Ive been playing with a class B assisted class A output stage whereby a lightly biased class A output stage drives class B biased current controlled current sources which provide current assist for the class A stage - which keeps it in class A mode. It can drive ~ 200W RMS into 4 ohms, never leave class A mode, and it only burns ~65 Watts quiescent. Below is the idea. The biasing of the current controlled current sources is accomplished by servoing the class A bias in the class A stage. Biasing tracks really well across temp. I also had to clamp the VAS stage to prevent the output from swinging up too close to the rails as this caused instabilities from a swelling Cbc in the CCCS stage. A schematic of the output stage is attached. The rest of the amplifier is your standard input diff pair and darlington VAS. It also has one of my discrete op amps in the very front end.
The concept is similar to the QUAD405 , but with drastically different implementation. SPICE says its awesome, but I lack THD testing gear. Anyone ever see or try anything like this before? Im trying to decide if I have invented something novel. Any feedback or thoughts are welcome. I was going to post a link to my website that has more photos / info, but as a new user I couldnt.
It can be found at sparkoslabs under projects.
