@@MarkTillotson - and there are numerous benefits of Class D for low power amps. The LM386 is so old, there's not really any reason to use such a device now days.
@@johncoops6897 Could you be specific with what part number would be preferable to the LM386? I am guessing that many other chips might be out there now... any which could drive 4 Ohm speakers?
@@BruceNitroxpro - One of my favorites is the PAM8403, which can be purchased as a complete STEREO module for less than $1 on eBay, etc Supply Voltage range / typical * LM386N-3 supply 4V to 12V / typ. 9V * PAM8403 supply 2.5V to 5.5V / typ. 5V Output @ Volts, Ohms @ 10% THD / Typical THD * LM386N-3 @ 9V 8 Ohm = 1 x 0.7W (max) / typ. 0.2% THD * PAM8403 @ 5V 4 Ohm = 2x3.2W @10%, 2x2.5W @ 1% / typ 0.15% So the PAM has heaps more power, uses a "easier" voltage (5V USB), can drive 4 Ohm loads, has lower distortion, massively higher efficiency (better battery life), is dual channel so far smaller PCB, and about the same price.
After four years, this is still one of my favorite projects! I have just reworked the board files and ordered another round of boards. Thank you, John, for having shared this great little circuit!
I remember building LM386 amplifiers as a kid on radioshack breadboards, I drove moms pair of Pioneer CS-203/CS-303 speakers with it. It actually sounded fairly well, and had a bit of bass. Man, brings back memories! Thanks for the nostalgia trip. Then I moved on to TDA2003 amps, (which is fun to use in a solderless breadboard), and LM380, etc... Eventually wound up making a super tiny little LM3886 amp which was very powerful.
have you ever connected this LM386 to the early "STEREO FM" (SONY or PHILLIPS)? Oh boy.. that was fun..I added 1000MFD capacitor connected pin 1 to pin 8. Even better. What a nostalgic trip indeed....
Ive used the LM386 since the early 80's in mostly radio project but its also found itself in most every project Ive made that needed a small audio amp. Ive never given it much thought about making it run any better then the circuits in the data sheet. I would pull them out of old PC modems (US robotics) that where being scrapped years ago so I have quite a few in my parts bin. Like others stated Ive never heard one sound as good as your circuit.
Breadboard done and sounds FANTASTIC !!! I wound up connecting the 47pF ceramic between pins 3 and 4. Now I'm going to build two circuits side-by-side for stereo. I'm sure it'll sound even better built (if that's possible). Volume is perfect for my use as is, no need for any volume control other than my source, the headphone jack on an old Sandisk Sansa I had laying in a drawer. I'm powering it with a couple of 18650's and it's only drawing around 35 mA with the music I'm playing, Joni Mitchell. This is my first amplifier, but now I'm hooked, and I'm sure it won't be my last. THANK YOU, John !!
The tda7052 is a great chip that's similar to the LM386 but it is more stable, doesn't require the decoupling capacitor, and puts out more power on less voltage.
For the gain regulation, I find you can get the widest span in amplification factor with a 10k pot, so your overall resistance will be 909 ohms when the pot is in maximum resistance. This will give you roughly 25 to 200 times amplification, instead of the 50 - 200 that you'd get with a 1k pot.
18:03 This is a balanced differential input, so if you have different circuitry causing a different DC level on the inverting input than you have on the non-inverting input you will create an offset. So when you strayed from the data sheet and added a capacitor to the non-inverting input, of course you need to do the same for the inverting input so that they will both be at the same DC level. Not a big surprise. Unless you really understand the circuit and calculate and simulate to confirm changes, following the data sheet, which is based on good design practices and design reviews, will give the best results.
@ Dana Olson = I am not an EE, so I am a bit confused by the way your reply is written... John placed a 0.27uF cap on the non-inverting input, to clean up the signal input. He then added a 0.27uF cap to the inverting input. I take it that this evened out the offset and that your statement simply explained how the offset occurred in the first place; correct? You are not suggesting that another cap is needed on the non-inverting input, are you? If so, exactly where should it be placed in the circuit? I'm sorry to create confusion, but it is always better to ask a question... Thanks.
great video! I built one for a small 4" 8ohm dayton speakers (1 amp per speaker) and love it! I really appreciate you explaining the different cap options etc...and really appreciate you taking the time to do this...
Thanks! Works great as a mic pre-amp for condenser/electet mic to line level input. I added a 1kOhm resistor from 9V to mic-in to power the mic. My mics have 900-1000 Ohm static impedance. Needed the optional 10 uF capacitor between pins 1 and 8 for gain. I only used 220 uF capacitor at the output. On the breadboard, there was lots of noise pickup but on a compact circuit (one shown is good) it only picked up radio. Probably could use a low-pass filter on it (I was getting good response up to 160 kHz and very slow falloff from there). I found that encasing in a metal cage eliminated the interference so I'd done that and I use without a filter.
I used to make headphone boosters with them. The only difference from your circuit was instead of a 470uf decoupling capacitor, I used 2,200uf for a better bass response. Also seemed to drive into 16 ohm loads better. Ran alright off a 9 volt battery with two lm386 chips
5:47 These damn LM386s made these days are real pains in the rear. The last circuit I made with them (Little Gem Mk II) required me to use a 10uf across Pins 1 & 8, and it had to be connected literally right at the pins or it would motorboat. When I used older chips (like 15 years old) I had no problem, but the newer ones are nothing but trouble.
Hello! I have never built your circuit. However, the LM386 is famous for the following: ① The LM386 has an old and unique design, and omitting the 10KΩ to 5KΩ input volume on pin 3 will result in poor sound quality. ② The design balances the differential amplifier circuit by grounding this volume to -GND via 10KΩ to 5KΩ and grounding pin 2 to -GND. ③ Therefore, if the input volume is not needed, it is necessary to insert a fixed resistor of 10KΩ to 5KΩ between pin 3 and -GND instead. And so on (from Japan).
