loading

I did it! I've always wanted to build my own amplifier, and now, finally, I made one. It's my first "serious audio" project ever.

Starting this project was daunting to me. There was a gap between buying the parts and actually making the amplifier... a two year gap, as a matter of fact. I know a thing or two about analog electronics and I can read and understand electronic diagrams, but I never made something from scratch that connects directly to a wall outlet. And I usually make stuff with electronics either for kids or as a "proof of concept", where accuracy and details matter only little. In other words, I was just insecure about me as a maker of Serious Amplifiers.

What helped me to get started is that I just needed an amplifier. I love to play music and the best sound system I had was my iMac. And that proved not good enough for me. So I had the choice between buying a cheap receiver while knowing I had all the parts for a much better amp or just get started building it. Oh boy, I am SO glad I did just that!

If you find yourself dreaming about building an amplifier now and then, but are not sure about how to get it done: Read on! I described the making of my Serious Amp as good as possible, including the mistakes and successes. The result is not a Perfect Amp, but it surely is a Serious One that is not that hard to build even if you never built one before.

This amplifier is the best one I've ever owned (that claim isn't worth very much) and I'm really, really happy with it. Even with my secondhand (€10,-) loudspeakers it sounds pretty nice. In the video, the sound is recorded with the on-camera mic. What you hear is far, far away from the real life experience.


The amp is a so called Gainclone, using one LM3886 op-amp per channel. The power supply is based on a (way too large) 230VA toroid transformer. It took me about 32 hours to build the amp and power supply.

In this Instructable, I'll describe how I made the power supply and amplifier. More important, I'll also describe what I would do different if I had to do the project again.

Step 1: Links: Get a grip on the Gainclone Universe

I'm not an audiophile and not an expert on amplifiers either. So when I started to look around on "how to build a Serious Amp", I entered the universe of Gainclone amplifiers. Looking at the sometimes gorgeous designs, I realised I wanted to build one of these babies myself. There are many different ways to build them, so browse around and get yourself familiar with the world of gainclones...

  • Mark Hennesy's website: Very, very nice amps and great info on gainclones and powersupplies. Great pics and comments on how he built it. The most inspiring resource I found.
  • Wikipedia article on Gainclone amps and where they come from.
  • The Chill Amp: Well documented site, including schematics and parts list.
  • Chipamp.com: I used their circuit diagram for the amplifier (the pdf is available on their site. Comes along with good tips on building amps). They also sell kits and parts.
  • DIY Chip Amps: Nice listing of several DIY gainclone amplifiers.
  • Audiosector.com: Stunningly beautiful gainclone amps.
  • Decibel Dungeon (that's an audio-nerdy name :-)): Schematics for a gainclone amp and a power supply.
  • Decibel Dungeon Gainclone Gallery: Nice collection of DIY gainclone amps.
On Instructables are just a few gainclone projects. ASCAS' DIY HiFi Gainclone Power Amp is an example.

You may have noticed that there are a lot of slightly different schematics and op-amp types that can be used. Not being an expert, I found it hard to make a choice. I turned to a friend who I consider to be an expert for advice. He convinced me just to choose something simple and trustworthy. The differences in the circuits are about details, which are hardly relevant when building your first amp. Novice mistakes will be made and those will have much more impact for the worse than the refined details have for the better.

