Actually this is not a new solution for that kind of problem. It`s state of the art. The designer of the IC makes the analysis of the circuitry whit consideration that the power supply connection (Pin 6) is a short to ground for AC signals. As you have discovered this is easily done by using a bypass capacitor between the power pins. Usually that cap is kept as close a possible to the IC. Without this capacitor,depending on the length of the power wires from the power supply, the amp can become unstable and begin to oscillate, what probably happened in your case. The usual technique is to put two caps parallel to the power pins of the IC. One cap is an electrolytic one which is able to bypass the low AC frequencies the other one is a sort of ceramic or other types which are able to by pass the higher frequencies. Butting only a electrolytic cap eliminates the low oscillations, which you hear through the speaker, but can sometime let oscillating the amp in higher frequencies modes even around 1MHz which your speaker can`t reproduce. The amp works properly anyway in his designed frequency range, you can hear some music but still oscillate on a higher frequency.

Thank you, Mr. Heilman. Very generous and informative video. You, sir, are most kind! I just received an order of a few LM324 (Quad) Op Amps, and am intending on building a similar stereo amp for use as I begin building simple -- eventually more complex audio oscillators, filters, etc. and one (or more?)16-Step Sequencers. I've seriously been bitten and smitten by the whole Euro Rack Synth Modules thing, late to arrive as I am. I will certainly apply your tips as well as some others offered by other kind, generous commenters. It's wonderful to show what you know, it's a true gift to educate. THANK YOU!

Hi HH, Just a little FYI, proper ground layout is really important with the LM386 and with any other chip amp. A separate ground path for input, power and output, the Star Grounding technique, as per John Audio Tech on YT. All the ground paths should junction right at pin 4, GND. Filter cap should be as close as possible to pin 6 and 4. Be good, C.

Dude, you save my life! Thanks!

Thanks for showing the problem and basic solution to handle it.

Thank you so much for posting this. Saved me (and my lab buddy) a whole lot of time!

The capacitor works even for voltage supply by the battery. Very good advice. Thanks

I didn't realize how bad the noise can be, and with a simple capacitor, it'll all go away. Now I understand why there are more capacitors than the basic circuit design.

u best man and your english is perfect to understand for me!

Hey, Thank you for your nice video. You explained it very nice way. Specially with the schematic....

good matt, i like lm386 and had same problem. nice video

Had same issue - this fixed it for me - thanks very much really helpful...

You can also connect a capacitor from pin 7 (Bypass) to ground, which ought to stabilise it a bit more and potentially remove a bit more noise. The datasheet doesn't give a value but since it's for smoothing, it can be as big as you like. 10uF seems enough; maybe even less.

God bless you! Thanks a lot for your help!

Me ayudaste mucho , tenía ese problema y no aparecía en ningún esquema . Ya estaba loco . muchas gracias , un me gusta y suscrito .

Your power supply is outputting chopped up DC, probably pulse width modulated with inadequate smoothing. If you run into a situation where a capacitor alone isn't enough without so much capacitance it isn't feasible, an inductor before it in series will push it over the top, however, it will trade spiky ripple with slow voltage drift. This is only an option if you are using dirty DC to feed a voltage regulator. Then the voltage out of the regulator will be smooth as a baby's bottom. Basically: Dirty DC -> 2.5-5mH inductor -> Capacitors -> LM317 -> Absolutely prestine DC with no noise.

Useful info. Thanks for sharing that solution!

Pin #7 is a Bypass for the Input Transistor's Collector... run a 0.1uf capacitor to ground. This will help with noise immunity. Another good practice, is to use an 100 ohm resistor between your power source and the Pin #6 power input, and place an electrolytic capacitor as close to Pin #6 and ground. You should also ground the Zoble Network and the Speaker as close to the LM386's ground pins.

Many thanks my friend!!!

"Problem solved" ... Not really, problem created removing it. The decoupling cap must be as close as possible to the chip with a 100nF ceramic cap. And by the way, it is not noise, it is oscillation.

Straight up knowledge bomb

hi. i know capacitors can be used to eliminste noises. but why vcc to grnd? can someone explain?

Thank you for that, mine didn't make the noise like yours, but it was getting a lot of interference from TV and radio, after I put the Cap. it worked fine. ( I had a 450uF cap on hand and that's what I used)

I am using a pam8403 class d amp module and when I put a capacitor across vcc and ground it only makes more noise than there already is now when I use an addresable led strip near that amp like the ws2812 it make some stronger humming noise unless the input wire is being disconnected and I wonder why?

If you want some more gain, add a 10uF cap between pin 1 and pin 8.

That was awesome!!!

If I have two speakers bridged, do I have to add a 250microfarad to both speakers or just the first positive?

that noise likely wont be a problem in the first place if you're using a battery as the power supply. it just sounds like switching noise from the power supply you're using but as noted below you should have a bypass cap "filtering" the input anyway because its an audio circuit and once you get past a few dozen Hz inductive effects on the ground plane ( just a wire here) can no longer be ignored. also as noted below, you shouldn't need nearly as large of a cap. its also worth noting that electrolytics alone aren't necessarily ideal for this purpose but they'll do fine as long as you're not using this to pre amp for a much more powerful amplifier. nominally you would want a mylar or silver/mica in parallel with a decent electrolytic because they can respond much faster to the sudden power needs and have a much more stable esr over a wider range of frequencies, but thats going all audiophile and junk.

thank you matt!!!!

I have an application where I actually need the noise. How do you aggravate the noise?

Hey dude you are a genius! Thank you for this video, really!

deberianos dejarnos el diagrama o pcb para realizarlo...muy interesante gracias...

Thank you!

This is EXACTLY the awful noise I'm getting on mine, thanks I'm going to try it

For some reason if I bridged pins 1 and 8, then bridge it to pin 2 ( which is also grounded ) it also removes all the noise, well my circuit is slightly different from his but never the less the same without the second capacitor between VCC and Ground. Im here to ask someone, why did connecting pins (1 & 8) then to 2 removes all the noise in my setup? Thanks in advanced.

me re sirvio , gracias y likes y nuevo suscriptor

hello engineer do you have a page where this diagram and your list of materials ???

Thanks for this info... :)

Is that a high pass filter at the output ? Via 10ohm and .05uf?

Can u tell me a circuit that removes noise from the input signal and obtains rectangular pulse in the output

I had made a amp with this chip about a month ago and i had the same problem but i put a LC filter or just an inductor from pin 5 to ground and it removed all the noise to me!

A good power-supply filter circuit AND a differential-input to the amplifier could be a solution.

Thank you

Thanks!

I guess I've been lucky, but nothing I've built with a 386, even when I place a 10uf electrolytic between pins 1 and 8 for max gain (200x), has ever caused me noise problems. However, I like the cure you present; don't think I was aware of it, and don't believe it even appears on the IC datasheets I've seen.

I believe pin 2 is always grounded and another pin has a cap to cut out am radio interference, if you mess around with these chips and make a good crossover for your speakers and use all good components the sound quality can be hi fi quality, much better than most boom boxes and so on.

This is no noise. It is oscillation. Sufficient grounding is essential. The ground point with the lowest ac impedance has to sit as close as possible to the ground pin of the chip amp. This point is usually the minus pole of the electrolytic buffer capacitor. This buffer capacitor should be paralleled with a ceramic capacitor of 100 nF in order to short also higher frequency oscillations which are not audible but a load for the chip amp. And, as Cass Virgillo wrote, use separate ground paths from this ground point to the input and output ground.

awesome thanks pro

I've used a 450 pF it work as well.

So what caused the noise in the first place?

Damn, it was the noise from your PSU :( I have a battery-powered circuit with LM386, producing some pretty loud hissing noise. Putting a capacitor on the supply wires has no effect (not a guess, actually tried it). Filtering the output has some effect though, but muffles the useful signal as well.

what is the code of capacitor (from vcc to gnd ) ?

Sisters of Mercy - Yeah

Nice post

HI everyone, I'm working on this project> When i tried the cap as u said in proteus it feels like it increased the pitch. Can anyone put some light on it.

This is a cry for help not criticism. I love the video, it's exactly the problem I'm trying to learn about. BUT, I'm New, I'm trying hard. You move the old Board so fast I can never get a good look. Help Me, did you bring the VCC( the pos input voltage) straight into the bread board and not hook it to the +/- feed on the outside 2 rows of breadboard? Looks like you go straight from output speaker to the outside +/- seemingly defeating the pin 5 capacitor and resistor. It also looks like you have a POS feed going through the pin 5 Cap and R - > to the speaker as diagrammed. Also Pin 3 should hit the 10k R then move on. Is the jumper from pin 2->pin 4 allowing you to shift the pin 3 1k R to pin 4, in a straight line due to the 2-4 jumper? If I make No Sense, I completely understand. Sincerely, ...........NBSTATIC.

Supply capacitor is crucial but frequently not shown in schematic. Cheers from Indonesia

How bout distortion pedal noise as well ? Budd

hi, Can you please make a video to show us what is the difference between sounds with and without amplifier. This is very interesting. Nice video BTW 👍

Voy a probarlo

just power supply issue, are you using bench power supply or any sort of AC to DC adapter?

Looks like that noise was coming from the power supply, you should probe first with a battery 9v becouse is a low noise source of suplly

If the wires running from your power supply is longer than 10 centimetres you must use filter capacitor else it will motorboat like this. I faced the same in case of building tda2030A amp circuit

sorry that's not noise - that's your LM386 oscillating

0.1uf axial cap and 1000uf in output 100uf for input and gain 820uf for feedback pin P.s all caps are made in Japan I find that old stuff in my local store

put 450uf capacitor between pin 7 and ground to get rid of any noise and put high capacitor on the power gate

Whenever i plug the jack in my mobile while the circuit is powered, my mobile screen starts to act strangly, i have tried the circuit with different mobiles but still face the same problem. Its like some current flow into mobile through the ciecuit

my woofer has noise.how can i get reed of it

This was also my problem with the LM386.

nice video ,would you gie me your schematic this circiut and the lis off the komponen??,please

Tis your power supply ripple. If you run your circuit on a battery, this doesn't happen, because the power supply is perfectly smooth, though it's still advisable to bridge the Vcc and GND pins of the IC - with the shortest route possible - using a 100nF cap (1206 SMD ceramics are the best for this purpose as you can connect them nice and easy without introducing a parasitic inductance). This little cap will help provide short bursts of energy with low latency when a battery or power supply has a much longer response time to provide the necessary power.

well... you have problem you make your in 20 gain set if you want to make much hi gain set in 200 just plug mylar .2 uf in pin 1,8 and put some a resistor that has resistance 22k in input of ic parallel in the ground filtered much the sound stability and .1 uf electrolytic into input of you put in the pot it will much filtered block the unwanted noise in the device and make the input signal will much fidelity and in my opinion you make the amp much better in crank just put 2200 or 3300 uf electrolytic so it will boost the watts yet filtered pwr source and parallel again a .1uf mylar so the unwanted noise go out and much no noise and the last to balance the low to hi freq ratio to output just put a .47 uf mylar parallel to resistor 1 ohm to gnd so it will just the bass will came out yet in hi freq and last put a .47 uf maylar in pin 7 so the sound way perfect in 99% sure the amp will be in hi fidelity and no distortion at all promise you make what I mean so it can be the pure and potential of that ic will came out promise to you but da next secret is I almost forgot is parallel again a 1k uf in 330 uf in output so the signal from amp is so filtered that I can tell you can fabricated and sell a radio amp no noise no distortion even in full volume of the amp

How can you forget the cap that provides the gain?

I guess some noise could come from the audio wires not being insulated.

Can you please post the schematics

Please sir Dolby NR circuit uploaded.

Naice

10uf cap from pin1 to 8, then take the cathode of an SB140 schottky rectifier and connect it to pin 8, connect the anode to pin 4. Noise gone.

And here i was losing my mind

Thankyou ihave this probleme but now Has disappeared

muito bom 3

Much better, still not so good yet though.

Diodes, who needs them! Arrive at: 'Circuit Solver' by Phasor Systems on Google Play.

not fix. circuit just need it

Bauhaus

looks like that noise was caused by switching within your power supply

pin 2 aint ground, pin 4 is.

I know the best configuration for the LM386. Try this one: (# 1) a 10uF CP across pin 1 and pin 8. (#2) connect pin 2 and pin 4 to the negative terminal of a electrolytic CP (1000uf) connect pin 7 to the positive terminal of the same capacitor. Than you you know the rest

😱😱😱😱😱😱😱😱

your video quality and annoying music are reminiscent of Indian people electronics assembly videos. Next.