I think that is the best design. Never heard an LM386 amplifier like this. I tried to make it many times but I never was able to make it sound clear like this.
Nice to see someone make a decent circuit on a prototyping pc board with the LM-386. So many knock this chip because they have no idea how to properly build the circuit in the schematic. They think they can build a high gain circuit on a plug in proto bread board (what you call socket board) using hook up wire and have no clue about mutual induction between wires and stray inductance and capacitive coupling between all the busses on a "socket board". LS Short, don't let any of the output signal get back into the input. To do that effectively you have to use twisted pair at the least, like you have here on the input and output and keep all leads as short as possible and to get the most performance out of the chip, use SHIELDED audio cable on the input and output with the center lead as short as possible and connected to good ground. Build the circuit on a etched double sided PC board with the input and output separated by a large ground plane as the chip was intended to be put on by the designer of the chip, International Semiconductor. Electronics Engineers like myself and knowledgeable techs like you can get away with a good circuit mounted on a proto board like we've both done with a few tricks. I have a two way intercom between my living room and outside privacy fence gate using the LM 386N-4 driven by a TL-082 OpAmp fed to a Radio Shack Mic that has been working since 2008. I love this combo and I can pick up sounds from blocks away. I've thought about making a vid on how to properly build an audio amp with this chip for a long time. It's amazing how loud and clean this little device will drive an 8 Ohm speaker and will fill a whole room with sound.
Very interesting what you say. It would be nice if you want and have the time to do it, to show how you do it in order to cover all the points you mention. I have problems with the smokey amp, and I am quite curious about what I am doing wrong to get that squeals and horrible noises, while the original ones do perform good or very good.
I've just built this amp on breadboard using your cap values, it sounds excellent, thanks! Stay away from Amazon and the cheap "lm386's" they have... they're fake. Sounded terrible when I was swapping them in and out. I used real Ti one's in the end.
Have built loads of LM386 amplifiers, usually for 🎸 practice amps. I didn't know about dumping - input via capacitor rather than a short to ground, much better. For guitar use, I often use a one stage BC549C preamp to get a bit more welly and use high pass filtering as the sub harmonics from a standard guitar has a lot of bass. LM386 withstands abuse, well. I use a 12 volt supply into 3 ohms. It's been working like that for years but isn't a good idea. Helpful upload, thank you.
This worked amzing! I have built many mini amplifiers using the lm386 and none of them have even come close to the pure disturbance free audio, this circuit and layout gave me.
The best results I've had with these LM386 designs were the ones using a FET as a buffer for the input -- the Ruby from runoffgroove mainly. I recall using a J201 to best effect, but apparently the through-hole package for those are out of official production now? I also liked a bigger output capacitor than most of the LM386 designs (they usually called for a 220uf), I think I usually went 470uf like you're using here, or 680uf, especially for guitar use. I remember finding a 10k volume pot to cut too much high end for HiFi audio, and a 50k audio taper working best for me there. I'm not an engineer though, I just made semi-blind substitutions. I remember socketing the transistors and the gain resistors to try different things. Thank for this design, it's simple and compact and if I get back into it I'll try it out for HiFi. But I still have a stereo buffered LM386 board I never bothered to put into a box and use yet, lol.
John, Thank You Very Much for your videos. I learn something new every time, and your discussion helps me a great deal on what to look for and things I should consider when I'm working with electronics. Wishing You and Snickers the very best.
The frequency sweep shows a hint of crossover distortion appearing above about 10kHz, a combination of inadequate bias on the output (to give low quiescent current) and starting to run out of open loop gain. As you mentioned, configuring these little amps for high gain makes performance worse in pretty much every respect and should be avoided for any application were reasonable audio quality is desired.
2:20 You create a capacitive divider at the input. This is not a low pass filter at all. Check it with LT Spice or something to see what I mean. You will create a standard RC low pass filter if you add the normal series input resistor. Something like 1 kOhm would work really well and not change the gain much at all as the input resistance is huge at 50 kohm. Resistors are practically free.
Ever since I tried this little amp when I got it from Radio Shack in 1986 it was such a disappointment I never tried it again. Your circuit surprised me!
I'm having a good time with this little chip ^^ I used it to make my practice little guitar amp and I'm trying to improve the circuit, hence why I'm here :) For guitar amps the distortion the chip can produce is VERY nice, it's almost a fuzz type of distortion and it's very simple to do. However I can not get clean tones from the amp, for clean tones they are pretty bad, at least for guitars :(
Hi, John. Please tell me what type of capacitor you actually used for the 47pF input cap, across the LM386's two input pins, pins 2 and 3. I watched the video, again, to catch this detail. I thought that you had commented on it, but, if so, I did not catch it, this time. Around 14+ minutes, there is a quick glimpse of the component. It sort of looks like a yellow ceramic, but I am just not sure. Teachable Moment = In addition to telling me what you actually used, please comment on what would be the best type of cap to use for this application and why so. Also, what might be substituted, in a pinch... Such as, maybe what appears to be a common ceramic is actually a Tantalum? This is very much the audio signal, so I was expecting some type of a film cap. I checked at Mouser on 47pF film caps and they only seem to be available in really high voltages. I have a nice 47pF/1KVDC-250VAC/Polypropylene/AC and Pulse Film cap on hand. And, I that you taught me that, if the cap voltage rating greatly exceeds the signals voltage, the cap will not form up correctly and the rated capacitance will not be achieved? Thanks, again... I have learned so much from you!
The capacitor is a NPO ceramic type. It is a low pass filter to help keep RF out of the amplifier. Ceramic is fine for this duty. Value can be changed depending on needs.
@@JohnAudioTech "NPO" = I haven't heard that one. Common ceramic - DIP, gum drop, etc. I know of these. I take it that a common ceramic will do. Thanks for the reply and the information. Cheers!
Adding the 10uf capacitor across pins 2 and 7 quietens the sound a bit, but playing the audio through a azotom stereo speaker the sound is pretty good thanks
Thanks for taking the time to explain the capacitor on the negative input of the LM386. I've seen it done before, but as a relative noob to electronics, it wasn't really clear why.
To short out the noise due to the internal 50k resistor - both Johnson noise and flicker noise(*), and reflected current noise from the transistor base due to the 50k. (*) integrated resistors are not metal film, not the best performing resistors.
Because of internal current limit, the amplifier would actually produce more power into a 16Ω load, than 8Ω with a 12 volt supply. I consider that to be an absolute maximum voltage, and would not use that much in an amplifier design. At 9 volts supply, it really is best at 8 Ohms. According to the graph, the amplifier output seems to current limit at 375 milliamps. If the amp would maintain this current limit with a 4 volt power supply, it would fully drive the 4Ω load.
Thanks very much for sharing your expertise. Everything you said made good sense to me. One area where I’d very much appreciate your comments regards the choice of capacitor types; specifically why the one ceramic cap and why the two mylar caps? Thanks!
Excellent! So many LM386 (and other IC amp) projects misuse them. You not only talked about why, but showed what happens. Another thing to consider is that a volume pot on the input side should have a cap between the wiper arm and the amplifier input, because bias currents through a wiper arm can cause extra noise.
18:55 The harmonics you see at clipping are the reason why you blow your tweeter during clipping. This is why "underpowering" is a common cause of breaking your speakers.
@H A R Yes it can. When the amplifier starts clipping you get tons of harmonics and those harmonics are unexpected HF that can stress your tweeter and damage it (probably the main reason of tweeter failure). See : jblpro.com/en-asia/site_elements/danger-low-power This can happen if you clip any stage of the audio signal.
About 3 years ago i build my own circuit like this,with lm386 Chip,all on my own so it played very well therr wasnt like a pot on for volume but i used a aux cable through phone so it played very nice,also the cap that i use give more bass
Nope, at least not when loaded with 8 Ohms. As explained in the video @8:00 the amp can hardly provide any more current. The increased load would lead to a significant voltage drop at each output stage, only causing more distortion.
Great explanation of this circuit, I've a couple of those 386 chips and I've done this circuit similar to yours(minus a few caps), and I use it for my guitar. I got a tube driver adaptor that plugs direct to the guitar and then I put the 386 amp on after that, where I connected with the phono male and female then out to the speaker. Your guide has solved a bit of puzzle why I couldn't directly connect the 386 amp into the guitar and then to the speaker ie. it needs a load for it to work(I think). I've used a dimmer switch for a push on and off, and another switch for the 200 gain, with a pot to adjust the loudness and it's powered by a 9v battery and it lasts for ages. Thanks a lot for this and I subbed :)
Hi Nick Sparks; I would like to ask you something. As far as I am concerned you used the LM386 to make a guitar amplifier. Does this circuit usefull for a guitar amplifier ? If it is not is there any modifications that I can make on this circuit to make it usefull? I am looking forward to hear from you. Have a nice day.
Thank you so much for posting this! I have tried several of these circuits and come away disappointed. This one is promising for a project I am working on. I really appreciate your explanations. I am assuming uF for each that don't specfically say p for pF. Is that right?
Great video as always, just watched this again, having made a couple of your lm386 circuits. These are so instructive and clear, thank you. Just one query...was that music by a band I used to listen to called Gryphon?
Thats the right way to Design a good running preamp, thank you for the Video! Your Video save my Time to Explain how to do it.These Old school Method is much more Precise than the CAD Way.Thank you, God bless you. Now i Subscribe, because you are very clever.
My first audio amplifier is using LA4440 12years ago.. Above half Volume and boom.. Too much noise and now i love TDA for low profile 2.1 Speaker setup
Thank you for your clear explanation of well designed release. Also for your teaching sharing of your study free of charge. I will duplicate your circuit for my hobby s.w. radio.
I used to make these as headphone booster amps for Walkmans. But I found the bass response was better if i used 2200uf decoupling capacitors on the output. Got bloody loud with some JVC over the ear headphones. They were rated nearly a watt of input power !
Thank you for the explanations, dear John, but I got a doubt: near 17:00min. you change something until the wave get round, not clipped. I'm too amateur to have an ocilloscope.And as I'm lookin for a good - not excessive but not poor - gain for a diy reusing of a pair of very small laptop speakers (but yet a 8ohm speakers), I'd like to know how to get a decent gain. Are you using that 10uF capacitor between pins 1 and 8? If so, which resistance is used to get no distortion? I'd like you to excuse me if you said that in the video but I'm not too fluent listener in English. Well... actualy, as you can see, not a fluent writer-speaker too. Thank you.
Az első általam épített (2x)LM386-os erősítőm az egykor kiégett PC hangszóróim helyére készült el, 2009-ben. Azóta csak egyszer kellett végfok kondenzátorokat cserélnem benne, így azóta is rendben működik. A második egy LM386-os monobox amit egy ételtartó dobozba építettem be, és 2011-óta működik, de ennél is a végfok kimenő kondi ment tönkre elsőnek. azóta is rendben működik. Szerény véleményem szerint az LM386-ot az örökké valóságnak készítették.
I can't wait to build this. I was planning on using it as a small guitar amp. In regards to the pin 1 and 8 gain stage could I use 1k pot in series with the 10mf cap to adjust the gain? And if so, what order should I have the components?
Have watched your vids for some time and finally subscribed. Have time to experiment and play. Good vid and good explanation. Much appreciated. Thank you.
The input is a differential pair so having both inputs capacitive coupled is great, but the 47pF capacitor should be on the same side of the coupling capacitors. As you show it, it is inside the + terminals coupling capacitor and outside the - coupling capacitor. It's not obvious why you would want to unbalance your input this way...
@@YTANDY100 It is connected from the positive input of the opamp to ground. It should be connected from the positive input of the opamp to the negative input of the opamp, or across IN. Both of those are balanced configurations. As shown, it is NOT balanced. That is pretty strange given that John correctly added the 0.27uF capacitor between the negative input of the opamp and ground to maintain a balanced input...
The .27uf caps would essentially be a short to the frequencies the 47pf cap would see as high impedance so you can consider that when placing it in the circuit.
@@JohnAudioTech At many very high frequencies that might cause oscillations, the 0.27uF capacitor will be past resonance and may have a higher impedance than the 47pf capacitor... But I concede this is a minor detail. Thanks for the great video!
Ah I finally got this IC to work! so many poor circuit diagrams and instructions out there and no explanation of which part is filtering / doing what - thankyou!
if you want more power or ability to use other loads you would have to graduate from the LM series to the TDA branded chips especially if you want a few watts in stereo unless you want to pile on LM chips on a PCB..
I built it with SMD components on double sided PCB and it's small. Works great! Question - can I put a capacitor with 10kOhm pot in series between pin 1 and 8 to control gain?
Lm386 N-1 has 325 mw @ 6V 8 ohm, N-3 has 700 mw @ 9V 8 ohms and N-4 has 1Watts @ 15V 16 ohms according to datasheet. So which of these flavors we are actually getting now in the open market as there is nothing printed after Lm386 on the chips ?
Ceramic capacitors are notorious for they lack of stability, they also shift capacitance with vibration. That's why you avoid using them in anything here value stability is more important that lower ESR.
I'd like to use this amp, but I want to add a volume control, a reverb circuit and an overdrive circuit to control them with pots as well. Can you direct me to site where I can learn that? Thanks 👍
Hello buddy. Iam trying to build an amplifier using TDA7265 amplifier chip with single supply design. The music is playing but don't know why the volume is so low. I have put input volume on full. For the circuit diagram, you can check the datasheet. Iam using fig 2, single supply design.
Nice video. Thanks for sharing. Which speakers are you using? 8 ohms, but how many watts and ways? It seems 2 or 3 ways, and they seem powerful. Speakers are also important for the final result as you mention in the video.
Hi John - Thank you for this helpful video. I am puzzled by one thing. You added a .27 cap from the inverting input to ground to balance the circuit. However you then add the 47p cap from the non-inverting input to ground. Doesn’t this unbalance things? - Jim
It’s a shame that todays replacement (The Pam 8403 amp) cant be configured for headphone boosting as its a simpler amplifier with more output and at a cheaper price
John, in your opinion, would a tda2822m bridge would make a good mini guitar amp? i don't have many LM386 if any, but I do have many many many... tda2822m
Never seen or heard a 386 perform this well. Always excited to see people getting the most from cheap and readily available components.
The LM386 is EOL'd by TI, perhaps will be by other manufacturers at some point, since class D chips are readily available for this power level.
@@MarkTillotson - and there are numerous benefits of Class D for low power amps. The LM386 is so old, there's not really any reason to use such a device now days.
@@johncoops6897 Could you be specific with what part number would be preferable to the LM386? I am guessing that many other chips might be out there now... any which could drive 4 Ohm speakers?
@@BruceNitroxpro - One of my favorites is the PAM8403, which can be purchased as a complete STEREO module for less than $1 on eBay, etc
Supply Voltage range / typical
* LM386N-3 supply 4V to 12V / typ. 9V
* PAM8403 supply 2.5V to 5.5V / typ. 5V
Output @ Volts, Ohms @ 10% THD / Typical THD
* LM386N-3 @ 9V 8 Ohm = 1 x 0.7W (max) / typ. 0.2% THD
* PAM8403 @ 5V 4 Ohm = 2x3.2W @10%, 2x2.5W @ 1% / typ 0.15%
So the PAM has heaps more power, uses a "easier" voltage (5V USB), can drive 4 Ohm loads, has lower distortion, massively higher efficiency (better battery life), is dual channel so far smaller PCB, and about the same price.
Agreed. Sounds pretty nice.
After four years, this is still one of my favorite projects! I have just reworked the board files and ordered another round of boards. Thank you, John, for having shared this great little circuit!
Your channel deserve millions subscribers dude I'm impressed by your hard work for your viewers
This was my first amp I ever buillt back in the 1990's :)
Same here.
I remember building LM386 amplifiers as a kid on radioshack breadboards, I drove moms pair of Pioneer CS-203/CS-303 speakers with it. It actually sounded fairly well, and had a bit of bass. Man, brings back memories! Thanks for the nostalgia trip. Then I moved on to TDA2003 amps, (which is fun to use in a solderless breadboard), and LM380, etc... Eventually wound up making a super tiny little LM3886 amp which was very powerful.
have you ever connected this LM386 to the early "STEREO FM" (SONY or PHILLIPS)? Oh boy.. that was fun..I added 1000MFD capacitor connected pin 1 to pin 8. Even better. What a nostalgic trip indeed....
This is absolutely the best video I have even seen on using the LM386 the best way possible. Thanks very much for posting this !
I really appreciate how you explain what individual components do. I'm an absolute beginner and that really helps me.
Ive used the LM386 since the early 80's in mostly radio project but its also found itself in most every project Ive made that needed a small audio amp. Ive never given it much thought about making it run any better then the circuits in the data sheet. I would pull them out of old PC modems (US robotics) that where being scrapped years ago so I have quite a few in my parts bin. Like others stated Ive never heard one sound as good as your circuit.
I read & followed Radio Shack book..then connected to the early "STEREO FM" gadget. It was fun back then..
Breadboard done and sounds FANTASTIC !!! I wound up connecting the 47pF ceramic between pins 3 and 4. Now I'm going to build two circuits side-by-side for stereo. I'm sure it'll sound even better built (if that's possible). Volume is perfect for my use as is, no need for any volume control other than my source, the headphone jack on an old Sandisk Sansa I had laying in a drawer. I'm powering it with a couple of 18650's and it's only drawing around 35 mA with the music I'm playing, Joni Mitchell. This is my first amplifier, but now I'm hooked, and I'm sure it won't be my last. THANK YOU, John !!
Make a video of that sweet amp, i wanna see.
@@analogaudiorules1724 I just followed John's excellent instructions exactly.
Great build. I like the cap on the grounded input to get better symmetrical clipping - clever design.
The tda7052 is a great chip that's similar to the LM386 but it is more stable, doesn't require the decoupling capacitor, and puts out more power on less voltage.
For the gain regulation, I find you can get the widest span in amplification factor with a 10k pot, so your overall resistance will be 909 ohms when the pot is in maximum resistance. This will give you roughly 25 to 200 times amplification, instead of the 50 - 200 that you'd get with a 1k pot.
18:03 This is a balanced differential input, so if you have different circuitry causing a different DC level on the inverting input than you have on the non-inverting input you will create an offset. So when you strayed from the data sheet and added a capacitor to the non-inverting input, of course you need to do the same for the inverting input so that they will both be at the same DC level. Not a big surprise. Unless you really understand the circuit and calculate and simulate to confirm changes, following the data sheet, which is based on good design practices and design reviews, will give the best results.
@ Dana Olson = I am not an EE, so I am a bit confused by the way your reply is written... John placed a 0.27uF cap on the non-inverting input, to clean up the signal input. He then added a 0.27uF cap to the inverting input. I take it that this evened out the offset and that your statement simply explained how the offset occurred in the first place; correct? You are not suggesting that another cap is needed on the non-inverting input, are you? If so, exactly where should it be placed in the circuit? I'm sorry to create confusion, but it is always better to ask a question... Thanks.
great video! I built one for a small 4" 8ohm dayton speakers (1 amp per speaker) and love it! I really appreciate you explaining the different cap options etc...and really appreciate you taking the time to do this...
Thanks! Works great as a mic pre-amp for condenser/electet mic to line level input. I added a 1kOhm resistor from 9V to mic-in to power the mic. My mics have 900-1000 Ohm static impedance. Needed the optional 10 uF capacitor between pins 1 and 8 for gain. I only used 220 uF capacitor at the output. On the breadboard, there was lots of noise pickup but on a compact circuit (one shown is good) it only picked up radio. Probably could use a low-pass filter on it (I was getting good response up to 160 kHz and very slow falloff from there). I found that encasing in a metal cage eliminated the interference so I'd done that and I use without a filter.
I used to make headphone boosters with them. The only difference from your circuit was instead of a 470uf decoupling capacitor, I used 2,200uf for a better bass response. Also seemed to drive into 16 ohm loads better. Ran alright off a 9 volt battery with two lm386 chips
After struggling to get good sound out of my LM386 I found this video. Nicely done and thank you! I'm a subscriber now.
5:47 These damn LM386s made these days are real pains in the rear. The last circuit I made with them (Little Gem Mk II) required me to use a 10uf across Pins 1 & 8, and it had to be connected literally right at the pins or it would motorboat. When I used older chips (like 15 years old) I had no problem, but the newer ones are nothing but trouble.
Dave Smith - Could maybe be why
Hello! I have never built your circuit. However, the LM386 is famous for the following: ① The LM386 has an old and unique design, and omitting the 10KΩ to 5KΩ input volume on pin 3 will result in poor sound quality. ② The design balances the differential amplifier circuit by grounding this volume to -GND via 10KΩ to 5KΩ and grounding pin 2 to -GND. ③ Therefore, if the input volume is not needed, it is necessary to insert a fixed resistor of 10KΩ to 5KΩ between pin 3 and -GND instead. And so on (from Japan).
My desk sound system uses LM386 running on 5V driving a 4 Ohm speaker. It sounds sweet at moderate levels where it's only outputting 250mV or so.
I think that is the best design. Never heard an LM386 amplifier like this.
I tried to make it many times but I never was able to make it sound clear like this.
Nice to see someone make a decent circuit on a prototyping pc board with the LM-386. So many knock this chip because they have no idea how to properly build the circuit in the schematic. They think they can build a high gain circuit on a plug in proto bread board (what you call socket board) using hook up wire and have no clue about mutual induction between wires and stray inductance and capacitive coupling between all the busses on a "socket board". LS Short, don't let any of the output signal get back into the input.
To do that effectively you have to use twisted pair at the least, like you have here on the input and output and keep all leads as short as possible and to get the most performance out of the chip, use SHIELDED audio cable on the input and output with the center lead as short as possible and connected to good ground. Build the circuit on a etched double sided PC board with the input and output separated by a large ground plane as the chip was intended to be put on by the designer of the chip, International Semiconductor.
Electronics Engineers like myself and knowledgeable techs like you can get away with a good circuit mounted on a proto board like we've both done with a few tricks.
I have a two way intercom between my living room and outside privacy fence gate using the LM 386N-4 driven by a TL-082 OpAmp fed to a Radio Shack Mic that has been working since 2008. I love this combo and I can pick up sounds from blocks away.
I've thought about making a vid on how to properly build an audio amp with this chip for a long time. It's amazing how loud and clean this little device will drive an 8 Ohm speaker and will fill a whole room with sound.
Very interesting what you say. It would be nice if you want and have the time to do it, to show how you do it in order to cover all the points you mention. I have problems with the smokey amp, and I am quite curious about what I am doing wrong to get that squeals and horrible noises, while the original ones do perform good or very good.
Thank you for the explanation behind this circuit! Alot of channels just say put this together and it works. Again thank you!