What I really like about the Gainclone is the minimalistic approach. The general rule with Gainclones is the less components, the better the amp. This comes at a prize, of course: The quality of the components and their specs become more important, since there are so few components to create the sound.
<p>Thank you again for this great instructable. I've finished my first gainclone this weekend and can't count the hours I've stared on this page. Also, thank you again for your help. I forgot to ground the secondary coil during my first attempt which (according to a friend) causes a &quot;floating&quot; of the current. However, it finally works perfectly fine, without any noise. I've attached two pictures, the wiring is not as nice as yours, but works for me :)</p>
<p>Wow! Thanks so much for your pics and nice comment! Your amp looks GREAT, thumbs up! You added a second line-in, didn't you?</p><p>Cheers, Y.</p>
<p>Yes, it's a second input. I am using the amp for CDs and PC and since I already had the parts I thought it's more convenient.</p>
<p>I've prototyped power supply but there's no voltage between &quot;&ndash;&quot; terminal and GND, but only 40V between + and GND/&ndash;. Is this okay? Shouldn't it be like 25V between + and GND and between &ndash; and GND?</p>
<p>Going to be attempting this in the next week or so. I am going down the route of separating the signals with banana plugs due to it being a work project. Will be using a hammond enclosure and transformer(Work). </p>
<p>I'm surprised TI still makes these amps. They were pretty good in their day, and found their way into many stereo receivers and TVs (along with many competitors' products). Nowadays you'd use a class D amp, and save yourself a bunch of money on power supply and heat sinking. </p>
<p>You seem to be saying that there's a better way to go than this design, then? Could you direct me to an example I could follow, perhaps? I was thinking of making this, it'd be a first for me, but if there's a 'nowadays' thing that saves a heap of money.....</p>
Sorry, I was in Guatemala and just got back, long story. <br><br>I don't think a class D amp should be your first. Layout is tricky when there's a lot of square waves circulating around. Learn how to make something that works. If you just want to save money you can buy Chinese boards from ebay, but you won't learn much.
<p>You seem to be saying that there's a better way to go than this design, then? Could you direct me to an example I could follow, perhaps? I was thinking of making this, it'd be a first for me, but if there's a 'nowadays' thing that saves a heap of money.....</p>
<p>great instructable ynze, I'm gunna need it! I'm a complete novice and this will really help with my project. Just wondered if you'd be able to let me know how I'd go about wiring in 3 inputs as opposed to one with a selector switch and also a pre-amp for phono channel. Thanks, Jon.</p>
<p>BTW, the LM-3886 is rated 38W Continuous Average Output Power into 8&Omega; at Vcc = &plusmn;28V with 0.1% THD+N from 20Hz&ndash;20kHz. <em>(Source Texas Instruments)</em></p><p>Since you're only supplying &plusmn;25V, it will be more like 35W which is still plenty of power for most people.</p>
<p>Great Instructable. I have built many audio amplifiers in my day (way back before there was an internet and Instructables). </p><p>Your DC voltage <em>should</em> be 1.4 x AC voltage. It is the result of rectification of the AC wave itself from the point where voltage is being coupled across the diodes + the small amount factored in by capacitive action. KBU-1005 Vf (forward voltage) is 1.0 V for 5 A so when the amp is drawing 5 A, the diodes are in pairs dissipating 5 W x 2.</p><p>Your DC current will be the reciprocal (.71 x rectifier input current minus the heat wasted in the diodes, capacitors, and transformer).</p>
<p>Awesome build, congrats! After a long time when I saw only Arduinos and digital electronics :) Once I saw someone on Instructables adding a simple audio Bluetooth connection (A2DP) to such amplifier, perhaps it will make it more attractive to be built by young people (n.b. you can connect it to your smartphone - tadaaaam) - </p><p>http://www.ebay.com/itm/3-5mm-Stereo-Audio-Music-Speaker-Receiver-Adapter-Dongle-USB-Bluetooth-Wireless-/381282219096?hash=item58c630f458:g:zVEAAOSw3xJVbpYm</p>
<p>That 47uF cap between the 680 ohm resistor and ground. It is there to block DC from the negative feedback of the amplifier. What it essentially does is allow the negative pin to float as seen from DC, which will help prevent your amplifier from applying an amplified DC offset value which would hurt your dynamic range and put unnecessary current through your speakers causing undue stress.</p>
<p>Also, see here for more technical descriptions of what everything in the circuit does. I just googled for the chip LM3886 and found datasheets and application notes.<br></p><p><a href="http://www.ti.com/lit/ds/symlink/lm3886.pdf" rel="nofollow">http://www.ti.com/lit/ds/symlink/lm3886.pdf</a></p><p><a href="http://www.ti.com/lit/ds/symlink/lm3886.pdf#page=8&zoom=auto,0,705.4" rel="nofollow">http://www.ti.com/lit/ds/symlink/lm3886.pdf#page=8...</a></p>
