How can I determine if my LM386N amplifier chip is defective?

I'm following the National Semiconductor datasheet for an approximate 200X gain (very minor modifications), but I get no gain at all. I do get sound though, which makes me believe the chip may be working just fine. The reason I know I have no gain is if I hook the audio out directly into the speaker I get the same sound volume/quality. I'm using a 9V battery for power, a 10 microfarad capacitor to connect pins 1 and 8, and a 220 microfarad capacitor between the 5 pin and the speaker positive input. Here are some differences in the modification:

1.) I'm using a 4 Ohm speaker and not an 8 Ohm speaker, but from what I understand that shouldn't matter since this is just a matter of less resistance and not more.
2.) I don't have a bypass or a .05 microfarad coming out from pin 5.
3.) I'm using a standard RCA cable from my computer headphone jack to test, and clipping a 20 gauge wire to each to run into the breadboard.

I'm really new to all this, and I've tried like 3 or 4 different configurations, but they all generate the same sound volume/quality. Any suggestions would be greatly appreciated.

sort by: active | newest | oldest
Check your power rails, measure the voltage on the supply pins, and don't trust the breadboard.

I'd like to see what you think you have wired before commenting any further.

giantjamsandwich (author)  steveastrouk6 years ago
When you say supply pins, do you mean the battery? It is a 9V, but when I tested with my multimeter it is showing 7V.
No, the ones on the chip itself. Breadboards shouldn't be trusted to make good connections.
Agreed, when troubleshooting a circuit, imo and ime, regardless of whether its on a breadboard, a pegboard, or a true PCB, one should measure the power supply voltage at the power supply pins of the IC. It's the only way to be sure you actually have the juice at the place where it's needed. And by that I mean measuring by touching the power supply pins of the IC with the DMM or Scope probes directly, not just the wires or copper leading into it.
giantjamsandwich (author)  seandogue6 years ago
I had this working to some degree before, but now I can't get any signal, and I'm wondering if I shorted out the chip somehow. seandogue or steveastrouk, if I touch the multimeter leads to pins 3 and 5 that should give some sort of voltage reading, correct? Thanks again for your help!

Here is the datasheet for reference:

a measurement between pins 3 and 4 should show the input voltage. Without a scope, you'll have to use DC values, since AC on a DMM (digital multimeter) is based on a pure sine wave and is an RMS measurement. that is, it's not the peak to peak value but a fraction thereof.

A measurement between pins 5 and 4 will show the output voltage * with a caveat...

Note from the datasheet
"The inputs are ground referenced while the output automatically
biases to one-half the supply voltage"

that means that 0 VDC in will result in 2.5 vdc out for a 5V supply.

And between pins 6 and 4, the power supply voltage (as measured AT the IC)

If you are able, I'd input a 0.1 VDC signal into the input, then measure the output. with no external gain resistor, for a 0.1V input, you should see a 4.5 V output. (the ncie thing about this amp is that you can input use a unipolar supply and still input negative input voltages...for instance...audio signals...

{well technically, since you're using a 9V battery for power, that would be half the supply (4.5V) plus the multiplied input (.1V x 20 = 2V) or 6.5V}

Now.........here's where things get a bit iffy. You said you have no bypass cap. I haven't used the LM386 in quite a while, and honestly, although I do recall having used it, I can't remember what for.

Having said that, my suspicion is that since the supply is unipolar (5V and GND, 10V and GND, etc), you need a bypass cap on the output to block the DC component noted above and in the datasheet. You really don't want to push DC onto a speaker, since it in general wants to see AC only.

It IS possible that you smoked the output stage by applying the DC component to a 4ohm load. However, try tacking a .1 or 1uF ceramic cap between the output and your speaker wire (the positive one I mean) and see if things don't improve before you hang your spurs and buy a new amp.

If you did, well, lesson learned. erm..stuff happens. I've blown quit a few chips over the years, many of which cost far more than an LM386...try blowing a 100W power opamp...not nice when you have to go to the boss and tell him you need another. ;-)
giantjamsandwich (author)  seandogue6 years ago
Thanks for the great feedback! Other than knowing how to check the pins based upon what you wrote it will take me a little bit to dissect everything else. Like I mentioned, I'm new at this, so I need time to process :)
Understood. that's kinda why I wrote it like I did, hopefully not to confuse but to educate. take it slow if you need, and if what I wrote confuses, let me know and I'll try to clarify the confusion point(s)
giantjamsandwich (author)  seandogue6 years ago
Looks like the amplifier is still passing through power, so now I just need to try to design the circuit on the breadboard again with the cap included, and see if that works. Thanks again!
Hey, np.

BTW, check a few pages into the datasheet (I think they start on page three or four) to see the reference to the cap, you'll see the schematics they suggest for use as speaker drivers. Illuminating...

best wishes
giantjamsandwich (author)  seandogue6 years ago
Alright, so I finally got around to trying the simplest configuration and taking some pictures. You can see I'm trying to follow the diagram for a 20 gain, leaving out the cap altogether. That is a 220 microfarad capacitor. You can see that pins 2 and 4 are grounded, 3 is the power supply from a 9V battery, 5 goes to the positive lead on the capacitor, and 6 is the incoming signal from the computer. The negative lead from the capacitor goes to the positive on the speaker. I know, probably horribly wrong, but I'm trying an almost literal interpretation of the diagram.
A literal interpretation of the nat semi schematic is the best way to duplicate their suggested circuit.
giantjamsandwich (author)  seandogue6 years ago
Indeed you are correct. Until I put the cap in place, it would not work, and even if I put a ground with no resistor and cap in place it works. It just needed that pin to be grounded. Thanks again for all of your help!!!
You're very welcome. I'm glad that you got it sorted out :)
Whoa, just checked the data sheet again. you'll need a far larger bypass cap on the output...250uF is what they recommend.