I've just built this amp on breadboard using your cap values, it sounds excellent, thanks! Stay away from Amazon and the cheap "lm386's" they have... they're fake. Sounded terrible when I was swapping them in and out. I used real Ti one's in the end.
Hi John, never seen the 386 perform so well!
Have built loads of LM386 amplifiers, usually for 🎸 practice amps. I didn't know about dumping - input via capacitor rather than a short to ground, much better. For guitar use, I often use a one stage BC549C preamp to get a bit more welly and use high pass filtering as the sub harmonics from a standard guitar has a lot of bass. LM386 withstands abuse, well. I use a 12 volt supply into 3 ohms. It's been working like that for years but isn't a good idea. Helpful upload, thank you.
This worked amzing! I have built many mini amplifiers using the lm386 and none of them have even come close to the pure disturbance free audio, this circuit and layout gave me.
Dude that's so nice I've never seen this performance of lm386 before
This is EXCELLENT !
I am currently using a 386 amp to add power to my 1.5v powered alarm (piezo) to turn on a 1.5v relay. Thank u for doing this.
The best results I've had with these LM386 designs were the ones using a FET as a buffer for the input -- the Ruby from runoffgroove mainly. I recall using a J201 to best effect, but apparently the through-hole package for those are out of official production now? I also liked a bigger output capacitor than most of the LM386 designs (they usually called for a 220uf), I think I usually went 470uf like you're using here, or 680uf, especially for guitar use. I remember finding a 10k volume pot to cut too much high end for HiFi audio, and a 50k audio taper working best for me there. I'm not an engineer though, I just made semi-blind substitutions. I remember socketing the transistors and the gain resistors to try different things. Thank for this design, it's simple and compact and if I get back into it I'll try it out for HiFi. But I still have a stereo buffered LM386 board I never bothered to put into a box and use yet, lol.
so, you can amplify the signal before the ic amplifies it?
John,
Thank You Very Much for your videos. I learn something new every time, and your discussion helps me a great deal on what to look for and things I should consider when I'm working with electronics. Wishing You and Snickers the very best.
Thanks for the kind comment!
The frequency sweep shows a hint of crossover distortion appearing above about 10kHz, a combination of inadequate bias on the output (to give low quiescent current) and starting to run out of open loop gain. As you mentioned, configuring these little amps for high gain makes performance worse in pretty much every respect and should be avoided for any application were reasonable audio quality is desired.
2:20 You create a capacitive divider at the input. This is not a low pass filter at all. Check it with LT Spice or something to see what I mean. You will create a standard RC low pass filter if you add the normal series input resistor. Something like 1 kOhm would work really well and not change the gain much at all as the input resistance is huge at 50 kohm. Resistors are practically free.
thank you very much. I updated by Archer engineer's mini notebook ref 279-5011 by Forrest M. Mims the 3rd I bought at Radioshack last century.
CaptainDangeax I have that book. Still wanna wanna build that diode cascade multiplier that goes on for days
thanks for a great vid, you filled in a couple of missing pieces of the puzzle for me.
That music sample was frickin' fantastic !!! WOW !!! I have to try this build. Great job and thanks so much for posting.
the sound was incredible for a few dollars worth of parts - best I have heard from a 386.
Ever since I tried this little amp when I got it from Radio Shack in 1986 it was such a disappointment I never tried it again. Your circuit surprised me!
Me too, never had a good experience with LM386.
If you don't setup the grounding very well, it will distort badly or even oscillate. It can sound good properly configured.
I'm having a good time with this little chip ^^
I used it to make my practice little guitar amp and I'm trying to improve the circuit, hence why I'm here :)
For guitar amps the distortion the chip can produce is VERY nice, it's almost a fuzz type of distortion and it's very simple to do.
However I can not get clean tones from the amp, for clean tones they are pretty bad, at least for guitars :(
@@JohnAudioTech Do one on the venerable tough as nails Sanyo STK086
@@JohnAudioTech PS, Sanyo Chip attempt ruclips.net/video/5WSQSX5GMvc/видео.html
Honestly impressive I've learned a lot watching this
Hi, John. Please tell me what type of capacitor you actually used for the 47pF input cap, across the LM386's two input pins, pins 2 and 3. I watched the video, again, to catch this detail. I thought that you had commented on it, but, if so, I did not catch it, this time. Around 14+ minutes, there is a quick glimpse of the component. It sort of looks like a yellow ceramic, but I am just not sure.
Teachable Moment = In addition to telling me what you actually used, please comment on what would be the best type of cap to use for this application and why so. Also, what might be substituted, in a pinch... Such as, maybe what appears to be a common ceramic is actually a Tantalum?
This is very much the audio signal, so I was expecting some type of a film cap. I checked at Mouser on 47pF film caps and they only seem to be available in really high voltages. I have a nice 47pF/1KVDC-250VAC/Polypropylene/AC and Pulse Film cap on hand. And, I that you taught me that, if the cap voltage rating greatly exceeds the signals voltage, the cap will not form up correctly and the rated capacitance will not be achieved? Thanks, again... I have learned so much from you!
The capacitor is a NPO ceramic type. It is a low pass filter to help keep RF out of the amplifier. Ceramic is fine for this duty. Value can be changed depending on needs.
@@JohnAudioTech "NPO" = I haven't heard that one. Common ceramic - DIP, gum drop, etc. I know of these. I take it that a common ceramic will do. Thanks for the reply and the information. Cheers!
Adding the 10uf capacitor across pins 2 and 7 quietens the sound a bit, but playing the audio through a azotom stereo speaker the sound is pretty good thanks
Great tip for perf. board and 0.1'' graph paper.
Thanks for taking the time to explain the capacitor on the negative input of the LM386. I've seen it done before, but as a relative noob to electronics, it wasn't really clear why.
To short out the noise due to the internal 50k resistor - both Johnson noise and flicker noise(*), and reflected current noise from the transistor base due to the 50k.
