A DIY amp. Opamp (TL071) pre-amp and LM386 power amp.

[brett]

After some experimenting, I arrived at an amp design that I quite like.

These days I live in an apartment, so low volume (70-80dB, aka 'very loud talking') is useful to me. Hence my interest in an LM386 amp.  It was hellish loud (90+dB, aka 'shouting'), so I put a power soak on it for 0.01W which is just right.  Who would have thought that a 10mW amp would sound so sweet?  Not me, even though I should understand how a log scale works.

Features include a pre-amp that (1) runs clean, (2) clips via the feedback loop (tubescreamer style) and (3) clips to ground (rat style).  Somehow, I got it to do all three quite well.  I think the 2.4k 'clip-to-ground' diode has a fair bit to do with it working quite well.  The 'clip-in-feedback' is quite crunchy, even though the gain is quite low.
Anyway... here's a schematic (and vero layout of the preamp), in three pieces.
You'll need to do your own wiring of the pots, switches and diodes.
All suggestions are most welcome.
Bretto

[antonis]

TL071 non-inverting input isn't biased at half-supply..!!
(or I need more coffee..)
Disconnect upper 1K resistor leg from +12V/22μF cap and connect it to GND while raising 100nF cap value..
(better make resistive divider values 10k, or so..)

I'd connect a 1M, or so, resistor between 1μF cap and GND for anti-pop purpose..

A 100nF ceramic cap (disk) across pins 7 & 4 (as close as physically posible) should "clear" IC supply..

Also, a 100pF cap across 100k feedback resistor should prevent HF oscillations..

And a 100R resistor to form a LPF with 22μF smoothing cap..

P.S.
What's the purpose of two caps (1μF & 22μF) between hard clipping diode pair.)?? 

edit: What I mean..

[Ben N]

Brett, I'm guessing you didn't feel the need to add any kind of tone stack or eq? Also, a 2p4T or 2P6T rotary and a suitable selection of resistors would give a range of power soak settings. BTW, what power rating did you use on the power soak Rs?

[Vivek]

Have you tried attenuating the input of the LM386 rather than power soaking the output ?

That could work if you have set up your LM386 to not distort.

[Ben N]

Have you tried attenuating the input of the LM386 rather than power soaking the output ?

That could work if you have set up your LM386 to not distort.

I'm guessing the idea was to overdrive the 386.

[Steben]

If overdriving the 386 is the target here, I'ld try asymmetrical clipping in the TL071 stage.