JFET Lineamp with Gain

[wrenchone]

The preamp sitting my home setup consists of JFET/MOSFET building blocks utilizing no global feedback .Threads that describe the varios functional blocks that have been in and out of the preamp over the past few years include "Open Loop Follies, Part 1", "JFET SRPP RIAA Amp", and "All American" Phono Preamp. Thus far, I have been quite satisfied with the listenability of the various circuits I've tried. For a line amp, I pretty much always used some sort of MOSFET or JFET-based follower. There have been times, however, that I could have used a little gain in the line stage.

Line amps are funny customers. Trying to build a simple JFET circuit that will do just a little bit of boost and satisfy requirements for drain voltage centering for max signal swing is difficult. In the tube world, one could just grab a low-mu triode and use that. One possible solution would be to borrow the concept of SY's Heretical Line Amp and drive a follower of some sort with a step-up transformer, or park the transformer at the follower output. I may try that eventually. One other solution would be to give up and allow some feedback. The circuit shown below does just that. It is designed around the single positive voltage I use to supply my preamp chassis. Simulation results are encouraging. I had tried another line amp with gain a couple of years ago, but ripped it out after only a few hours because it just didn't sound right (the change in sonic characteristics was very obviously unpleasant). I may give this one a shot, pairing it up with a fast/quiet shunt regulator. As shown, the gain is ~10X - a little hot for my taste, but still usable. With some juggling of values and input stage bias currents, I could probably get lower gain with similar distortion characteristics. We'll see.

[wrenchone]

I'm bumping this thread, as I'm actually going to build this circuit (or something close to it) and try it out. More later.

[qusp]

I look forward to it, buffers are one thing, but when you need a touch of voltage gain this type of thing is the go, I have been toying around with various Borbely articles and have plenty of ideas, but have not put anything to use yet. I have a bag full of toshiba signal jfets, some lovoltechs and a quad of SS R550 and some (OK heaps of) mosfets so i'm fairly well armed

[wrenchone]

Actually the Toshibas (at least the J74 and K170) don't look as if they're really optimal for the job here - more suitable for something like a gain of 50. I want to keep the gain for this line amp down to 10 or less. So far, the circuit in my "Liniac Revisited" thread seems more suitable for high amplitude drive than anything in this vein I've simulated just yet. I'll keep trying, though

[wrenchone]

I was able to do a circuit somewhat similar to the one first posted that gives a THD at 0.5V input almost the same as the circuit in my "Liniac Revisited thread (a tad less than 0.005%, with an attractive harmonic spread). Things are still in simulation-land - I'll post more details when I have a circuit that I like better. I think this one is going to be in the "advanced" catagory, just like the Liniac.

[WuYit]

wrenchone,
what a nice topology! I would modify it like this:

[wrenchone]

Why don't you continue that thought and post a real circuit with voltage and component values and distortion results. The body is a lot more than just the skeleton.

[WuYit]

wrenchone,
the CFP reduces the measured distortion. How much voltage gain do you need? Do you need any at all?

[wrenchone]

I'm looking for a voltage gain of ~5. I already have loads of circuits for unity gain that work very well - that's easy. A gain of ~50 is relatively easy. A gain of 5-10 with proper voltage centering and acceptable distortion at line levels with a single supply of +30V is another matter. My "Liniac Revisited" circuit meets the criteria and is listenable. I'm looking for other alternatives hoping for a simpler circuit, but the constraints I pose make the job difficult. When I'm satisfied with the circuit variant I mentioned here, I'll post.

[Steve Eddy]

Originally Posted by wrenchone

I'm looking for other alternatives hoping for a simpler circuit, but the constraints I pose make the job difficult.

Just go back to your first post:

One possible solution would be to borrow the concept of SY's Heretical Line Amp and drive a follower of some sort with a step-up transformer...

CineMag's CMMI-5C (1:5) and CMMI-10C (1:10) would be my recommendation.

se