(*) integrated resistors are not metal film, not the best performing resistors.
3
I use these amps in my transceiver circuits. Nice video and I will incorporate your tips in my builds.
You´re the best! I can´t believe you got that quality from that ¨thing¨
Because of internal current limit, the amplifier would actually produce more power into a 16Ω load, than 8Ω with a 12 volt supply. I consider that to be an absolute maximum voltage, and would not use that much in an amplifier design. At 9 volts supply, it really is best at 8 Ohms. According to the graph, the amplifier output seems to current limit at 375 milliamps. If the amp would maintain this current limit with a 4 volt power supply, it would fully drive the 4Ω load.
Thanks very much for sharing your expertise. Everything you said made good sense to me. One area where I’d very much appreciate your comments regards the choice of capacitor types; specifically why the one ceramic cap and why the two mylar caps? Thanks!
Excellent! So many LM386 (and other IC amp) projects misuse them. You not only talked about why, but showed what happens.
Another thing to consider is that a volume pot on the input side should have a cap between the wiper arm and the amplifier input, because bias currents through a wiper arm can cause extra noise.
18:55 The harmonics you see at clipping are the reason why you blow your tweeter during clipping.
This is why "underpowering" is a common cause of breaking your speakers.
@H A R Yes it can. When the amplifier starts clipping you get tons of harmonics and those harmonics are unexpected HF that can stress your tweeter and damage it (probably the main reason of tweeter failure).
See : jblpro.com/en-asia/site_elements/danger-low-power
This can happen if you clip any stage of the audio signal.
About 3 years ago i build my own circuit like this,with lm386 Chip,all on my own so it played very well therr wasnt like a pot on for volume but i used a aux cable through phone so it played very nice,also the cap that i use give more bass
i used a 5v split supply and eliminated the output cap slightly adjusted the bias and got 3v into 8 ohms flat down to 10Hz
beautiful design and performance , speaker also counted for this
Hello John as you know there are 3 different versions of the 386 each with different max output levels, which version are you using here ?
How did you calculate the values of capacitor?
Always recomend supply decoupling capacitor approx 10uf near to the chip as possible.
it would be cool to see you build the circuit component by component if you ever get feeling to do that :)
true
It sounds so good. Fantastic work.
Hey John love you stuff
Should do a 3 lm386 chips in a bridge stereo circuit
Can get 3 watt
A bridged amp uses two amps in anti-phase, why would you use three of them?
or use a proper TDA branded chip that can deliver that power in stereo..
PAM 8xxx is really enough
ElfNet Designs what’s the fun in that?
Nope, at least not when loaded with 8 Ohms. As explained in the video @8:00 the amp can hardly provide any more current. The increased load would lead to a significant voltage drop at each output stage, only causing more distortion.
much better than my LM386. great !
may be i should redo mine.
Dont use the chinese diagrams to build it and you will have something worth using.
Great explanation of this circuit, I've a couple of those 386 chips and I've done this circuit similar to yours(minus a few caps), and I use it for my guitar. I got a tube driver adaptor that plugs direct to the guitar and then I put the 386 amp on after that, where I connected with the phono male and female then out to the speaker. Your guide has solved a bit of puzzle why I couldn't directly connect the 386 amp into the guitar and then to the speaker ie. it needs a load for it to work(I think). I've used a dimmer switch for a push on and off, and another switch for the 200 gain, with a pot to adjust the loudness and it's powered by a 9v battery and it lasts for ages. Thanks a lot for this and I subbed :)
Hi Nick Sparks;
I would like to ask you something. As far as I am concerned you used the LM386 to make a guitar amplifier. Does this circuit usefull for a guitar amplifier ? If it is not is there any modifications that I can make on this circuit to make it usefull? I am looking forward to hear from you. Have a nice day.
Perfect for a headphone amplifier of course 1/channel., I also like using it as a preamp, so it doesn't need a dual power supply, relatively simple
Thank you so much for posting this! I have tried several of these circuits and come away disappointed. This one is promising for a project I am working on. I really appreciate your explanations. I am assuming uF for each that don't specfically say p for pF. Is that right?
Great video as always, just watched this again, having made a couple of your lm386 circuits. These are so instructive and clear, thank you.
Just one query...was that music by a band I used to listen to called Gryphon?
Unbelievable. I never thought LM386 sounds so great !
Hello, John! Could you make a video of how to make preassembled lm386 module, that you may find cheap on Amazon or Ali, sound best?
Thats the right way to Design a good running preamp, thank you for the Video! Your Video save my Time to Explain how to do it.These Old school Method is much more Precise than the CAD Way.Thank you, God bless you.
Now i Subscribe, because you are very clever.
My first audio amplifier is using LA4440 12years ago.. Above half Volume and boom.. Too much noise and now i love TDA for low profile 2.1 Speaker setup
Thank you for your clear explanation of well designed release. Also for your teaching sharing of your study free of charge.
I will duplicate your circuit for my hobby s.w. radio.
I used to make these as headphone booster amps for Walkmans. But I found the bass response was better if i used 2200uf decoupling capacitors on the output. Got bloody loud with some JVC over the ear headphones. They were rated nearly a watt of input power !
1 amp to rule them all!
Thank you for the explanations, dear John, but I got a doubt: near 17:00min. you change something until the wave get round, not clipped. I'm too amateur to have an ocilloscope.And as I'm lookin for a good - not excessive but not poor - gain for a diy reusing of a pair of very small laptop speakers (but yet a 8ohm speakers), I'd like to know how to get a decent gain.
Are you using that 10uF capacitor between pins 1 and 8? If so, which resistance is used to get no distortion?
I'd like you to excuse me if you said that in the video but I'm not too fluent listener in English. Well... actualy, as you can see, not a fluent writer-speaker too.
Thank you.
