FET/BJT hybrid amplifiers

[angel]

Have any of you tried using bipolars for the output stages of an Aleph, X or SoZ?

I'm looking to build a low power amp, like 5-10W into 4 ohm, for example using the Aleph topology. However, at these power ratings, the Vgs losses of MOSFET transistors get to be really noticeable. So, I'm wondering if it might make sense to use a N-channel MOSFET input, and a PNP output to cut the losses a bit, or whether this would be too hard to do without impacting sound quality.

 

[MikeW]

Build a Mini-A (mosfets), JLH (bipolars), or Le Monstre (Fet input Bipolar output). They are all nice amps. Or build all three.

 

[LineSource]

COMPLEMENTARY JFET FRONT-END, BIPOLAR OUTPUT

Hi Angel,

There is a thread "25 watt class A into 1 ohm" active that uses complementary JFET front-end and bipolar output for modest class A power into low impedances. You may want to simulate and give your comments on this design.

 

[Nelson Pass]

You could do all this stuff with bipolar, but since the current
gain of the average bipolar is limited, you usually find yourself
using a Bipolar Darlington arrangement instead of a MosFet.
Once you get the gain up, it works fine.

 

[angel]

Re: COMPLEMENTARY JFET FRONT-END, BIPOLAR OUTPUT

There is a thread "25 watt class A into 1 ohm" active that uses complementary JFET front-end and bipolar output for modest class A power into low impedances. You may want to simulate and give your comments on this design.

Had a look at it, and my initial thought was "is all this necessary".. It looks like it should measure pretty well, but I'm wondering if anyone has actually built it yet. Complicated circuits tend to yield complicated sonics.

I'll think about it anyway =)

 

[angel]

You could do all this stuff with bipolar, but since the current
gain of the average bipolar is limited, you usually find yourself
using a Bipolar Darlington arrangement instead of a MosFet.
Once you get the gain up, it works fine.

I was kind of thinking about using the 2SK389 for input, and a darlington with constant current loading, or a compound follower.

The darlington sort of defeats the entire point, though, which was to reduce losses due to Vgs/Vbe.

 

[Nelson Pass]

Vgs will be about 4.5V, and Vbe will be about 1.4V