LM386 headphone amp suggestions

[onboard]

The GGG LM386 headphone amp sounds good and probably gets a "10" for simplicity. "Besides a volume control, how can I mod this for some flexability....?" The Headwize Headphone Amp and Pocket Rockit builds look great, but had plenty of parts that aren't in my drawer right now.

I wanted to
1)to put something together with what I have
2) keep the parts count low
3)try a Big Muff tone control I tweaked with the Duncan Tone Stack calculator (sweet utility!)
4)come up with something that can handle guitar or bass with the absolute minimum amount of value changes/switching

That last one is where I'm going in circles at this point. The best "catch-all" value for R1 seems to be 100k, the headphone speakers don't clip with guitar or bass, and both have a full sounding bottom. Switching R2 to 10k for guitar tames the high end without turning the volume down. With these values, the mid notch is plenty deep (maybe too deep!), although a tad wide. I can post the graph if anyone wants to see it.

Maybe the answer is more along the lines of a fixed value tone control, and a switchable low pass filter for use with guitar, or high pass for use with bass. At any rate, here's what I've got on the breadboard so far. Fire away!

(I'm using a mono output jack with a mono-to-stereo adaptor, not sure if the two 100R's on the stereo jack lugs would be needed)

edit: changed post title. Anyone have suggestions/critiques?

[onboard]

I keep fiddling with this and taking one step forward and two steps back. Here's the current changes (the schem's updated)
-R1 increased to 220k
-C3 changed to +1uF
-220R taken out between pin 6 and +9V, added +100uF to ground for housekeeping purposes
-might be good practice to include rf blocking, I'm picking up local sports talk radio when I touch R4 :roll:

Man, what a sweep with the tone pot. Nice and punchy. You could tweak a Big Muff tone control forever...

[thomas2]

cool! i've been thinking of adding the big muff tone control to my little gem combo  :lol:  little gem rocks! i also need a headphone amp so.. i don't have a radio, so rf interference would only be a positive thing  :twisted:

[smoguzbenjamin]

Remember, if the LM386 is giving you nasty distortion, add a resistor in series before the input, that will decrease the distortion but do almost nothing to the volume level

[onboard]

OT-----

i don't have a radio, so rf interference would only be a positive thing

There could actually be a use for deliberate RF interception in a pedal. Imagine mixing in AM talk radio for a cool lo-fi overdub effect. Maybe that's too twisted...  :twisted:

Remember, if the LM386 is giving you nasty distortion, add a resistor in series before the input

I don't know if the tiny amount of distortion is coming from the chip, or my headphone speakers. There's a very faint buzz as the notes decay, not the kind of distortion you get from high gain but just the opposite. Isn't it true that speakers can distort if they're *under*powered?

Being used for headphones, tone is more important than volume. I'm surprised at how quiet the chip is with this arrangement so far.

[smoguzbenjamin]

If speakers would distort while being underpowered, you could get distortion by turning down the volume, so that's not the problem... That buzzing could be gating somewhere in the IC. Try adding a series resistor and ezxperiment, alternately if you're using a resistor between pins 1 & 8, lower the gain a little. The IC should be able to give you all the power you need, even with lower gain.

[RDV]

The LM386 is incapable of 100% distortion free operation. You can minimize it, but it's just not a pristine clean amp chip. This is a dandy if you leave out the .0047 in the loop. http://headwize.com/projects/showfile.php?file=guitar_prj.htm

RDV

[onboard]

smoguzbenjamin- If speakers would distort while being underpowered, you could get distortion by turning down the volume

I totally agree, that's why this was odd. With pins 1 & 8 open, tinkering with a series resistor in front of the input helps the most. I'll poke around with the gain. Does gating in the chip have to do with biasing? Similar to misbiasing a transistor and getting a "splat"?

RDV - the Headwise build must then massage the chip enough to get it to run the cleanest it can? Working with headphones on is like putting the circuit's performance under a magnifying glass.    

Thanks you guys, everybody's willingness to explain is great. I'm still pretty mystified in the electronics department. But I'm too addicted to do anything else besides actualy learn what I'm doing...

[smoguzbenjamin]

Not so much biasing, more like that it is trying to amplify a sound that it can't. The LM386 self-biases the output to 1/2 the input voltage so no troubles with biasing. Try that current-reducing resistor in front of the IC and you should get pretty good results

[RLBJR65]

I have built quite a few (15 - 20) mini amps and half a dozen or so pedals based on the LM386. I also live 250 yards from an FM transmiter and tower so I'm always fighting RF noise.

They are NEVER totaly clean but their are a few things you can do that will help.

1st - I use a 2.2K resistor in line after the input cap a trick I learned recently thanks to someones sugestion on the forum.

2nd - This may seem odd (it did to me before I tried it) but try changing your input from the non inverting (+) to the inverting (-) pin of the chip then ground the non inverting pin. For some reason the LM386 just works better hooked up that way. It should help the distortion your getting as the signal drops off.

3rd Normal precautions short wires, good soldering and a metal enclosure to mount the circiut in.

Alot of these small amps like the Smokey amp were built using the JRC - NJM386 they are no longer in production but if you go to google and type NJM386 pdf you can get the data sheet it also has some amp schematics.

It's a mystery to me why the older chips work better with the input on the Non inverting side of the op amp and the newer LM386 the inverting side but that has been my experience.

I just posted a tremolo amp schematic using an LM386 on my web site. If your intereseted.

[David]

The LM386 is incapable of 100% distortion free operation. You can minimize it, but it's just not a pristine clean amp chip. This is a dandy if you leave out the .0047 in the loop. http://headwize.com/projects/showfile.php?file=guitar_prj.htm

RDV

I vote for this one too.  However, mine does seem "pristine clean".

[onboard]

Glancing back and forth between a few different schematics, I did notice that some were inverting while others non. And they were all for the same LM386 application... There was deffinately a noticeable difference between the two setups when I tried them both, I don't remember which was which right now.

I just looked at a couple of schems, the GGG EZ Heaphone Amp uses the inverting input, whereas the Headwize build uses the non-inverting. I think I went with the non-inverting and had improved results, just like you said.

My first go at it included a 1.5K resistor in series with the + input. I'll have to try it again and up it to the 2.2k and check it out. (Right now there's an LF353 3-band active eq taking up the whole breadboard...pretty sweet, though, and right out of the data sheet app. notes)

As for keeping short leads, my breadboard jumpers are probably waaaay too long. Every one of 'em is an antenna. :lol: