The Definitive NOOB Guide to Building an Amplifier Enclosure out of Wood

Let’s make a gravity mounted enclosure for an amplifier. Usually this is for a tube amplifier, but it’ll work for anything. If you’ve seen a Bottlehead Crack, you know what a gravity mounted enclosure is.

So purty. Let’s make one!

The wood base is called a skirt and the metal sheet that the components ultimately are mounted to simply sits on top of it. Gravity keeps it in place, though taking a bit of time to secure it to the skirt with screws is a good idea. We’ll focus on the wood part – creating a nice top panel by hand is another post entirely.

Before we get any further, we have two important lists to talk about. One makes sure you are not wasting your time and the other is so you can reasonably do this and, thus, not waste your time.

What you should expect:

  1. It will not be perfect. Things will be cut incorrectly, things will be a millimeter or two too long, too short, or of the incorrect slope. You’ll miss a tiny bit of wood glue and it will show through the stain. Accept this will happen.
  2. You have to want imperfection. If you are the person that cannot work with imperfection, you want to just buy something from a company where a machine does it.
  3. It involves time and a lot of waiting. Days worth.
  4. You will make mistakes ranging from “oh, that’s kind cool” to “Eh, I think I can work with this” to “shit, time to go back to the lumber yard”.
  5. You need to have sane tools. You can get away with really basic tooling but the crappier your tools, the crappier your end product and the worse precision you’ll end up with. However, embracing those imperfections will go a long way to a unique and awesome end product. But, no way around the shit tools and shit product correlation.

What you need:

  1. A miter saw. A miter box is probably gonna suck, I do not recommend it. I also do not recommend a hand miter saw. I’ve tried that and found it difficult to get accurate angles, especially if trying to put together a skirt requiring angle cuts.
  2. An orbital sander and sanding block. You might be able to get away with only a sanding block, but say good bye to more time and very sore arms/hands.
  3. A quality caliper.
  4. Wood glue, any will do.
  5. Wood screws (optional, dependant on the kind of skirt)
  6. Wood dowels (optional, dependant on the kind of skirt)
  7. Stain, get a few different stains and test them on scrap to see what looks best to you. I like Minwax Early American as it looks great on a variety of woods. A small can of stain will go a long way.
  8. Some sort of finish. Shellac or spray-on polyurethane are my go-to. Spray on is going to produce superior results vs. trying to brush it on, unless you’re way more talented than I.
  9. A drill.
  10. An electric screw driver (be nice to your hands/wrists and get an electric one).
  11. Painter’s tape.
  12. Clamps (at least 4 long-ish ones, more the better).
  13. Angle clamps (at least 1 of them, more the better).
  14. A speed square.
  15. A drill press.

That’s the sane list and if you don’t have that stuff already, you’re putting up a decent cost up-front. I consider that list to be the minimum for a decent product. You can get a lot of other things that will help and increase the awesome-ness of the finished product, but are not necessary IMO.

A tale of two skirts

There are two skirt designs I like. The first is a box of 4 equal length pieces of wood with 45 degree angles on each end that fit together to create a box with seams only at the edges. This creates a very clean and classy look. If your metal sheet is 12″ x 12″ and 1/2″ thick, you’re looking at 4 pieces cut to 13″. The downside to building this is that if you’re not able to cut accurate angles, you’ll have issues fitting it together nicely at the end. That is work-around-able depending on how mis-aligned it is, but that’s the biggest risk. We’ll call this the Pretty One™ and is the configuration Bottlehead uses. I’ve used it a number of times and it looks great. It’s probably the most popular skirt DIYer’s make.

A “small” tube preamp using a pretty skirt. Made of oak. This one actually kinda has some flaws, but still looks good to the non-insane.

The other consists of two lengths that are the length of two sides of your top panel and two that are the length of the other two sides, plus the thickness of the wood. So, if your same metal panel is 12″ x 12″ and your wood is 1/2″ thick, you’re looking at two 13″ lengths and two 12″ lengths. Seams on the sides, not the edges. This is easier to put together and will result in much more noticeable lines. There are also ways to incorporate decorative physical seaming or lines with this method. We’ll call this The Rustic™.