Az első általam épített (2x)LM386-os erősítőm az egykor kiégett PC hangszóróim helyére készült el, 2009-ben. Azóta csak egyszer kellett végfok kondenzátorokat cserélnem benne, így azóta is rendben működik.
A második egy LM386-os monobox amit egy ételtartó dobozba építettem be, és 2011-óta működik, de ennél is a végfok kimenő kondi ment tönkre elsőnek. azóta is rendben működik.
Szerény véleményem szerint az LM386-ot az örökké valóságnak készítették.
On Pin 3 of other circuits, I see a 470pF rather then your 47pF. Why?
I suggest LM386-4, since it's rated at 1Watt output!
I can't wait to build this. I was planning on using it as a small guitar amp. In regards to the pin 1 and 8 gain stage could I use 1k pot in series with the 10mf cap to adjust the gain? And if so, what order should I have the components?
Better use 10k
Have watched your vids for some time and finally subscribed. Have time to experiment and play. Good vid and good explanation. Much appreciated. Thank you.
The input is a differential pair so having both inputs capacitive coupled is great, but the 47pF capacitor should be on the same side of the coupling capacitors. As you show it, it is inside the + terminals coupling capacitor and outside the - coupling capacitor. It's not obvious why you would want to unbalance your input this way...
@
ats89117
it is connected to ground not - input :-)
@@YTANDY100 It is connected from the positive input of the opamp to ground. It should be connected from the positive input of the opamp to the negative input of the opamp, or across IN. Both of those are balanced configurations. As shown, it is NOT balanced. That is pretty strange given that John correctly added the 0.27uF capacitor between the negative input of the opamp and ground to maintain a balanced input...
@@ats89117
across in would probably be better but to both + and - inputs would maybe negate higher input frequency's :-)
The .27uf caps would essentially be a short to the frequencies the 47pf cap would see as high impedance so you can consider that when placing it in the circuit.
@@JohnAudioTech At many very high frequencies that might cause oscillations, the 0.27uF capacitor will be past resonance and may have a higher impedance than the 47pf capacitor... But I concede this is a minor detail. Thanks for the great video!
Ah I finally got this IC to work! so many poor circuit diagrams and instructions out there and no explanation of which part is filtering / doing what - thankyou!
If possible in the next video please show us some low voltage mic amplifiers circuits.
i also would like to see that using electrets and dynamic
Nice
Can I build one on breadboard?
Also where is volume control potentiometer?
At 6:30 you discuss an output coupling capacitor of 470 “ohms”. Expecting to see something like 470 uf. Is that correct?
if you want more power or ability to use other loads you would have to graduate from the LM series to the TDA branded chips especially if you want a few watts in stereo unless you want to pile on LM chips on a PCB..
I just built this great little project! It's great! Now, I will be building a MP3 player.
What an amazing explanation! But why did you connect bypass capacitor at dc 9v?
its all about the ground plane with this IC.
I built it with SMD components on double sided PCB and it's small. Works great! Question - can I put a capacitor with 10kOhm pot in series between pin 1 and 8 to control gain?
i have always wanted to try that, pity no answer yet
Nice explanation of the LM386, thanks
Lm386 N-1 has 325 mw @ 6V 8 ohm, N-3 has 700 mw @ 9V 8 ohms and N-4 has 1Watts @ 15V 16 ohms according to datasheet. So which of these flavors we are actually getting now in the open market as there is nothing printed after Lm386 on the chips ?
Hello John, from about 19:40 to 19:50, what is happening around the zero crossing? There's a bulge on the sine wave :)
I'll have to check that. It seems to be on the falling edge only.
@@JohnAudioTech Use an analog scope for this purpose. These sh... digital scopes are not good for this measurement
@@JohnAudioTech CAN A LM386 BE USED IN AN FM ( BUG ) TRANSMITTER ?
Great video! I see you’re using film capacitors. Is there a reason I shouldn’t use ceramic caps? I’m asking as a beginner. :)
Ceramic capacitors are notorious for they lack of stability, they also shift capacitance with vibration.
That's why you avoid using them in anything here value stability is more important that lower ESR.
I'd like to use this amp, but I want to add a volume control, a reverb circuit and an overdrive circuit to control them with pots as well.
Can you direct me to site where I can learn that?
Thanks 👍
Beautiful. Thanks John.
Hello buddy. Iam trying to build an amplifier using TDA7265 amplifier chip with single supply design. The music is playing but don't know why the volume is so low. I have put input volume on full. For the circuit diagram, you can check the datasheet. Iam using fig 2, single supply design.
sir what is 1KHz normalising while measuring audio amplifiers.thank u
Nice video. Thanks for sharing. Which speakers are you using? 8 ohms, but how many watts and ways? It seems 2 or 3 ways, and they seem powerful. Speakers are also important for the final result as you mention in the video.
I have connected the one 8ohm two way speaker. SPL would rate in the low 90dB range at 1watt / 1 meter.
Hi John - Thank you for this helpful video. I am puzzled by one thing. You added a .27 cap from the inverting input to ground to balance the circuit. However you then add the 47p cap from the non-inverting input to ground. Doesn’t this unbalance things? - Jim
the 470 output component is a capacitor but you called it out in ohms. That makes it unclear what value you want it to have 470uF, 470pf?
I misspoke, it is 470uf.
If if was pF you'd hear nothing, if it was nF only the high frequencies, so it must be µF or greater, mF would be unnecessary.
Thanks, but I have received an answer from the author. He clarified it without even a small hint of condescending tone. I like that.
It’s a shame that todays replacement (The Pam 8403 amp) cant be configured for headphone boosting as its a simpler amplifier with more output and at a cheaper price
Wish you had this as a kit to put together... With a pcb
John, in your opinion, would a tda2822m bridge would make a good mini guitar amp? i don't have many LM386 if any, but I do have many many many... tda2822m
It sounds really good!