even the taps on an ultralinear transformer can go out of phase at high frequencies. This causes the amplifier to ring on one half of the square wave. In an autoformer the entire primary is used all the time. Though not widely talked about, we who design amplifiers are very familiar with this problem. Wiggins realized that if he put the transformer in a bridge circuit that the primary would act as a whole and this problem would go away. That is the essence of the WIggins Circlotron. Since Wiggins wanted to keep the amplifier efficient, he used a high ratio transformer with conventional taps. Of additional note, we do not put taps on an amplifier to "match" the impedance of the speaker as we know it varies. We put them on there to deliver the proper ratio of voltage and current to make the amplifier happy. You can always use a lower tap and enjoy lower distortion, better damping, lower noise, and extended tube life. You also extend the Class A region. The only reason to use a higher or matched tap is to get the most power out of the amplifier if you play it loud. In the RM-10 manual I suggest this strongly and have termed it "Light Loading".
IN THE SPEAKER: In the application of making a speaker easier to drive we use the autoformer in a step-down configuration. For example, stepping down the voltage by a factor of two doubles the current. A step-up configuration would be a rare but still valid usage. Although this discussion is based around impedance, I would like readers to consider voltage and current. Impedance is just a simple way to combine the two into one term. By presenting a higher impedance to the amplifier it will have MORE control over the speaker, and control equals damping. In addition, the autoformer extends the Class A region when used to increase the impedance as in this discussion.
Now, what is an autoformer going to do for you? Well, if you have an OTL amplifier you should know that the power is greatly reduced into low impedance loads. Even worse is that low impedance loads will overheat the tubes at high power levels as most of the power supply voltage is being dropped across the tube not the load. So low impedance loads are hard on the tubes and cause higher distortion. All these ills can be solved using a proper autoformer.
For OTL amplifiers that have high output impedance and produce their best performance into 16 to 32 ohms one needs a 6 or 8 to one step down ratio. This will make the speaker and amplifier very happy and still preserve the qualities of the OTL. A 4 to 1 is not enough. This is no problem to make, and I have been using mine for many years. An autoformer can also be used in reverse if one has a low voltage, high current amplifier like a Mark Levinson ML-2 which is 25 watts into 8 ohms but 100 watts into 2 ohms. Again a 4 to 1 will get you 100 watts, but now into 8 ohms and an 8 to 1 even more. Remember the impedance ratio is the turns squared. So, even a 9 to 1 impedance is only 3 to 1 turns and 1/3 of the signal is direct through the primary.
I hope this clears up the differences in these two very different types of transformers and we can stop considering them as the same. While some may consider an autoformer a band-aid, I consider it a device that makes the problem go away.
[Source: August 31, 2018]