Aleph X with BJT, anyone?

[h_a]

In my simulations I have found that I can get a lot closer to the 12v rail voltage using a 2N3055 BJT in the CCS (constant current source)in the top half than I can using the FET's.

You get closer to the rail as BJTs limit rail swing by 1 volt or so (VBE plus little extra) and mosfets eat VGS (3.5 V or so).

Junk box amps are a nice playground! Just keep in mind not to go lower than the 12 V rails as the capacitances in the fets are very non linear below 12-15 V or so.

Hannes

[geraldfryjr]

Yes,that was was what I had figured out if I were to use a higher power supply voltage this problem goes away.

I am working on the circuit for use on a 12v pc supply so I needed to get as close to the rail that I can.

My first version uses opamps to drive the output Device,So this compensates for most of the non-linearity's even though it is still operating in class A mode.

I also have a fully transistorized version in mind and that will be the one that I will present here in the Pass-Labs Forum when I get it working.

jer

[Nelson Pass]

Originally Posted by 3nity

Don't know if it works properly.

Possibly the current source would work better as a "mu follower" as the Vce
across the little npn transistor which regulates it is low, and you will see some
non-linearity there. ZM or someone here will point you toward such a circuit
if you don't already know it.