I have just purchased the latest Sound Lab M1

Hi Ralph,
It looks like it’s time for me to get involved in doing the best I can to explain the ZOTL technology so more people can understand the real issues. It is good that you pasted my patent abstract into your post, and from the complexity of the description, it is easy to see why the ZOTL technology is hard to grasp. In the near future, I hope to write as non-technical an explanation as I can and post it on the Berning web site to help people understand a little better the implications of the patent, and why it works the way it does.

I maintain the term OTL in the description in the patent and in other discussions of the ZOTL because, simply, there is no output transformer in these amplifiers. No such devices are called out in the patent. What are called out in the patent are converter transformers. This is not simply calling one thing by a different name; it really is very different in form and function.

Since it will take me some time to pen an explanation of patent, I will just say for now that the interested person can visit my web site and look for the review of the ZH270 that Chuck Hansen did in Glass Audio some years ago. Chuck is a switching power supply engineer and was able to fully understand the technology and tried to explain it in terms that non-engineers could understand.

But for now, I think that it may be instructive to make some comparisons between your amplifier and my amplifier. When I look at your amplifier, I see that not only is the speaker connected to the tubes, but it is also connected to a pair of floating power supplies. Is it not true that these power supplies have ss rectifiers and transformers in them? Is it fair play now for me to call your transformers output transformers? I would not do that because it is silly. Your transformers operate at a fixed line frequency that has nothing to do with the audio frequency. Likewise, the converter transformers in my ZOTL amplifiers operate at a fixed frequency that has nothing to do with the audio frequency. Remember, also, that the ZOTL amplifiers are dc coupled except for the input stages, where ac coupling may be used to prevent dc from being passed to the speaker. Clearly, this would not be possible if there were output transformers. Likewise, your output and driver sections are dc coupled.

The real difference between your amplifier, Ralph, and my ZOTL is the power supply. You have simple fixed power supplies that are designed to strike the best balance between allowing the tubes to pass enough current to the speaker and not overcooking the tubes too much. On the other hand, the power converters in the ZOTL perform optimum impedance matching between the speakers and the tubes because they can independently furnish the operating power to the output tubes while providing the current to the speaker via an arbitrary tube voltage step-down ratio, which is the same as the speaker current step-up ratio! In other words, the ZOTL frees the amplifier designer from the requirement that the same current that goes through the speaker must go through the tubes. The ZOTL further adds the wonderful feature of isolation between the high voltage of the tubes and the speaker, and this isolation simplifies dc coupling in that there is no requirement that the output tubes be operated at the same dc level as the speaker.

With the ZOTL, the tubes can be operated at their best efficiency. In the ZH270 and the new Quadrature Z, I choose to operate with a dc plate voltage of around 1400 volts. This would be too high for an audio output transformer, as the insulation would fail if required high fidelity tight-coupling winding procedures were followed. Operating the tubes at this voltage brings the efficiency of the tubes up to what would be expected of transistors in a conventional amplifier. This is why the Quadrature Z can boast 200 watts out of only four modest-sized output tubes (33JV6) at 8 ohms. This goes up to 270 watts at 4 ohms continuous without the tubes showing color. Ralph- that’s 270 watts of OTL tube power into 4 ohms!

David Berning

Some would conclude that my description above in many ways describes the characteristics of an audio output transformer, except for the dc-coupling attributes. In fact, the ZOTL can accurately be described as an ideal output transformer emulation circuit. When I claim that it is OTL, I am saying that the ZOTL lacks the physical item known in the audio industry as an audio output transformer. The physical object known as an audio output transformer is made using an iron-based core that is wound with perhaps a kilometer of magnet wire to obtain the required inductance to couple the full range of audio frequencies from a primary winding (tube side) to a secondary winding to which the speaker is normally connected.

The audio output transformer is a good way to couple tubes to speakers, and makes a lot of practical sense. It is the most economical way to make a tube amp, although I think the ZOTL could compete if made on a high-volume scale. For this reason, almost all tube amps ever made use such a transformer.

But let’s face it, the audio output transformer, with its hysteresic iron core and perhaps a kilometer of wire is plagued by parasitics that adversely affect sound reproduction. The first major OTL pioneer, Futterman, and later persons like Ralph of Atmasphere realized the sonic limitations of output transformers and went to great lengths to find a way to use tubes but eliminate the output transformer. (Apparently Futterman developed his earliest amplifiers with the goal of reducing cost, not achieving better sound). It should be noted that not everyone likes the sound of the OTL amp and some prefer the output transformer coupled sound.

For many years I made transformer-coupled amps, and I am very familiar with the sonic issues of these amplifiers. Almost any amplifier will sound good on reasonably well-recorded music of the pop and jazz variety. But I worked very hard to make some of my difficult recordings sound right. These were still well recorded, but the music itself was difficult to reproduce convincingly. One of my test records was Shostakovich String Quartets on L’oiseau-Lyre DSL011. I struggled with this record for years. I could tweek one resistor this way and make one instrument sound right, but another instrument would then sound wrong. The first ZOTL I built was soon put to the test with this record, and suddenly all of the instruments were reproduced correctly. My output transformer based amplifiers were suddenly obsolete in my mind.

The point that I am making with all this is that the audio output transformer is a sonic filter that while some prefers having it, there is a reduction of reproduced transparency particularly at the frequency extremes. Both traditional OTL designers like Ralph and designers like myself who come up with a new way to beat the colorations of the output transformer have the common goal of integrity of the sound reproduction. In my view, an amplifier that eliminates the audio output transformer, both physically and in sprit, falls into a class that is simply known as the OTL amplifier. In other words, either the amplifier has an audio output transformer characterized as described above, or it doesn’t. The sonic attributes of the ZOTL and the traditional OTL amplifiers are far more similar than different, although differences can be expected based on speaker matching.