A Rustic™ mid-build.

Which method to use? It depends. If I’m looking for a more sleek look, the first option is the best. If I’d like to build something with lines and perhaps a more “worn” or “hand made” look, the latter is best. Both will produce a very sturdy and safe product, which is just as important as beauty.

Greg, Power Supplies, Dim Bulbs, and Not Getting Electrocuted

First, it’s been a while since I last posted. COVID-19 and all kinds of other personal, work, and other stuffs have been making it hard. I hope y’all are ok, well, and dealing as well as you can.

Ok, so after the O2 it all gets a bit hazy. It was around this time a number of projects happened, another Pete Millett amp, I think I built a quick CMOY for some reason, and probably something else – but I remember building a ESS ES9038Q2M dac from a board I bought from AliExpress. The particular board I purchased has lots of variants and there is an extensive discussion of that family of Chinese dac boards on diyaudio.com.

It was my first time actually building a power supply and using a transformer. This was scary. Things like grounding were no longer about eliminating noise but about safety and not being electrocuted. Obviously, this was scary and just reading about what to do on the internet seemed a bit risky.

My day job was working in tech. Much of the company was remote so we all used a chat tool called Slack to communicate. What was great about this is Slack allowed for various channels or rooms that were work related and whatever else you could want. Remote worker water-cooler chats happened in places like “The Danger Room” or “Pancakes”, the latter being a place to post pictures of delicious pancakes. One of those channels was dedicated to music and I had already started talking about amplifiers, dacs, headphones, and all that in there. I was likely being annoying to everyone who was there to discuss Taylor Swift.

But it was through that Slack channel I met Greg. Greg was in to DIY audio as well and was much more accomplished. He’d built speakers and power amplifiers – you name it, he’d done it. He had a crazy awesome home lab complete with measurement gear, a CNC, and loads of other goodies. He was not only soldering things together but milling aluminum enclosures and fabricating the entire product.

He is/was my idol.

Shortly after discovering we were doing sorta the same thing, we started a “DIY Audio” channel and, though I felt silly, started asking him questions about wiring for my dac. Windings, secondaries, center tapped – stuff I didn’t know about and he’d explain to me. I told him I needed 15V AC to the dac board, which apparently had a it’s own DC power supply. He showed me a R-Core transformer that would give me the 15V AC I needed and walked me through wiring it correctly, using a fuse, and where to ground things.

The enclosure I purchased was one from Hammond Mfg. and had some nice wood side panels. I’d take a picture, send it over slack, he’d say “the green wire should go to ground, twist those red/white and black/white ones together”, and it’d do it. Funny enough, if I’d just looked at the wiring diagram that came with the transformer I proably could have figured it out. However, since we’re dealing with AC electricity, I was scared to death. And death is possible.

He also suggested I spend some time building a dim bulb tester. What that does is lets you, relatively safely, test high current gear without a ton of risk of damage. If there is a short, the light bulb will light up and there shouldn’t be a fuck ton of current flowing through the gear. Now, a small dac is not quite super high voltage and testers like this come in to play big time with anything involving tubes, but AC mains are still involved and I didn’t want to risk fire, death, or any sort of bodily harm. I built one out of a wall electrical box and a plugin light bulb receptacle after a quick trip to Home Depot.

There were some things in the middle here – the board had mounted RCA jacks on it and I wasn’t sure enough of my ability to drill correctly aligned holes in the expensive enclosure, so I opted to de-solder them, panel mount them to the enclosure, and run wire back to the board. I did this and ruined most of the solder pads. This lead me to having to trace those pads back to their origins and attempt to solder the wires to those points. I did that, felt accomplished, and re-checked continuity with my multimeter at least 5 times.

Alright, the moment of truth. I’d drilled out the chassis I purchased with my hand drill (which probably deserves it’s own post), wired everything, secured everything, shrink wrapped any wires where something conductive could possibly touch or even move, plugged the dac in to the one end of the dim bulb tester, the tester in to a power strip in the off position, and inserted a light bulb. With fingers crossed and standing a more than a few feet away from the dac, I turned on the power strip.