<p>thats a great project. consider wire wrapping instead of &quot;P2P&quot; soldering. It easy with a cheap wire wrap tool and makes a nice connection that can be unwrapped if you make a mistake. Many old Zenith components from back in the 50's were wire wrapped by hand. Mount the components through a rigid medium like 2.5mm plastic and then drill tiny holes for the leads. then if you want to get really fancy, you can wrap the wires with an insulated ground to shield them after its all done, justremember to tie it to the ground plane. Very nice. </p>
Great presentation. Your written dialog is refreshingly clear and even pleasant to read. A similar, but refined version of my old, &quot;How to fix your VW for the Complete Idiot&quot; It, and you, make a seemingly impossible task, a simple, and logical task of common sense. <br> Thanks. Please post any other projects that you may attempt. I will gladly read about the journey. <br>Thanks, John Millington<br>
<p>Excellent design, congratulations.</p>
<p>When I saw the word &quot;imac&quot; I quit reading...JK!!! LOL!!</p>
<p>You did a wonderful job-congratulations. I have one mono gainclone and am very pleased with it.</p><p>Re: breaking out of wood when drilling: I put tape over the to be drilled area and introduce the drill through the good side of the wood panel through the tape. Or drill from the wood backside while holding the wood panel firmly against a flat piece of scrap wood.</p>
<p>Maybe it's not the best idea to put something you attach to the powergrid in a flammable wooden box?</p>
<p>That crossed my mind as well, but then I thought: 50 years ago most boxes containing electronics were made from wood. The casing can withstand some big sparks for sure. If you want a really fireproof box metal sheets seem the only way to go...</p>
<p>For almost all my projects I use acrylic boxes. It cuts as easy as wood. If you find an acrylic supplier you can usually pick up pretty decent sized scrap pieces for about a $1US(&euro;0.93) per pound (0.45kg) The only tricky part is the acrylic glue is really runny and if not applied carefully can look crappy. But if you use metal L brackets to hold the box together they can absorb a pretty decent sized explosion. Acrylic can also be tapped with thread holes for mounting heat sinks or chassis grounds.</p>
<p>Nice Job.</p><p>I like the way to built the power supply, although I would think that the heatsink if much too small for the amount of power that you use.</p><p>The cases look incredible. You did a fantastic job of putting these together. A finished case like this makes the whole project look professional. I wish I had that type of talent!</p><p>I do have a bit of a concern with the amplifier itself. As it is, it isn't very robust and you will have to make sure that it won't get bumped hard, or be in an area with a lot of vibration. P2P wiring such as this will eventually rear it's ugly head.</p><p>One thing that came to mind was your use of aluminum tubing as the heatsink. This is brilliant. Another idea would be to cut more slots into the tubing and then put a fan at one or both ends to really get the airflow happening. This will keep the amps reasonably cool. Having the slots on the side would allow air to escape on both sides as well.</p><p>All in all, a beautifully made project. I have been looking at building a tube guitar amp for a year or so myself but haven't had the time to get to it, yet. I have had all the parts and more for over a year! You have given me some great idea's and better yet, the inspiration to do something about it!</p>
<p>THIS is the kind of project that got me into electronics longer ago that I care to remember.THANKS!</p>
<p>Nice work! I know this is quite old now but it just got featured in the Instructables newsletter. The 25V DC is a result of the AC peak voltage being significantly higher than the value always given (larger by a factor of the square root of 2) - under low loads or with no load, the voltage output from a transformer, bridge rectifier and capacitor can be quite high.</p>
can i use .1uf mylar instead of mkp? whats the difference?
got my first seriour amp. thanks for the instruction.
<p>Very cool! Cheers!</p><p>Y.</p>
<p>Hey man, very impressed with this amp. What would be your thoughts on scaling up the output wattage to around 200/500 watts?</p>
<p>Uhhhmmmm... You would need either an entirely different op-amp, or design a grid of at least four LM3886's to distribute the power. And then take care of the heat that comes with 200 to 500 Watts. The design of my amp is not suitable for that kind of power...</p>
