PassLabs Zen V7 free download

File name PassLabs-ZenV7 pwramp.pdf

Zen Variations 7: More fun with Son of Zen and SuperSymmetry By Nelson Pass, Pass Laboratories Introduction One of the performance issues raised by the original Son of Zen (Audio Electronics, #2, 1997) was its efficiency figure, which was charitably described as 4% (500 watts in, 20 watts out). You may recall that this was dictated by the original requirements - no feedback, no capacitors in the signal path, and a single gain stage. Zen Variation 6 relaxed the requirements on feedback and capacitors in order to provide a tutorial exercise about "super-symmetric" feedback. The performance was improved in distortion and output impedance, but the efficiency was only slightly improved, largely because we used most of the original circuit. In the first part of this piece, we are going to raise the efficiency from about less than 4% to nearly 6% simply by eliminating the negative power supply rail and biasing the differential gain pair with a constant current source. This will also give improved distortion performance with a single-ended signal source over the circuit of ZV6. In the second part of this article, we will additionally raise the efficiency to over 16% by driving coupled inductors instead of resistors. This will about quadruple the efficiency of the original Son of Zen, but with much better performance in all areas except for the input impedance. Of course, it is also a slightly more complex circuit, and will keep in mind that the Zen and its brethren are all explorations in achieving good amplifier performance as simply as possible. What constitutes "good" is observed from both an objective (measurement) and a subjective (listening) viewpoint. It is my personal opinion that the simpler the circuit, the more similarity there tends to be between these two, but not always. Biasing With an Active Current Source The transistors of the original Son of Zen were biased up from the negative supply and did not contribute to the available output of the amplifier in a direct way, so there was considerable waste of energy. They needed to see relatively high impedance in the biasing network and they also needed to run at the full output current, and so they ran near the full rail voltage and used up about half the power. An active current source can operate at quite low voltages (as low as 2 volts for a power Mosfet without significant degradation). At low voltage it can carry the full bias current of the amplifier, dissipate a much smaller amount of energy and still offer a much higher impedance than the original network. Figure 1 shows such an arrangement applied to a Son of Zen, along with the feedback networks R1-R4 and P1-P2 that we applied in ZV6. I am going to call it ZV7-R. Q3 is the current source, and it sinks about 7.5 amps of bias current split equally through Q1 and Q2, which are the gain transistors for this single-stage circuit. The Gate of Q3 is fed control voltage by a network of resistors R5, R6, R7 and P3. This network senses the DC