A small LED lit up on the dac board. No smoke. I was still terrified.

Using a multimeter set to measure AC, I carefully touched the probes to the enclosure to make sure no current was flowing. They said none was. I flipped them to DC, touched one to the enclosure and the other to a spot on the board where I knew DC flowing. It measured the correct voltage. Holy shit.

So, at this point you have a few options. Keep testing various points to make sure things measure as they should or YOLO and plug something in. Because I live on the edge, I plugged in my laptop and beloved Butte. It worked. I heard sound, music, and no noise. Holy shit.

That dac was a lot of terrifiyingly fun to make. It was my first experience with a transformer. Thinking back on it, I could have just bought a 15VAC wall wart and went that route, not even dealing with the transformer, IEC inlet, and grounding would have been a lot easier. But it didn’t even occur to me. At the time I’m not sure I knew it was even an option.

Three things were learned in this build. I learned about transformers, windings, what a center tap is, primaries, and secondaries. I really met Greg and got to know someone who actually knew what they were doing, though I’m sure he’d say he doesn’t know. I also learned that anyone who says a dac makes a significant difference in how something sounds is full of crap. I’m sure there may be exceptions to that, but my experience with this dac, compared to everything else and the onboard dac on my computer, is they all sounded good. Or rather, was just there. Couldn’t hear a difference. I’d learn that as long as the actual analog stages (IV, gain, output buffering) of the digital to analog conversion were properly implemented, it shouldn’t matter.

So, just like Pete, thanks Greg. It’s been a lot of fun to have someone to almost get electrocuted with.

The Objective 2 and NFB-11

I had finished up building Pete Millett’s Butte and really enjoyed it. I thought it better than the other amplifiers I’d had at the time, or at least on par.

Previous to the Butte, my high end amplifier was from a company called Audio GD (pronounced Audio Gawwwd, according to me) which is a small boutique manufacturer in China. Their website is amazing and exactly what you’d expect from something poorly translated from Chinese to English. My favorite page is titled “The babies in gestating” and describes their high quality build and QA process. All joking aside, the build quality of their gear is quite good.

The Audio-GD NFB-11

I feel this amplifier deserves a few paragraphs of thoughts, as it really did leave an impression on me and I learned something from it.

The particular model I owned was the NFB-11 and it was another combination headphone amplifier and DAC, like the Schiit Fulla and Fulla 2 I owned previously. There would be a number of revisions to the NFB-11 over the years, but I bought the original version. I had chosen it due to a lot of folks going on and on about how great it was. Words and phrases like “musical”, “excellent transients”, “hard hitting”, “fast”, and “natural” were used in the reviews I watched and read about it. Plus Zeos said it was awesome. Pretty much there was kool-aid being passed around the headphone community and I drank it.

A big highlight of the amplifier was actually ordering it. Ordering it involves emailing an @vip.163.com email address with what you wanted, waiting for a few days, getting a quote back with a possibly suspect PayPal link, and then PayPaling money. After that, you waited and prayed. A month later, a package should arrive from China and you have an amplifier. I’m not sure what would happen if you needed to return it for warranty reasons, but I’m sure that it was a similar process. I call their process for ordering The Indian Jones Process.

I went through the ordering process and after about a month, a large package containing an amplifier showed up at my door.

After spending some time with the NFB-11 I thought it sounded good, but still much like all the other amplifiers. The DAC portion actually was a pain in the ass as it required the installation of special drivers and those drivers were not easy to install. As expected, everything was in Chinese and was not as simple as opening up an installer and hitting “next” a few times. After an evening of mucking around with Google translate on some non-English speaking audio forums, I had them working and I could listen to 384khz music files. Not that I’ve ever actually heard an audible difference between standard CD-quality 44.1khz lossless files and a unicorn native 384khz file.

However, if I needed to, I could. That’s pretty rad, lol.

I feel temped to talk a bit about DACs and my thoughts on them, but I think that’s it’s own post.

