LM386 chip get hot on audio amp circuit

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

I just Breadboarded a LM386 audio amp circuit using the schematic below.
It works ok, but the LM386 chip get very hot, so hot i can't touch it.
I'm only using a 9v battery, so I'm not sure why it's getting so hot.

I'm using it to power a 6 ohm 2.5w speaker.
My input is a 4093 nand gate and also a 40106, schematics here:
http://vidiotsquad.com/?page_id=10

my_low_power_amp.gif

Description:

Filesize:

4.41 KB

Viewed:

29694 Time(s)

Inventor · Posted: Sat Sep 05, 2009 8:29 pm Post subject:

I don't know but I think that the minimum circuit configuration for an LM386 has an RC shunt from the output node for some reason. Perhaps you need to add that. Also the Power Output specification for the chip is 125 mW, not 2.5W! Zounds! You might be toasting that little puppy. Datasheet attached for convenience.

Les

LM386.pdf

Description:

LM386 Datasheet

Download (listen)

Filename:

LM386.pdf

Filesize:

264.42 KB

Downloaded:

331 Time(s)

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

Actually its 250mw, but damn! your right, I didn't even consider that.
I had no idea using 2.5w speaker would make the chip run hot.
So I'm off to find a another speaker, I Think Sparkfun has some.

BTW, what would be a good chip to use for 2.5w speakers?

Inventor · Posted: Sat Sep 05, 2009 8:56 pm Post subject:

I don't know, I would cruise the www.national.com site for audio amplifiers. I do not know much about them, but surely you will need a heatsink of some sort due to the maximum power condition which states that at best the source must dissipate as much as the load. Maybe someone with more experience will chime in soon.

Les

Thomas_Henry · Posted: Sun Sep 06, 2009 1:17 am Post subject:

Hello,

Inventor · Posted: Sun Sep 06, 2009 3:07 am Post subject:

I stand corrected. Thomas_Henry seems to know what he is talking about. I'd check those higher frequency oscillations, and look into the RC shunt in the datasheet, maybe that's what it's for?

Les

RF · Posted: Sun Sep 06, 2009 8:03 am Post subject:

Hi acidblue -
I've used this circuit many times - it's right out of the applications handbook....
Never gets hot, always works well...

bruce

lm386_20_956.gif

Description:

Filesize:

10.12 KB

Viewed:

29615 Time(s)

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

Thanks Tom for clearing things up, yippie! I don't have to buy more speakers.
BTW I don't have a scope, it'a on my xmas list.

MMM. RF my schematic dosn't have the cap and resistor going to ground on pin 5.
I wonder if that's the reason im getting hot chips.
I'll adjust my breadboard circuit and give it a go.

I'll get back with the results.

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

Ok I'm still getting a hot LM386 chip.
I've included pics of my breadboard circuit.

I didn't have a 10 ohm R nor a 0.05uF cap, so I used 330 R and a 0.1uF cap.
I don't think that will make too much a difference.

The larger chip is the 4093.

Birds eye view LM386 amp.JPG

Description:

Filesize:

68.34 KB

Viewed:

313 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

Close Side shot LM386.JPG

Description:

Filesize:

58.42 KB

Viewed:

303 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

Inventor · Posted: Sun Sep 06, 2009 6:49 pm Post subject:

Cool man, detailed photos and stuff. I would imagine that your component values are not appropriate for the application. If it were me, I'd get the right parts, even if I had to go to Radio Shack for some quickie parts. Also you probably don't want little parts for this, especially the resistor. I'd get a 1/2 watt or bigger resistor if they have it. That's because your speaker is 6 ohms and you're using a 10 Ohm resistor sort of in parallel, so you don't want to toast a 1/4 Watt resistor.

Those are just my inexperienced thoughts, I'll let wiser folks more in the know critique them for you. Best of luck with it.

Les

DGTom · Posted: Sun Sep 06, 2009 6:52 pm Post subject:

I might be wrong.. but to me it looks as tho the resistor on the right with the little cap won't actually be doing anything will it?

