Project history & acknowledgementsI have been dabbling in audio design and DIY work for over 25 years, and in early 2004, I decided to build a headphone amplifier to power my newly acquired Sennheiser HD600 headphones. I had long been fond of using power MOSFETs as amplifier output devices. I also wanted to try some of the newer high performance opamps as the driver for the MOSFETs. While doing some web research I came across the SDS Labs MOSFET headphone amplifier by Sheldon Stokes, and it was almost exactly what I had in mind. I built the amplifier with Sheldon's board and made several modifications, and called it the AMB β18. I presented photos and descriptions of this amplifier at the headwize and head-fi forums, and it to a local meet where it received very favorable comments.
This caught the attention of Morsel, who is active in the DIY community and had been the driving force behind the PPA, Pimeta and Meta42 amplifiers. As it turned out, Morsel is also a MOSFET fan, and had wanted to do a headphone amplifier topology similar to the PPA, but with a MOSFET output stage. Morsel contacted me to see if I had an interest in working with her on such a project.
I thought it was a great idea. What's more, Morsel and I live within a few miles from each other, so all the planets were lined up for a great collaboration. We laid down some basic design criteria, and sketched out several possible output stage topologies featuring MOSFETs. I acquired parts and built experimental prototypes on breadboards, testing and listening to them. We quickly arrived at a very elegant and simple solution that is essentially a hybrid of the PPA and the SDS Labs amplifiers, with some added enhancements. We had also toyed with a discrete diamond buffer topology using MOSFETs as the output devices, but found that the simpler, direct-drive MOSFET design to perform better.
Our project was announced to the head-fi DIY community in late August of 2004. While I was still testing on the breadboard, we invited a few local "golden-ear" head-fi'ers to come and listen to the prototype, and made some minor design adjustments based on the outcome.
Morsel meanwhile began working on the circuit board layout. We posted work-in-progress snapshots in our forum thread and solicited feedback, and where it made sense, incorporated the suggestions.
Our first printed circuit board prototype was finished in January of 2005, and again we invited the local golden-ears to audition it. We also made some minor revisions based on our experience with this prototype and further feedback from the forums, and went to "production" phase.
Since the M³ amplifier would not have been what it is without the other designs that inspired it, credit must also go to ppl and tangent, who are part of Team PPA, as well as Sheldon Stokes for the SDS Labs MOSFET amplifier.
Look for more exciting DIY projects from AMB in the future!
Main: M³ Main | Prev: Specifications | Next: M³ gallery