The other thing I do remember about this amplifier is that it had a great volume knob and potentiometer. Silly as it sounds, I still grin when a large knob turns with just the right amount of tension and the music volume adjusts accordingly, reacting perfectly. It’s a “feeling” thing that I do believe is important in the music experience. It’s the same reason I enjoy vinyl – you must take some physical action and if that action “feels” right, it’s enjoyable. The NFB-11 definitely did that part right. It was the first amplifier I used that gave me pleasure to operate. It felt well engineered and sexy. Remembering that feeling would become important as I started to build more amplifiers and customize them to my liking.

Never fuck up the knobs. Never ever.

However, at the end of the day, it sounded like everything else I’d heard in most regards. Just it physically was bigger, looked cooler, and gave me increased audio geek cred. Plus the volume knob.

I finished building the Butte a bit after acquiring the amplifier and, due to that, had moved on to thinking about the Objective-2 amp.


This is what I was waiting for. The DIY amplifier that disrupted the headphone game. Uber disrupting the taxi cab business. Apple disrupting the cell phone. The Objective 2 (O2 from here out) disrupting headphone amplifier.

Needless to say, I had very high expectations for this one.

The O2 is interesting. NwAvGuy gave out the schematic for it freely, though partnered with manufactures to have the PCBs built. This was all done under the Attribution-NoDerivs 3.0 Unported license, which I find disappointing. Not quite Open Source, but not quite proprietary. You can make it and sell it, but can’t modify it (and then share your improved version) even with attribution. So, here is this thing I made that’s pretty cool but fuck you if you improve it and then share those improvements. None the less, a small outfit near St. Louis, JDS Labs, sold PCBs for it so I picked one up. JDS Labs would go on to build their own products, and to great success.

Again, I had a bill of materials for the O2 and ordered the parts from that weird electronics store called Mouser. For those that don’t know what a bill of materials is, most electronics projects like this will have a list of all the parts and their part numbers that are needed to build the project. Good projects will have labels on the PCB that correspond to the bill of materials so you just look at the marking on the board and match it up against the part number on the bill of materials to understand what goes where. Some projects may or may not have that last bit and the builder gets to figure it out, but in general, most of well established projects will.

Once the parts arrived I stuffed the PCB. Stuffing just means putting the right components in the right places. After that, I soldered it all in place.

My earlier fear of the components being close together and being more difficult to solder was partially right. They were closer together, but my skills were better now. I avoided any huge mess-ups and double checked, with a multi-meter, that none of the solder pads with solder caked on top were making a connection they shouldn’t. I knew this time to check polarity and the PCB had markings describing how to position anything with positive or negative pins.

The O2 also uses some Operational Amplifiers or opamps if you’re cool. Opamps are another topic that deserves their own post, but I do want to touch on them briefly.

The O2 under construction. The black rectangles with little dots on them are the opamps. Bill of materials in the background and reference IDs on the board are visible.

The opamps were in a DIP-8 package. That means the component (like an opamp) comes with a bunch of small pins (8 for a DIP-8 package) in a standardized size and shape and must positioned correctly on the PCB. Most have a small dot on one end of the part and most PCBs will have a corresponding dot printed. Line the dots up and the component be positioned correctly. Each pin on the opamp has different things going to it, like positive and negative voltage, ground, the audio signal in, and the audio signal out. Installing it upside down could cause something like DC voltage flow to a pin for audio output and, well, we could blow something up. We didn’t want that.

Everything was now soldered so I checked any solder spots that looked sketchy with a muti-meter (which I had learned to use from the almighty Google), attached the power brick, and hit the power button. A small LED lit up, it passed my “is anything burning” sniff test, and I plugged in a music source and some cheap headphones.

Not to derail the story, but one thing about the O2 is that it was meant to be “portable”. In the real world it’s not, though it can be run off two 9 volt batteries if you choose. I don’t know anyone who actually has done this outside of seeing if it works or thinking that technically batteries are a “quieter power source” and thus will make things “sound better”, as it’s just not practical given it’s size. Due to it being designed for “portable” use, the inputs are 3.5mm – headphone jacks in and out on the front panel along with the power connector. In my opinion, this is one of the more awkward designs I’ve seen and goes against “the feel” I got with the NFB-11. Using the amplifier was not “fun” and instead clunky and a pain in the ass. It looked the opposite of sexy. The default Alps potentiometer felt alright, but not as good as the larger and more sturdy one the NFB-11 had.

This is what the O2 looks like in use. Yeah. It’s a mess.

Ok, so yeah – the power light lit up and I connected a source. I was using the Sennheiser HD 600 headphone at the time and cranked the little O2 up, waiting for some magic.

Like any new toy, I listened long and hard. The amplifier certainly ticked boxes – no hiss or any sort of sound in the background. In audiophile speak it had a black background, and it was that way with all my headphones, including some earbuds and in-ear monitors. That’s a very good thing as many earphones have a very low impedance, which means many amplifiers will produce a level of noise or hiss if they themselves do not have a low output impedance and correspondingly low noise floor. For reference, the Butte produced a very small, but audible, hiss with very low impedance earphones.

Other than that, it was powerful enough for the HD 600, which is the opposite of an low impedance and highly sensitive earphone – it had a much higher impedance (300 Ohm) and thus required more voltage to produce higher volumes.

No noise, check. Power for higher impedance headphones, check.

However, the sound was actually a bit different. Not in any sort of significant way and only the kind of thing you would “feel” after spending a lot of time listening to music and then reflecting on it. NwAvGuy said his intention was for it to be transparent and get out of the way of the music. I think it probably does, but to a point of only what I can only describe the O2 as “sterile”. Things sounded as they should, I suppose, but the O2 lacked any and all character. It was surgical compared to the Butte and even the NFB-11. I wasn’t sure it was actually causing me to enjoy the music or if I was more now just trying to dissect everything with my new surgical instrument. Analyzing, rather than enjoying. I’d later realize, though I’d heard folks say it before and knew it all along, that music is art, subjective, and it’s really all up to the listener. I was learning that I preferred some character.

Why do I think most things look best in natural outdoor light? Because it is not perfect white light. It’s light from a giant ball of gas many millions of miles away that has been burning away for an eternity and traveling great distances before it reaches my eyes. It is imperfect.

The O2 is a fine amplifier. I believe it achieves it’s goal – it measures well, doesn’t really color the music, and pretty much works great for anything you can reasonably throw at it with a few notable and rare exceptions (like the AKG K1000, but lol those are special). However, it had zero character and I found myself enjoying the good ol’ Butte a lot more.

This sterility I’m talking about – is it like something where average person is going to say this? No, I’m talking about some 0.05% thing here. If I was a normal person, I’d have given zero shits, not noticed, and kept it since it ticket all these great boxes. But I’m likely not quite normal and am on a journey. No, not a journey, but a quest. Not a quest of chasing the audio dragon, but trying to figure out the various audio dragons and what that meant to me.

I ended up gifting the O2 to my mom for her birthday. As before, I 3D printed a case for it with well wishes inscribed on the bare back plate and a joking “Yerhot Audio, Love John”. She enjoys it to this day. As well as a small collection of other gear I’ve built for her.

Fin. Love you Mom.

Dammit, Pete Millett.

After more rounds of amplifier upgrades and side-grades, I kept hearing about this headphone amplifier that was a giant killer. This mysterious person from the Pacific Northwest designed it to prove a point that super expensive boutique audiophile gear was mostly just hogwash and could be beat, objective performance-wise, by some rather cheap parts and good engineering. It was the Objective 2 and I thought it was so punk rock that I could build my own “better than everything else” amplifier. Super David v. Goliath.

The mystery around the creator – a person who went by NwAvGuy – who posted to a blog for about a year, raised hell, and then disappeared was almost too good.

I spent a few evenings looking at pictures of it and watching videos of folks building it. It looked ridiculously complicated, especially given my only experience thus far was the small Frys kit I built a while back. This was supremely more complicated and involved lots of parts, most of them small and crammed next to each other. It seemed a step past where I was.

Like a lot of folks, I decided it was daunting and I didn’t have time to muck around with it. But, I had to build something now that I had the itch.