State Machine · Posted: Sun Sep 06, 2009 6:57 pm Post subject:

No doubt, the thing is oscillating way beyond audible levels as Thomas mentioned a few posts back. I would hang a 0.1 uF capacitor, AS CLOSE TO THE CHIP AS YOU CAN GET, from Vcc to ground. That will clean things up and it will most likely run cool as a cucumber. lol ........ would not hurt to also put a 10 uF bypass capacitor near the 0.1 uF also. Wink

Bill

Inventor · Posted: Sun Sep 06, 2009 7:04 pm Post subject:

Duh, I didn't even think of that Bill, nice work. I suspect that the 10uF cap is going to be the thing that's really needed in this application and the 0.1uF only makes it better, but what do I know? Probably too much for my own good, lol.

Les

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

magman · Joined: Feb 04, 2009 Posts: 363 Location: Liverpool, UK

As DGTom stated (and looking at the large Cap as well), these components are actually being shorted out by the connections on the breadboard (the 5 connections in a row are shorted together - that's how these breadboards work).

You will need to make these connections in adjacent rows to give them separate connections.

For example, move the negative (short) leg of the electrolytic cap from from I49, to I50, also move the output connection from J49 to J50. Then move the long leg of the resistor from H52 to H51 and the capacitor from I52 to I51.

Try this first then lets see how the circuit behaves, but the decoupling cap already mentioned would then be a good addition.

I suspect the reason the chip is getting hot is that you have effectively put a low resistance to ground on the output (the speaker), because the output capacitor is being shorted. Here's hoping this hasn't cooked the chip.

Regards

Magman

aerogramma · Posted: Sun Sep 06, 2009 11:49 pm Post subject:

Inventor · Posted: Mon Sep 07, 2009 12:43 am Post subject:

Oh so if I get this right, you folks took a close look at the photos and figured out that there was a wiring error? I think I see it too, on pin 5 of the LM386, right? How cool is that? DIY help debugging by photograph now, nice! This could be a new angle on helping each other here on the forum. Your circuit don't work? Take a photo of it. Neat.

Les

State Machine · Posted: Mon Sep 07, 2009 7:49 am Post subject:

Magman,

Great eyes there man !! Just to explain a bit further, even with no signal input, this amplifier centers it's output at 1/2 of Vcc. This means that if you have a 9V supply, the output will be sitting at 4.5V. The decoupling capacitor that is now shorted allows this to be applied to 8 ohms to ground. The amp is trying to deliver 2.5W to the speaker .. ouch . but it can't and thus gets HOT Rolling Eyes

It's a good chance that the LM386 is destroyed so if you rearrange things and it still does not work, it's may just be a bad amplifier chip.

Les wrote:

fluxmonkey · Joined: Jun 24, 2005 Posts: 708 Location: cleve

i've got a step-by-step tutorial for building the "little gem" lm386 amp here: http://www.fluxmonkey.com/electronoize/386amplifier.htm... starts out with an explanation of how breadboards works (as alluded to above), and winds up with something like this:

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

State Machine · Posted: Mon Sep 07, 2009 10:36 am Post subject:

Good reference bbob Very Happy

This configuration will allow one to dial in up to 46 dB of gain so the it will be easy clip the amplifier if one inputs too much signal level.

Ref Data Sheet:

http://www.nari.ee.ethz.ch/wireless/education/PPS/PPS02/doc/LM386.pdf

Bill

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside

Success!
No more hot chip
Alas, I had to replace the LM386 with a new one, since the old one
was burnt out.
I did what was suggested and moved the large cap and output down a peg.
I also "hanged" a cap between pins 4 and 6. Don't know if I hanged it right
but it seems to be ok

Thanks for all your help guys, I'll have to give that 'Little Gem' circuit
bbob mentioned a try.
I figured taking a pic of a circuit that doesn't work would be better than trying to explain
everything.
Every one should do this, it's much easier to troubleshoot.
Perhaps taking a pic could be mentioned in the FAQ, we have a FAQ right?
Or maybe a mod could make a sticky post or something.

New Circuit with Hanging cap.JPG

Description:

Filesize:

54.9 KB

Viewed:

335 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

State Machine · Posted: Mon Sep 07, 2009 5:00 pm Post subject:

Great news AND great collaboration by all I must say !! It's good to see the teamwork involved in helping others ...

Now it's time to amplify something Wink

Bill

PS: Perfect capacitor hanging to ... Cool

acidblue · Joined: Jun 26, 2009 Posts: 226 Location: The Darkside