As a no-feedback follower circuit with less than a gain of 1, the B1 design elegantly bypasses most of the mechanisms by which many assume less-than-ideal music reproduction can manifest. Once you want to add gain to the equation, "Pandora's box" is opened. A significant portion of the DIY community feels that the "magic" is lost when global negative feedback is employed... 
Part 1 of Nelson Pass's A75 project (http://www.passdiy.com/pdf/a75p1.pdf) illustrates a path to a complementary JFET/MOSFET greater-than-one gain stage (gain is about 20 as presented in the article). Nelson recommends an alternative gain-stage configuration in which the signal from the input transconductance stage is fed to the MOSFET stage via the MOSFET sources, instead of via the MOSFET gates (essentially a grounded-gate configuration); the total gain is reduced, but the sonic character is improved as a result. First, you could run the circuit open-loop (i.e., eliminate the feedback path), then you could throw-away additional gain by lowering the impedance of the MOSFET drains, although loading overall gain down (possibly tying the output to ground via a 100-1000 ohm resistor) to 5 might dramatically increase the aerobic workload of the circuit (you'd likely need to heatsink the MOSFET's...). I believe that Nelson himself once observed that, "a tired puppy is a good puppy..."
Charles Hansen, Ayre Acoustics, employs a similar topology for his highly-regarded Ayre V-Series zero-feedback preamplifiers and power-amplifiers.
Part 1 of Nelson Pass's A75 project (http://www.passdiy.com/pdf/a75p1.pdf) illustrates a path to a complementary JFET/MOSFET greater-than-one gain stage (gain is about 20 as presented in the article). Nelson recommends an alternative gain-stage configuration in which the signal from the input transconductance stage is fed to the MOSFET stage via the MOSFET sources, instead of via the MOSFET gates (essentially a grounded-gate configuration); the total gain is reduced, but the sonic character is improved as a result. First, you could run the circuit open-loop (i.e., eliminate the feedback path), then you could throw-away additional gain by lowering the impedance of the MOSFET drains, although loading overall gain down (possibly tying the output to ground via a 100-1000 ohm resistor) to 5 might dramatically increase the aerobic workload of the circuit (you'd likely need to heatsink the MOSFET's...). I believe that Nelson himself once observed that, "a tired puppy is a good puppy..."
Charles Hansen, Ayre Acoustics, employs a similar topology for his highly-regarded Ayre V-Series zero-feedback preamplifiers and power-amplifiers.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.