JLH Liniac Revisited

[AndrewT]

Which part of the circuit makes it "Liniac"?
And what is it doing to "improve/change" the operation?

[wrenchone]

Read the posts - I'm now calling it "Liniac Inspired", as I'm sick and tired of all the quibbling that really has nothing to do with the circuit. The idea I stole from the Liniac articles (besides the basic feedback pair) was stiffening up the drain of the input device with a current source. This is all explained in the thread.

The problem this particular circuit addresses is to get a low gain buffer (gain <10) with good output centering and low distortion . A similar circuit with a current source at the source of the input device makes a heck of a unity gain buffer, and I used said buffer until I came up with this circuit. The first incarnation of this circuit is in my living room system. The second one will be possibly getting some measurements with my 8903A this weekend.

I could possibly ( make that certainly) do a circuit using
2SK170 and 2SJ74 that would work as well or better than this circuit, but the very purpose of this exercise was to be totally stubborn and eccentric and avoid doing that. Also, I was intrigued at the possibility of using one of the Supertex devices in a gain context, as nobody else seems to be using them for anything besides current sources. I used the TO-92 flavor in the circuit I just presented.

[AndrewT]

Found much of it in post59.

Judging other comments in this Thread, you may have thought I was having a dig at you.
That is not the case.

I simply wanted to know what made a "liniac", it seems the CCS load on the first and the second Darlington make it such.

[wrenchone]

This little fit of pique on my part had one good outcome - I actually went and simulated a 2SK170/2SJ74 lineamp using small-signal Darlington outputs. The circuit is a mirrored complementary version of the basic skeleton of the one I presented here, minus the extra current sources and doodads. It took <1/2 hour to get something that balanced and simulated well. I strongly suspect that the circuit or something close to it has already been invented by someone else years ago (probably JC or Erno) - it was too easy to be otherwise.

Just to be perverse, I also ginned up a version using the J176 and J113 for input jfets, with mosfet outputs (Zetex small-signal devices). The J176 and J113 need some small helper current sources to perform at their best, which is about twice the THD of the fancy-pants Japanese solution. This is not a bad place to be for "Joe Blow" fets like the ones I used.

If someone asks me nicely, I'll post what I did in a separate thread.

These symmetric circuits also present the possibility of using an opamp servo to ditch the output capacitor. Despite violent opinion to the contrary, I use an input cap on everything, as you can't trust anything in the outside world. I have too many commercial products with DC offset on their outputs (one would be too many). I suppose I could fix them with an output cap or a little more hacking, but I'd rather spend my time building new circuits.

I did a little casual searching, and both JLH and Hiraga have some similar circuits. The JLH circuit is a unity gain buffer, and the Hiraga circuit is "Le Monstre". Hiraga cascodes his input jfets, something I was tempted to do. Maybe later...

[mad_z]

Originally Posted by wrenchone

Just to be perverse, I also ginned up a version using the J176 and J113 for input jfets, with mosfet outputs (Zetex small-signal devices). The J176 and J113 need some small helper current sources to perform at their best, which is about twice the THD of the fancy-pants Japanese solution. This is not a bad place to be for "Joe Blow" fets like the ones I used.

If someone asks me nicely, I'll post what I did in a separate thread.

Wrench... I would be more than grateful if you post it, I am following your threads with biggest interest for some time. Stashing the fets in the mean time...

Thanks in advance!
mad_z