JLH Liniac Revisited

[wrenchone]

Circuit, please- I want to compare operating scenarios. What is good in one situation may nor be so good in another, usw... I'm building up the circuit I simulated (with some component differences) and intend to sub it into my living room test bed. So far, the discrete single-ended lineamps with gain I've tried have been less than satisfying from a listening standpoint. I'd like to break that particular jinx.

[Joachim Gerhard]

would that work ? does it make sense ?

[Joachim Gerhard]

values are only default so i did not put it in simulation, too busy, sorry

[wrenchone]

My Liniac is almost built. A couple more solder joints and I can start bringing it up. It has an on-board high speed shunt regulator.

[revintage]

Hey Wrenchome,

I´m sure your circuit OK. But where is the Liniac topology?? The Liniac was a inverting one-stage configuration with FET-CCS and emitterfollower. I built it -71 when it was presented and liked it. Hadn´t so much to compare with though.

Changing the BJT to a cascoded JFET is not a bad idea. But changing the emitterfollower to a gainstage makes it not being a Liniac anymore. Below is the original article.

http://anacon-tech.com/IPdocs/JLH_LINIAC.pdf

[wrenchone]

It's not strictly a liniac, but one of the direct variants he came up with. I got the bare bones of the circuit out of the Paul Kemble compendium of Hood circuits at A Paul Kemble web page - John Linsley Hood preamp designs. . If you want to be a stickler about it, I could call it a "Hood Feedback Pair", but the Liniac was definitely its father.

[wrenchone]

Puppy is up and running after fixing some bonehead stuffing errors. Some compensation is required for optimal square wave response. Without compensation, there is damped undershoot with 10kHz square wave excitation. A 220 pF compensation cap clears this right up. Either I'll settle for an NPO cap, or go scrabbling around in my assorted polystyrenes for something that will do the trick. I also want to try driving a 6ft output cable terminated by a 10k resistor to see if a modest amount of capacitance can be tolerated without the electronic equivalent of a hissy fit...

[wrenchone]

Attached is a picture of the line amp so far. The LEDs are the reference for the on-board discrete shunt rugulator. Schematic will follow after final tweak-up.

[revintage]

I might be a stickler about the nomenclature but the only thing in common with the Liniac is that it was last in the same article. Think Kemble mentioned that to.
The nice thing about the Liniac was that it only had one gainstage making it act like a triode when feedbacked. Lets leave it by that. Have you vere tried the JLH 10W? It was the first transistor amp I built back in -69 and I still remember the sweet sound of it.

You HF problems might be due to the high gain of the two cascaded gainstages. Hope you fix it and that it sounds!

[wrenchone]

Anyway whatever you call it, here is its respose to square wave excitation. The smaller waveform is the input, the larger in amplitude is the output. The leading edge is well-behaved. Remember that this is a single-ended amp so that the leading and trailing edges will have different slew rates. Also, since this is a feedback system, a little bit of high frequency compensation is not at all unheard of. I'm using 130 pF, which leaves the trailing edge a tiny bit undercompensated, but it's something I can live with for the time being. I want to see what it looks like driving a 6ft output cable before I make any final decisions on compensation.