On reddit.com/r/headphones someone posted about this amplifier they’d built for their cans. It was called a Butte and I read more about it on this website that looked like it was from 1996. The designer was apparently named Pete and sold stuff on eBay. He said he had a background in electronics and his website seemed like he did. I mean, being so awful looking and full of technical information, he must know his stuff. Honestly, it all seemed sketchy as hell but he had posted stupidly detailed instructions for how to build the Butte and it seemed significantly less complicated than the Objective 2. I figured what the hell – Reddit says other people have done it and that means I can. I ordered the Butte PCB from his eBay page and ordered the pre-configured shopping cart from this weird website called Mouser. I’d never ordered parts like this before, so even that portion was an adventure.

A week later everything arrived.

As usual, I tore it all opened, printed Pete’s assembly instructions, and started on my way inserting resistors where the manual said to and putting on various connectors. After an hour or two I had it assembled and soldered. I decided to power it on with the wall wort that came as part of the preconfigured Mouser shopping cart and hoped for the best.

A little green light lit up and I got excited. A few seconds later I noticed a funny smell. Looking a little closer I noticed something weird, one of the capacitors seemed to be larger than the other. Not in a “good larger” way either. Quickly I unplugged the wall wort.

After some more looking I noticed that in Pete’s pictures the capacitors looked a little different. There are these stripes on them and they went the other way in the pictures. I guess that matters.

This is how I learned about polarity. Turns out polarity is important, folks.

I didn’t really know what a capacitor did, but I did learn that one side is positive and one side is negative. And feeding the wrong side with DC+ isn’t a good idea. I de-soldered the incorrectly placed capacitors and re-soldered them the other way around, though one was bulging out a bit.

Again, I flipped it on, the light turned on, did a quick sniff test for anything smelling funky, and finally connected it to my computer and cheap headphones. Omg it worked. And it sounded good (with super sketchy capacitors in the power supply).

The finished PCB on top the pinnacle of my non-DIY headhone amps, the Audio GD NFB-11

I also noticed that I had a few parts left over. Pete includes an optional CCS circuit that can be installed. I had no idea what this meant or really did, but I decided to install it. Basically, you’re installing a CPC3703 mosfet to make the opamp run at full tilt all the time. Class-A. Moar powwwwa is better than less powwwwa so that was a no brainer. The parts were these small little poopers and I fumbled my way through it. But, it continued to work afterwards and I felt super accomplished.

I even made a 3D printed case for it after some time. Not amazing, but it was fun and cheaper than the $50 or whatever Pete wanted for a really nice aluminum pre-drilled chassis.

Yes, I did replace the capacitors with non-blown up ones. But, wow, I really did think this amp sounded great. It had zero noise and had authority. That’s the best way I can describe it – authoritative. Am I biased because I made it? Probably. But, to this day I still think it’s a damn fine amplifier and for a total cost of less than $100 excluding enclosure, it’s right on the money.

This particular amplifier ended up going to a different home, but I love Pete Millett’s Butte so much I made another one a few years later just because. Well, not completely because – I had started to think about building chassis out of wood and needed an easy project as an excuse to try making a chassis from wood. It turned out alright and is still sitting on my audio rack today.

For a first time DIY audio project, I recommend the Butte big time. And Pete Millett, if you’re reading this, I love you. I really do. I didn’t know it then, but I’d end up building a number of your projects and each and every one is special in different ways. Thank you for being the amazing Colorado guy running a 1996 looking website full of marvelous information that you are.

This project taught me about polarity, AC/DC, and so much more. After almost blowing up (err, really I did) the capacitors and then fixing it, I spent about a week reading and testing various things out. I referenced every single component I put in that PCB and learned that resistors are not polarized, the capacitors were (but not all capacitors are), what a pin out was, where the audio signal went and how it fed into the opamp, what the mosfets I added did, and how a potentiometer is like an adjustable resistor. I feel like I learned so much building, screwing up, and then fixing the Butte that I can’t even express it. I had so much fun.

But, I still wondered about this giant killer – the Objective 2.