My A.N.T. "Amber" headphone amplifier

[x-pro]

As I plan to release an updated version soon, here is a complete circuit of my SE Class A and "zero negative feedback" headphone amplifier design from 2004, using my idea I've published here:

Non-feedback parametric linearisation in FETs

It works well with most medium to high impedance headphones (from 60 to 600 Ohm). For low impedance headphones the output FET could be replaced by a device with a higher transconductance at 100 mA.

This circuit and a short description also now available as a PDF file on my website here: http://www.ant-audio.co.uk /Theory/ANT Amber Headphone Amplifier.pdf . As usual - free for a private builder, not for production. If somebody would like to build a copy, I would be happy to explain in this thread how to select the FETs for it and what replacement parts could be used.

Enjoy!

Alex

[dimitri]

Hello Alex,
Glad to see one more implementation of Richard Knapp (Spatial Audio) patents, dated 1985 and earlier. Probably it will be nice to show some appreciation while "borrowing" his ideas.

[x-pro]

Originally posted by dimitri
Hello Alex,
Glad to see one more implementation of Richard Knapp (Spatial Audio) patents, dated 1985 and earlier. Probably it will be nice to show some appreciation while "borrowing" his ideas.

First of all, Dimitri, I did not "borrow" anybody's ideas. If I know about the previous work I would always mention it. However in this particular case I can see no relevance to my circuit at all, except some similarity in configuration. There is no capacitor in my circuit . Q5 and Q7 in my circuit only provide for a current source load, temperature compensated to match the amplifing pair below. And the cascoded combination of two FETs was known well before 1985 and described in "The Art of Electronics".

Secondly, I really don't like your tone. Sorry.

Alex

[Tim__x]

There's a NatSemi app note somewhere (I'm not sure of the date) with a circuit similar to the on Dimitri posted, they called it a mu-follower. I can't recall if they used the Greek U-like mu, or just "mu".

[Gordy]

Thanks for the post and the info Alex.

[Tim__x]

A quick google shows that it not a NatSemi invention, but dates back to tubes, and is certainly older than any patent from the 80's.

[x-pro]

Originally posted by Tim__x
A quick google shows that it not a NatSemi invention, but dates back to tubes, and is certainly older than any patent from the 80's.

Yes, the "mu-follower" is well known for the tube circuits and there is a certain connection between that and the circuit Dimitri posted. However, as I've pointed out already, my circuit has nothing to do with either of these . In my case the cascode is adjusted for a minimum THD and the top pair of the devices is just a current source, nothing more. The only reason to use a same matched combination of devices was a near-perfect temperature compensation. Actual load in the VAS is R15 and the gain could easily be made several times higher if required by just increasing value of R15. VAS here is a complete DC amplifier, with no capacitors in the signal path and no LF roll-off. The LF limit is set by the output capacitor and the headphone impedance.

Cheers

Alex

[Tim__x]

I never meant to imply your cicruit was "borrowed".
Just an off-topic comment on the patented circuit mentioned not being very original.

I see yours as nothing more than a cascode (not that there's anything wrong with 'just a cascode').

At this point, there are very few functional arrangments of three terminal devices that have not been done before.

[x-pro]

Originally posted by Tim__x
At this point, there are very few functional arrangments of three terminal devices that have not been done before.

True . In my circuit it is not the arrangement by itself but a combination of parameters and the working point that is important to get a much improved linearity without the feedback.

Cheers

Alex