LM386 Issues

[jrem]

Folks:  I've been looking around but haven't found any help so I thought I would post here.  I have an LM386 that hits the rails hard.  Any ideas?


The circuit is just an input cap (.005uf) and an output cap (1000uf), have tried other combos, same thing. 

Thanks in advance, John.

[jrem]

http://www.jrem.com/062112c.jpg

[R.G.]

The LM386 datasheet says that when set up like that, it will self bias at half the available power supply voltage, and pull its inputs to the most-negative voltage with 50K internally on each input.

In that configuration, it has a gain of 20x. You don't say what power supply voltage you're running from nor what load resistance you're driving with it; if it's 9Vdc, it can only drive the output to a peak-to-peak voltage of 6V for an 8 ohm load, and 3.5V for a 4 ohm load. It only goes to 8V peak to peak with an open circuit for a load.

Say it's 8 ohms, as a guess. The biggest input it can accept without clipping is 6V/20 = 0.3V peak to peak, which is 150mV peak, and that's only slightly larger than a typical single coil pickup puts out.

There could be other stuff wrong, but it could also be working as expected.

Have you read "Debugging; what to do when it doesn't work" ?

[jrem]

Hmm, thanks RG.  I'm building a small stereo amp and I figured it might be something like that.  I'll check my input signal voltage, I probably have to pad it.

[Jdansti]

BTW- After you finish this build, you might want to try the TDA2822M. I just built a stereo amp for my iPhone using the following schem and it is super clean.  Very low parts count and extremely easy to build.   It puts out 1W per channel. The only mod I made was I had to put a 220k resistor in series with each channel's input to tame the iPhone signal. It runs on any voltage from 1.8-15VDC.

(RL=speaker)

[Earthscum]

BTW- After you finish this build, you might want to try the TDA2822M. I just built a stereo amp for my iPhone using the following schem and it is super clean.  Very low parts count and extremely easy to build.   It puts out 1W per channel. The only mod I made was I had to put a 220k resistor in series with each channel's input to tame the iPhone signal. It runs on any voltage from 1.8-15VDC.

I second that one! They drive headphones beautifully, and speakers are great. The distortion isn't too much unlike the 386 when it's overdriven. Once you get the hang of playing with them, you can tailor the circuit to cut out some highs and kind of sim a speaker cab sound for headphones.

I got bored one day and started playing around with one for distortion sounds and posted a couple schems of distortions I liked HERE. One of the things I ran across was possibly using one of the outputs to drive for a compressor... I'll have to bust one out again and see if I can get it to work. These things have a lot of potential to be explored. It takes a little to understand how the inputs work for other operations (it's all in the AC coupled current to 5 and , but once you get the hang of it, you can tailor this thing to do all sorts of pony tricks.

[Jdansti]

BTW- After you finish this build, you might want to try the TDA2822M. I just built a stereo amp for my iPhone using the following schem and it is super clean.  Very low parts count and extremely easy to build.   It puts out 1W per channel. The only mod I made was I had to put a 220k resistor in series with each channel's input to tame the iPhone signal. It runs on any voltage from 1.8-15VDC.

I second that one! They drive headphones beautifully, and speakers are great. The distortion isn't too much unlike the 386 when it's overdriven. Once you get the hang of playing with them, you can tailor the circuit to cut out some highs and kind of sim a speaker cab sound for headphones.

I got bored one day and started playing around with one for distortion sounds and posted a couple schems of distortions I liked HERE. One of the things I ran across was possibly using one of the outputs to drive for a compressor... I'll have to bust one out again and see if I can get it to work. These things have a lot of potential to be explored. It takes a little to understand how the inputs work for other operations (it's all in the AC coupled current to 5 and , but once you get the hang of it, you can tailor this thing to do all sorts of pony tricks.

Thanks for the info.  I've got one or two that I'll have to try that with. Does the "last" resistor and pot keep the output low enough to go into an amp instead of speakers?

[jrem]

Awesome guys, thanks. Just what I was looking for!

[tca]

I got bored one day and started playing around with one for distortion sounds and posted a couple schems of distortions I liked HERE. One of the things I ran across was possibly using one of the outputs to drive for a compressor... I'll have to bust one out again and see if I can get it to work. These things have a lot of potential to be explored. It takes a little to understand how the inputs work for other operations (it's all in the AC coupled current to 5 and , but once you get the hang of it, you can tailor this thing to do all sorts of pony tricks.

Did you try the Bridge version of the circuit? I like the idea not having an output cap.

[Earthscum]

@John: The 1k in the octave up and the 100R in the distortion keep the output from sinking a gigantic amount of current to whatever comes after, including plugging the cable in and such. If you don't watch the outputs of these things, you can deal a deathblow to a feeble power supply pretty quickly, lol.

@tca: The distortion is cascaded, the octave up one is bridged. You would still want an output cap because, as I mentioned, these things will sink some current if you aren't careful. An accidental 100R to ground about made for a good heater element.

[Jdansti]

@John: The 1k in the octave up and the 100R in the distortion keep the output from sinking a gigantic amount of current to whatever comes after, including plugging the cable in and such. If you don't watch the outputs of these things, you can deal a deathblow to a feeble power supply pretty quickly, lol.

@tca: The distortion is cascaded, the octave up one is bridged. You would still want an output cap because, as I mentioned, these things will sink some current if you aren't careful. An accidental 100R to ground about made for a good heater element.

A 100R/1/4W resistor to ground at the bridged output would see 2W, correct?

I bought a couple of the 10W x 2 chips. I plan on bridging one of them for a 20W guitar amp.