Geting the most out of a LM386

[Voltzy]

Hello all. I have been playing around with all different kinds of distortion/fuzz circuits and I am enjoying with a design that uses a LM386 as a first gain stage (max gain with a 10uF connecting the gain pins). So far I have my input signal have a 1M shunt to ground, then through a 10uF cap, then into the LM386 non-inverting input.

I am trying to figure out how to build the best environment for the LM386 to work with. I am planning on building a power filtering circuit similar to the ProCo RAT and I know that this particular opamp doesn't need the input biased. I am wondering if it is a good idea to have a unity-gain input buffer before the LM386 and whether this would improve the operation of the op-amp. I was looking at the Klon Centaur circuit and was considering using one half of a TL072 for the active EQ at the end of the circuit, would the other half be suitable for an input buffer?

I understand it will output a low impedance signal, in the Klon Centaur this goes into another TL072 which has a ridiculously high input resistance, where as the LM386 has an input resistance of 50k so I am unsure if it would be a good idea. I have also seen some people suggest a J201 buffer into a LM386.

[diffeq]

50k is a rather low, input buffer would prevent tone sucking when connected directly to the guitar pickup.

[Voltzy]

OK thank you, I have put together an input buffer like the one suggested for the 1W amp at electrosmash:

https://www.electrosmash.com/images/tech/1wamp/1wamp-input-stage.png

I have put it together on the breadboard and so far I am getting no output from the pedal, but I have also had not time to play around and see what the problem is or what I have done wrong so I am not worried about that, but if this was working would this provide a better signal to the LM386?

[diffeq]

if this was working would this provide a better signal to the LM386?

Generally, yes.

A few notes about that particular buffer:
1. TP1 point is input for bias voltage*  that sets the gate somewhere in the middle between Vcc and Ground. If you omit that entirely, transistor will remain shut. Exact voltage depends on JFET itself.
2. No coupling capacitor on the input (bias will flow into whatever is connected to it)
3. No coupling capacitor on the output (will conflict with LM386 internal bias circuitry)

Does it have to be a JFET buffer? Could do with TL071/TL072 opamp.

* usually through large resistance to keep input impedance high

[GibsonGM]

Yup, what Tony said.  You can do this more simply without the need for the fancy zener diodes.  An opamp would be very simple. 
Look:  http://www.generalguitargadgets.com/effects-projects/boosters/buffer-ic/   

R1/R2 are your bias network to put the input at about 1/2 supply.   

[Voltzy]

Does it have to be a JFET buffer? Could do with TL071/TL072 opamp.

I have got a cap on the input but I don't have one on the output of the buffer, I will try it out. It doesn't have to be a FET was asking what the differences would be, that was the first LM386 input buffer I found online so thats the first I tried.

Would there be any audio benefit between using a JFET or an Opamp? I have some 741s and 358s lying around I can try out.

Yup, what Tony said.  You can do this more simply without the need for the fancy zener diodes.  An opamp would be very simple. 
Look:  http://www.generalguitargadgets.com/effects-projects/boosters/buffer-ic/   

R1/R2 are your bias network to put the input at about 1/2 supply.   

Awesome thank you I will give this a go. If I get the JFET buffer working as well then I might try to do some comparison samples.

[diffeq]

Well, they do sound different. It's subjective so you'll have to choose for yourself. Here is a relevant article: http://www.muzique.com/lab/buffers.htm

741 and 358 are a bit too arcane for the job, but if you want to test it, start with lm741:


(other jellybean opamps to try: TL071, NE5534, RC4558 (which is dual, so pinout will be different))

[Voltzy]

Well, they do sound different. It's subjective so you'll have to choose for yourself. Here is a relevant article: http://www.muzique.com/lab/buffers.htm

741 and 358 are a bit too arcane for the job, but if you want to test it, start with lm741:


(other jellybean opamps to try: TL071, NE5534, RC4558 (which is dual, so pinout will be different))

That is a really great resource thank you, I have built the 741 buffer and it works however it was late at night and I haven't had a chance to really listen to it.

A few notes about that particular buffer:
1. TP1 point is input for bias voltage*  that sets the gate somewhere in the middle between Vcc and Ground. If you omit that entirely, transistor will remain shut. Exact voltage depends on JFET itself.

I actually got this JFET buffer working by adding a 10uF cap between the buffer output and the LM386 input, and it sounds great, and then I re-read your comment here and understood it. You are saying that there needs to be a resistor at the TP1 point that has the other end connected to +9V (making a voltage divider with R1) which will boost the clean input to the JFET?

If that is the case then I don't know how the JFET is working at the moment because the only resistors I have connected are R1, R2, and R3 in the diagram. Shouldn't the JFET be shut?

[diffeq]

I actually got this JFET buffer working by adding a 10uF cap between the buffer output and the LM386 input, and it sounds great, and then I re-read your comment here and understood it. You are saying that there needs to be a resistor at the TP1 point that has the other end connected to +9V (making a voltage divider with R1) which will boost the clean input to the JFET?

If that is the case then I don't know how the JFET is working at the moment because the only resistors I have connected are R1, R2, and R3 in the diagram. Shouldn't the JFET be shut?

For some reason I assumed that N-channel JFETs are biased exactly like BJTs, but I was wrong! After some reading, turned out JFETs gate needs to be reverse-biased (relative to source), which is exactly what happens in your schematic.

A really good article on the subject: http://diy.smallbearelec.com/HowTos/BreadboardBareAss/BreadboardBareAss.htm