Mastroianni Single JFET Input Preamp

[lineup]

Attached schematic
shows a JFET preamplifier by Francois Mastroianni.

I would like to try it. Looks like it is very easy to build.

Input is one JFET 2N4392.
Single rail supply 35 Vdc .
Output is a single Class A follower 2N3252 operated at 15 mA.
No global feedback used.

My Questions:

1 - What modern JFET and bipolars should I use for best results?

2 - Any other modifications to improve performance?

3 - How would this preamplifier perform?

4 - What is major benefits of this circuit?

5 - Serious drawbacks?

[stoolpigeon]

Hi lineup,

Sorry I can't answer your questions but I wanted to ask if you have you tried this pre? If not any more thoughts on the design?

Thanks!

[lineup]

Originally posted by stoolpigeon
Hi lineup, Sorry I can't answer your questions, but I wanted to ask if you have you tried this pre? If not, any more thoughts on the design? Thanks!

stoolpig eon,
you are new in http://www.diyaudio.com/
Aren't you?
You will learn plenty of things, if you stay around for some years.

I have been here for more than 5 years, now.
We all have different levels of knowledge in this forum.
One day maybe I will be asking, to those being newbies now, some questions.
And they will know an answer to some issue I may have.


My question above
was more directed
to all our many experienced members
of our board.

Because An original Mastroianni Amplfier design is always something interesting
You can take my word for this, stoolpigeon
... as you, right now, can only have very limited knowhow in this very topic.


Regards
lineup
Lineup Audio Labs Ltd.
Europe

[Nordic]

guess something like j310 or maybe 2sk170?

[scott wurcer]

That 330uF non-polarized output cap might set you back a few euro.

[flg]

The input JFET needs to be 35V or more,maybe 2sk117?... And looks heavily degenerated? Gain might be modest... And while your there, you might think about making that regulator a JFET also... My first question would be, is the C1 necessary, but, after staring at the foreign language caption for a minute, I guess that's covered... No cap bypassing the 2 bias diodes??? A 330uF at the output seems quite large, generally you can get away with a very high quality cap of lower value, say <33uF, when driving 10K or more...

[analog_sa]

That 330uF non-polarized output cap might set you back a few euro

Not that there is any good reason for it being 330uF. Unless you're driving speakers.

[Nordic]

I tried simming it, and it seems to be unity gain or a little less... 840mv out for 1v in. Looks nice... and I might try it.

He probably meant a non-polar electrolytic, which could be had for little over a dollar.

[lineup]

Originally posted by Nordic

I tried simming it, and it seems to be unity gain or a little less... 840mv out for 1v in. Looks nice... and I might try it.

He probably meant a non-polar electrolytic, which could be had for little over a dollar.

Yes, it looks nice Nordic.
I have a bunch of TO-18, metall case, JFETs 2N4092
I don't know if I can use these here?
.... think they are not same type???

What would you use 2N4092 JFETs for??
They are higher currents than normal jfets.

======================

Nordic,
I think you have setup something wrong.
Gain below or around x1 is too low.
Check your input and output caps.

Gain is set by RD/RS in first stage + one output follower stage.
6.2k/1k would give gain like 5-6 for a bipolar input transistor.
For a JFET we get some Vgs dependent loss,
but like 3-4 maybe is a reasonable Gain figure - for JFET....

==================================

Two things, i like them both:

1. No global feedback
2. We may need no Input cap. Can put input directly across 100k resistor.
( at least if your source have only like +/- 0.200 Volt DC or less. )

A. JFET should be replaced with something that works good at 3mA
and can take the voltage of like 25-30V, at least
There are many we could have a go at!

B. For follower 2N3252 we should taget at 15 mA transistors.
BC338, BC337, BD135, BD139, BC639 or something like it.
Can be TO-92. But also of course TO-126

C. A lot of bypassing filter caps
and the use of Constant Voltage (=regulator) in input (most probably also 2N2219A)
and Constant Current Source in output (2N2219A)
... should make a fair and good power supply rejection ratio, PSRR


Regards
lineup

PS. might put this into my MultiSim for experimenting!
I am a bit busy with other things than Audio, right now
DS.

[Nordic]

Think Q1 is another 2n2219? I prepared a pcb file, but dropped a 2n2222 in there, not realy knowing what to put there...

I guess there is so much headroom, you could use lower rails , (with adaptation of resistors to maintain current and bias levels), even if you have to drop the gain. Then you can use lower voltage fets.

I was haveing a mindfart... I had to sim the circuit in two parts... too big for demo.