LM386 audio amplifier

[jim tylee]

Having used the LM386 amplifier ic a number of times to get around output stage non operational it is a ic that gets very warm in use ..too hot to keep a finger on but they never seem to fail even at high temperature use one in my R209 military receiver no problems even at full volume...
JIM G4REK

[Restoration73]

I recommend 1nF capacitor pin 1 to ground to prevent RF getting in or oscillation.
It's great for a 30 year old chip, but there are better new alternatives:
http://diodes.com/datasheets/PAM8403.pdf

That will be fine, as you have found out. For comparison, too hot for a finger is about 60..70 centigrade, silicon works up to at least 125C. If you assume an ambient of 30C (about right for inside a cabinet) 70C has a delta of 40C, 125C the delta is 95C, nearly three times as much.

[paulsherwin]

The PAM8403 is a very different chip. The LM386 is indeed a very useful IC if you just want to replace the audio stages of a mono radio and don't need a lot of output. They shouldn't get obviously hot unless you are pushing the ratings, but even then they're pretty tough.

[MrElectronicman]

If the LM386 is getting hot then it is most likely that it is oscillating. I have had the same problem and you need to pay close attention to the arrangements of the other components around the IC

[Restoration73]

correction re # 2 - should read "capacitor pin 3 to ground"
pins 1 & 8 are used for gain setting

[ThePillenwerfer]

I'm a huge lover of the LM386 but those PAM8403s look well worth investigating.

I don't do much in stereo but one of those is only about the price of two LM386s and are available as ready-built modules with volume control, which is another saving. Then there's the extra power and ability to work down to 3V.

They seem too good to be true but worth risking a quid on.

- Joe

[jim tylee]

Originally Posted by MrElectronicman

If the LM386 is getting hot then it is most likely that it is oscillating.

No oscillation at all scope on the output...they just get hot even when lightly driven
jim

[PE9ZZ_JO22KI]

That's weird. Are the DC conditions on its pins OK? What's the current drawn from your supply? From the datasheet this should be 4 mA quiescent.

Tjerk, 9ZZ

[jim tylee]

Yes thats correct 12.3 volt supply...quiescent current 5 milliamps......
190 milliamps drawn on a moderate input signal .....hot hot

[Ian - G4JQT]

Something is not right. They are designed for battery operation, so this much heat isn't good for battery work! I've often used them and mine don't get hot. 12V is at the top of the supply range for the LM386N-1, -3 but that should be fine.

Your quiescent current 5 milliamps looks fine.

Have you a zobel network 10R/0.047u pin 5 to earth?
Bypass approx 10u pin 7 to earth? (Not usually needed.)

You say no instability on the scope, so not sure what's wrong.

Anyway, here's the data sheet if you've not seen it.

Ian

[PE9ZZ_JO22KI]

With an input power of 2 W and an 8 ohm load it should produce 0.8 W of audio power (datasheet). That's a lot of noise. With only "moderate input signal" the same input power means something's very wrong. Maybe your load is too low? Or even short circuit? This is a very simple circuit so finding your fault should not be hard. Your scope should tell all.

Tjerk, 9ZZ

[G6Tanuki]

I'd check the load-impedance and also that the electrolytic between the chip and the speaker is polarised correctly!

I've seen a couple of similar 'homebuilt' amps where the capacitor's been put in the wrong way round, so causing a continuous leakage current. The biasing cct in the chip pushes more and more current into the output-pair to try and maintain the centre-point at mid-rail voltage and gets hot as a result.

Personally I've found the LM386 to be a rather 'hissy' chip - my preferred low-budget audio-output device is the now somewhat obsolete SGS-ATES TBA820