<p>Hello,</p><p>I finished both amps and instructions are very nicely written, but I have trouble with potentiometer resistors. Could you maybe post a bit clearer image or schematic how to hook it up and what values to use?</p><p>I know that your aim is not to make AMPs for Dummies 101, but I would appreciate so much if you could post the remaining part of the build in similar manner. At the moment I have real trouble figuring out the output safety mechanism and potentiometer resistors.</p><p>Thank you much, if you will have time, if no, wish you best with what you're doing. :)</p><p>Regards</p>
<p>Could you give us some computer generated schematics? I find them easier to read. Eagle is a good program.</p>
<p>this is a great instructable, very easy to follow. i have always wanted to own an amp that worked well, but not having a lot of money or knowledge i figured building one would be the best but never found a good, easy to follow guide. i will definitely be using this to build one for my next serious project.</p>
<p>ok if of you have a problem with that my i suggest!!!!!!!!!!! lm386 half watt ic!!!! i try that one man makes satisfie you self if you the 386 one</p>
<p>and less cheaper than to make and too powerfull and smokey !!!! you will love it</p>
<p>i will research the usage of this, if i can figure it out i may use this idea, thanks!</p>
<p>ill give a schematic in the breadboard i wanna make this!!! and work !!! there was a secret about my schematic that make distortion clear !!! figure it out !!!! so it will help you for first speaker and!!!! by the way!!!!!! there have some missing connection for it and is the!!!!......... hahaha figure it out!!!!. so it will may get you been un the challege!!!</p>
<p>your welcome!!! :D</p>
<p>Enjoyed your project and its inspired me to try to build one. My knowledge of electronics is quite limited but I have been reading up a bit on some of the sites you suggested. One question, where do you connect in your switch, LED and resistor? I don't see them in any of your schematics.</p>
<p>Awesome work! I read the article about your amp on hackaday.com and I've never heard about gainclones before. I haven't such a deep knowledge about electronics, but thought I give it a try. So I ordered all parts the next day (but skipped the Anti-interference module and the rectifier). I ordered the same toroid transformer from Conrad (Germany) and followed your instructions about the second coil, connected the red and purple wire and soldered them on the ground connector of the 220V power cable. I'm pretty sure that I've soldered everything else correctly and triple checked each connection. But after I powered it, the clamp that connected the second coil to the ground started smoking and the speaker made strange sounds!! I've no idea what I've done wrong here, so I can only guess:<br>Was it wrong to connect the secondary coils to the ground?<br><br>I connected every ground of the sketch (amp+power supply) to the ground of the 220V input, is that correct?<br><br>I forgot to wire the parallel capacitors (100uF and 0,1uF) on the V+ cable, but I don't think that was the problem and now I'm scared to retry without any changes. <br><br>I used a cheap 4Ohm instead of a 8Ohm speaker, just in case I break it.</p>
<p>Hi mfaust,</p><p>Thanks for the post. You obviously created a short somewhere, but I can't tell why and where... My guess is that you connected the secondary coils the wrong way.</p><p>Try this:</p><ul><li>Take out the toroid transformer.<li>Compare your toroid's label with mine (enclosed image). With my model, the first sec. coil has colors Black - Red (SW-RT on the label). The other sec. coil has colors Purple and White (VIO-WS on the label). Is that the same as with your model?<li>If so: Disconnect the purple and red wires. <li>Connect just the transformer's primary coil to 230 Volts (nothing can go wrong now, as long as you make sure that the secondary coils' wires don't touch anything). With a Voltmeter, measure the Voltage between the secondary coils' wires: So measure between red and white, red and purple, red and black, etc.<li>Between two pairs of wires you should read 18 Volts AC. <li>Write down what the positive and minus leads are of each coil (you can use the Voltmeter for that. Hook up the black probe to the Voltmeter's COM. If you read +18 Volts, the black probe is connected to the coil's minus lead.</ul><p>If you did this, please leave a comment with the results. We'll go from there :-)</p><p>Good luck!</p><p>Y.</p>
<p>Hello ynze,</p><p>I checked the toroid's model number, it's exactly the same you used. I measured the output AC current I on the black and white wires I connected the power supply with and got the 18 Volts. I also have the 25V output you mentioned in your instruction. So I thinks this works fine.</p><p>Anyway, thanks for your help! I just have to confess that my knowledge is not sufficent for bugfixing this circuit, so I will ask some friends for help. I let you know if I get this thing running some day ;)</p>

About This Instructable

208,885views

1,712favorites

License:

Bio: Also have a look at member Monster-Marit. She Rocks (I'm a bit prejudiced, though :-))
More by ynze:Make your first Serious Amplifier RC circuit for a Bibberbeest / vibrobot 3D printed Push Puppet named Eeps 
Add instructable to: