Dresser: … versus Humidity

It has been a little over 14 months since I finished my dresser. I’ve learned a few things that I’ll consider during my next build, like how epoxy doesn’t really stick to slate, and how heavy that much 4/4 cherry is. But the most surprising one has been the reaction to humidity.

It’s not exactly hard to find free advice on the perils of ignoring seasonal wood movement. We all know that wood expands and contracts as humidity fluctuates. But many of us also get to nearly ignore it. Kiln-dried wood, cut and assembled in a climate-controlled shop, and used in a climate-controlled home, doesn’t actually move much. I built this dresser in San Jose. Even without climate control, the environmental humidity hardly changed 10% throughout the year.

But I haven’t stayed in that environment. This past year, I moved to a home that experiences “real winter” (for two weeks in January, the outside temperature remained below zero Fahrenheit), uses force-air heating (without air conditioning), and also overlooks a lake. The air inside was so dry in January we bought moisturizing lotion at Costco. Spring has finally warmed up the place, and brought days and days of rain. I wouldn’t compare it to the mugginess that the southeastern US experiences … but in comparison to winter, my dresser might.

Humidity has caused the drawer face (cherry, the darker-colored wood) to bow outward, pulling the front of the drawer box away from supporting the drawer bottom.

The front of every drawer is bowed out. It actually doesn’t look terrible with all of the drawers in place. Since they’re all made of aligned segments of the same pieces of wood, all of the bows match. I accounted for some expansion, so the drawers aren’t jammed. But, the bottom panel of each drawer is no longer supported by the groove in the front, so they sag under the weight of clothing. The next-to-lowest drawer bottom sags low enough to catch on the front of the lowest drawer face.

Why did this happen? I used plain-sawn boards instead of quarter-sawn, mostly. But, I alternated the growth ring direction of each board, as is commonly advised. The theory behind that technique is that if each board swells away from its outer growth rings, alternating which side the outer rings are on will cause each board to cup in the opposite direction. The face still wouldn’t be flat, but it would have a few smaller waves instead of one big bow.

With screws removed, the drawer box moves back into place. Alternating growth rings (highlighted in white) didn’t correct the arch.

I didn’t get waves. I got an arch. One, rainbow-like arch. Why? Probably two things, both to do with the face being screwed to the drawer box. First, each board probably didn’t absorb humidity evenly. If they had, each board would have cupped its own way, producing the expected wave. But, the back side of each face, flush against the box of the drawer, likely absorbed much less than the face in the open air. So, the open-air face expanded, and the sealed face didn’t. Second, the drawer box is multi-layer finished birch plywood. It didn’t expand in the humidity at all. While I did oversize the screw holes to allow for some movement, I doubt it was enough to compensate for this much. So, the box front itself kept the back side of the drawer from expanding. I bet if I removed the drawer fronts from the dresser and let them stand free in the open air for a few days, the expected wave shape would show up.

But I’m not interested in getting a wave shape. I’m interested in continuing to use my dresser through the coming humid summer. There are a few ways to think about correcting the bow. The one I’ve decided to use focuses on the use problem, instead of the look problem. The real issue I have with the bowed drawer fronts is that they pull the drawer box front with them, which leaves the drawer bottom unsupported.

Stretchers force the drawer box front to stay against the drawer bottom, even when the tension of the face bow is reapplied.

What I’ve done is to make sure that the front can’t be pulled too far from the back, by installing stretchers between them. This is something I considered putting in the long drawers (40 inches side-to-side) anyway, to add some structure and divide up the space. I used a sliding dovetail on either end of the stretcher, to give it a strong grip on the front and back. I also drove a pan-head screw through the front and back into the stretcher, too keep the stretcher from sliding along the groove as I pull clothes from under and around it.

With the stretchers clamping the box front and back on the square bottom, the screws holding the face to the front were able to easily pull the curve back out of the face. Maybe this is kicking the ball down the road, and I’ll have to deal with a worse problem as humidity continues to soak into the drawer faces, or when it all leaves again next winter, but it seems like this is working so far.

Starlink: Tower

One reason I haven’t worked on fixing my obstructions before now is that winter makes the ground impenetrable, and the roofs treacherous. The second reason is that I didn’t want to attempt any solutions without having data to guide me.

“What data do you need? Just get the dish up as high as you can!” is the sentiment I’ve gotten from the Starlink subreddit. Forty-foot Rohn towers are the “put a bird on it” of that community. But the idea of just ordering $1000 or more of tower, and pouring a large cement pad for it, just to see if that fixes things, doesn’t sit right with me. What if all I needed was 20ft of tower at a different location? What if I actually needed something taller than 40ft?

Don’t even get me started on the, “Just cut down the trees,” crowd.

So once it got warm enough to be outside without thick mittens for more than a minute at a time, I did what any engineer would do: I hacked together a sextant, and mapped the positions and heights of my obstructions – trees, mostly.

Surveying with what you have: tripod + walking stick = reference, protractor + straw + string + weight = sextant.

Then, because my particular engineering specialty is computers, I created a 3D model of the property, and put Dishy in it.

Dishy’s original location, about 4ft off the ground. North is along the positive Y-axis, which runs to the right (view is toward WNW).

A cone with a 100º peak, rotated northward until its edge is 25º of of horizontal represents Dishy’s field-of-view. I was lazy about tree modeling – they’re just cylinders rising to the measured height, at the correct location. I assume that if I can get the cone above the peaks, it will also be outside of the rest of the tree shape. Comparing this model to the obstruction view in the Starlink app, I think I got close. Each tree I see inside the app’s window, I see inside the model’s cone.

Dishy 10ft above my roof. The blue cylinder near it represents my chimney. The tree in the cone to the north is far more in Dishy’s field-of-view than this model shows, because many tall branches extend southward.

With hope that the model is correct, I started moving the Dishy cone around. My first question was whether I should plan to put Dishy on my roof. Unfortunately, there is a tree very close to the north side of my house, not to mention the east and west sides. The cone wasn’t clear until Dishy was ten feet above my roof. To confirm, when the ice had cleared, I climbed on the roof and pointed the app around. The view from the app made me think this might even be optimistic, as some branches of the tree reach over the roof, in a way that makes the simple tree-cylinder model too simplistic.

Dishy a few feet southwest, and 16ft higher than its original location.

If the roof mount was going to require a tower anyway, how much tower would I need from the ground? I began moving Dishy’s cone up from its original position. When I reached twenty feet, moving it a little south and a little west seemed to almost entirely clear the cone. If I raise the aim or narrow the field at all, as is expected to happen when more satellites come online later this year, the cone is completely clear.

So I should have just ordered a 20ft Rohn and put it there, right? Maybe, but, I’m not quite that confident in my mapping. My hacked sextant gave very coarse readings, made worse by the fact that many readings were over 45º, where sine grows faster than cosine. My compass liked to swing 5-10º between eye level and ground level. I think the model is a decent start, but I’m not willing to risk the permanent installation of a steel tower on it yet.

But I do like building things, and even though lumber prices are higher than usual, it’s manageable for small projects. How close to 20ft can I get?

3D model of a 18.5ft wood-and-pipe tower. Adding Dishy’s 16in stem places the center of the dish 20ft off the ground.

Pretty close, it turns out! A tripod made of 12ft 4x4s, with 7ft of a 10ft pipe sticking out the top, plus Dishy’s 16in stem comes to almost exactly 20ft. If something like this will work, it offers a few nice features:

  • The pole can be raised and lowered to make installing Dishy easier.
  • The pole or the legs can be extended, if just a little more height is required.
  • If ground anchors are enough to keep it from tipping, it’s possible to reposition the tower.

That last point is half of an important question: is this design strong enough? Luckily, radio operators have been mounting antennas on pipes for decades, so the engineering isn’t hard to find.

Total Wind Torque < S40 Bending Moment. Hooray!

Dishy’s stem has a 1.5in. outer diameter. Schedule 40 1.5in. steel pipe has a 1.5in. inner diameter, so mounting Dishy on such a pipe wouldn’t even require the backordered adapter. Is it strong enough? I chose what seemed like the worst case scenario: Dishy’s flat face pointed straight into an 80mph wind. Our winds generally come from the west (parallel-ish to Dishy’s face), and average high gusts are 40mph, so this should give us a good margin of error. Luckily, even in the worst case I’ve specified, math says that 8ft. of 1.5in. S40 shouldn’t bend. Hooray!

Total Torque << Ground Anchor Torque. Hooray!

What about those ground anchors? If I use the wind force already calculated, an calculate the leverage at the pivot point at ground level, I get almost 25,000in.lbs. I found ground anchors that say they provide 2250lbs. of holding power “in normal soil conditions.” Given that they will be 4.5ft from the pivot point, that comes out to over 120,000in.lbs. of leverage. That’s so much more than the wind leverage that even if my soil conditions are abnormal, I think I’m safe. (Yes, I have ignored wind torque on the wood tower as well. Given that it’s much lower to the ground, and the anchor torque is so much higher, I’m not concerned.)

If my well pump hadn’t failed just a couple of hours into construction, I might have had the tower up in two days. It took three. The tower was stable enough for me to lean my ladder against while mounting Dishy. We’ve only had light breezes so far, but Dishy doesn’t seem to wiggle too much.

Dishy powering up in its new home. View is to the southwest – these trees were outside even the original field of view.

So, the did-I-save-over-a-Rohn question: did I fix my obstructions? Almost. We had rainy days following installation, so obstructions have bounced around a bit. In the days before relocation, I was seeing over 35 minutes of obstruction per 12 hr period. Since relocation, I’ve seen as low as 2 minutes of obstruction per 12 hr period, and no higher than 12 minutes (during the thickest cloud cover). So it seems like I fixed at least 60%, and maybe almost 95%. Our leaves are finally starting to come in, so I’ll probably let this setup gather data for a bit before deciding how much farther to push it. Come back in a couple of weeks for my May outage data analysis to find out what effect this tower had on my connection statistics.

If you’re interested in building a similar tower, I’ve published my plans here: http://woodworking-plans.beerriot.com/signal-tower/.

Dishy looking (mostly) over the northward obstructions.

Woodworking Plans and OpenSCAD

Every time I post pictures of a project I’ve completed, someone will ask if I have plans I can share. I never do. I have sketches with numbers near them, but I am confident that no one would be able to interpret them. If it has been too long since I made the project, I might have trouble interpreting them myself!

I’ve started an experiment to correct my lack of sharable plans. Diagrams and how-tos for many of my most recent projects are now available at http://woodworking-plans.beerriot.com/. Other projects from this blog’s history should show up there in the future.

While I gather sketches and measurements of past projects, I think it’s also a good time to explain what tools I’ve used, and why. Most of them are new to me, so I’m hoping this post might generate some discussion on better ways to approach this.

What I’ve settled on for diagramming is OpenSCAD. It’s a 3D modeler, controlled by a programming language that supports basic shape manipulation. I chose a CAD system because I thought that, if I had a full model of the project, I could generate component and assembly diagrams from different, partially-completed views of that model.

I chose a 3D modeler that is programmable because … well, let’s be honest, a good deal of it is because programming is how I interact with computers. But the secondary reason is that I believe the model, itself, is not enough to explain how to build a project. Sure, someone could pull apart a model in whatever tool I used, and inspect it for themselves. But if the point of making the model is to explain the project’s construction, then the product of my process shouldn’t just be the model, but should also include descriptions of the model: diagrams of sizes and angles, and natural language telling a person how to make it.

The model isn’t going to generate natural language build instructions, but if the sizes and angles it uses are available in code, they can also be templated into English written along with the model. To accomplish the templating, I’ve chosen the Jekyll website generator. Via a small script, I can export variable names and values from the model, making them available to include in a templated webpage.

An additional benefit of programmability that I’m excited about is standard version control. That’s exciting because I can develop models iteratively, and improve things over time … and you can help me! The models, the diagrams, and the how-tos are all open-source on Github.

Since you can see my source code, that’s what I’d like to spend the rest of this post talking about. I started learning OpenSCAD only about six weeks ago. If you look through the code repository’s history, you’ll see how I’ve adapted my approach over time. Overall, it has been amazing how quickly I could get useful results from the tool.

There are also places I still feel like I’m fighting the tool. Most of these are places where I would really like the model’s code to somewhat read like a natural description of the creation of the project (start with a piece this size, cut this much off here, attach that other part there), but the details of making the tool render that clearly get in the way (rotate this around x and z, move it an infinitesimal amount to the side to prevent rendering conflicts, color this here so the cut is colored like so, by the way this can be animated). Finding the right abstractions are taking time.

Some abstractions are simple things, like getting used to expressing most things in vectors, instead of individual scalars along (or around) each axis. You can see that I learned that in the perfume display, and then forgot it when I started the toddler tower.

Other things aren’t so much abstractions as they are conventions. For example, which orientation should a component be described in? The way I would think about holding it while making it seems most natural in some ways, but the way it fits into the assembly seems most natural in others. I think the currently popularity of CNC and 3D printing means that most CAD models are described in the orientation that the machine will operate on them. Should I endeavor to describe my components such that they could potentially be made via CNC or 3D printing? Muddled in this decision are which way is up, and where should the origin point be?

Some abstractions seem like more complex concepts. Take, for example, these few notes:

  1. Nodes in the scene cannot be referenced by variables.
  2. No introspection can be done on nodes (size, position, color, etc. are all hidden to the language after creation).
  3. Modules, which look a little like functions in some other programming languages, can create nodes in the scene, but cannot otherwise return values.
  4. Nodes can’t be passed to modules, but there is a facility called “children”, which allows the effects of modules to be chained together.
  5. Functions, which also look like functions in some other programming languages, can not create or alter nodes in the scene.

These notes have strong influence on composability. You can write a module that creates a cube of a certain size, and you can write a module that moves whatever its children are up and to the right, and you can chain them together so that you get a cube of a certain size that is moved up and to the right. But, the mover module can’t base the amount that it moves the children on anything about the children. You have to pass parameterization information like that as arguments to the mover module.

Examples of how modules and functions can and cannot be composed.

It seems like thinking about nodes in the scene similar to the way one would think about side-effects in other languages is the near the right model. My struggle with it is part of why you’ll see many modules and many functions in each model. Since I want to make diagrams showing each component at different stages of its completion, I need the ability to selectively apply each stage. The best I’ve found so far is to define each step of the creation as another module, so that I can apply them in different combinations. That solution came after being unsatisfied by parameterizing the modules with “do this step” or “don’t do that step” arguments. Functions and variables for every value help to make it possible to keep the many modules in sync without threading all of the information through arguments, though it does make for a lot of names to keep track of.

There are a hundred other little things I’ve learned and experimented with along the way, I’m sure, but I’ll save them for another spew session. The OpenSCAD code is only part of the repository. There’s fun things like Liquid templating and the Pure.CSS layout framework that made building the website relatively quick, which I may write about some day as well. For now, if you have time and interest to look around, read some of the code, and let me know what you think. Or better yet, if you have time, material, and interest, have a go at building one of the projects, and let me know what you think of the instructions!

Making a Kitchen Knife

I once watched a YouTube video about making a knife. Since then, YouTube has recommended more knife-making videos to me than I can remember. This is probably why, when faced with the question of what to make for this year’s family gift exchange, I suddenly thought, “I could try making a knife!”

I traced the profile of one of my favorite knives onto an old table saw blade, following in the footsteps of many a ‘tuber before me, and set to work.

In the course of a couple of weeks, I learned a few things:

And at the end of it, I have two knives I’m quite happy with.

If you would like to fall down the rabbit-hole of hobbyist knife makers, I can also offer you a brand new starting point: my own story of making these knives.


Turned Kitchen Tool Handles

Pizza cutter, pie server, cheese plane, and ice cream scoop

I wasn’t kidding when I said the shape of my turned tap handle would be a familiar tool-handle shape. I turned the handles above for a Christmas gift this year. They’re an inch or so shorter than the tap handle, with different end diameters to match each tool, but otherwise they’re all the same idea.

Pretty maple grain
Ray flake catches the light

These also have threaded inserts instead of the maple being tapped with threads directly. That will be important for cleaning. These can be spun off of their tool ends easily. Those ends can go through the dishwasher, while the handles get a lighter hand washing.

Turned Tap Handle

I wanted a nicer handle for the tap serving seltzer from my kegerator, so I grabbed a piece of scrap maple and turned one.

I decided to tackle this video differently. Instead of full instructions on how to make the handle, I described the process mostly at a higher level. To help push me through getting it done, I also limited myself to a clip length of four seconds – find something interesting, say something interesting about it, and move on. This video took probably 20% of the time, or less, to edit. It has some rough edges that could have been polished further, but I think it helped me learn that a lot of the feel of the project work can come through, even when many details are left out.

Build a Folding Laundry Rack

I’ve been enjoying watching other YouTube builders recently. So, instead of writing up my latest project, I decided to make a video about it.

We left our old laundry rack with friends when we moved. But with cold weather moving in, we have once again needed a place to dry all of our warm wool clothing.

This rack gives lots of hanging space, with plenty of clearance for long garments. It also folds near-flat for easy storage.

Plans can be downloaded for free below. I know that SketchUp is the popular tool these days, but I haven’t learned it yet. The plans are closer to classic architectural drawings, which is the style I’ve used forever.

Let me know if you build this rack yourself, and what you change if you do!

UPDATE: New plans with step-by-step build instructions are available here: http://woodworking-plans.beerriot.com/laundry-rack/.

Electric Guitar

I have crossed another long-time TODO project off my list. From the earliest days of learning to play guitar, I knew I wanted to make one. It took me a long time to get started because I wasn’t sure what style I wanted. Instead of cloning of a common model, I decided instead to consider each piece and plan my own design. This quickly led to far more choices and choice-dependencies than I could handle. But this summer, with a little nudge from a friend who also wanted to build a guitar, I took the dive.

Some of my choices had already been made for me. Shortly after I decided to build a guitar years ago, a friend of a friend gave up on his own build. He offered the components he had to anyone who would promise to complete their own build. I’m lucky he didn’t put a time limit on the promise!

These components resolved my choice of pickup, tuning machines, and tail stock. A bridge-position humbucker, 3×3 tuners, and a 12″ radius bar-style tailstock also pointed toward a few common models to mimic. In particular, the Les Paul Jr. caught my eye. My friend also building a guitar just happened to have a regular Les Paul that I could use as reference.

I’ll spare you the step-by-step, since there are already myriad videos of guitar builds to watch. But, I would like to talk about a few of my favorite parts.

To start, all of the wood in this guitar is scraps from previous projects. The neck is leftover maple from my sleigh bed, with inlay made of cherry from my dresser. The body is also cherry from the dresser, combined with mahogany from my coffee table. It was a fun challenge to figure out how to work with what I already had on hand, and makes me feel justified in having carted that wood across the country.

I’m particularly proud of the neck. I chose to make it out of a single, solid piece of maple, instead of gluing a fretboard and headstock onto a neck stick. That took quite a bit of extra planning. I ordered the work on the neck to make each step as easy as possible. For example, I cut the fret slots first, so that the stock was still square, which made lining up the very precise cuts easiest.

I installed the truss rod right after that, from the back, through far more stock that would remain later for the same reason. It was much easier to route the straight, deep slot through stock that would be removed to reveal the headstock than it would have been to route with a headstock sticking up near one end.

And so on. In fact, working on the neck at all was part of the planned order. I wasn’t using a standard neck template, so I didn’t have a template to route the neck pocket into the body. Instead, I finished enough of the neck to define the heel end, then used the usual mortise-and-tenon techniques to mark and chop the pocket out of the body blank before cutting out the shape of the body.

I have mixed feelings about the inlay in the neck and headstock. On one hand, the end-grain of the cherry darkened with the finish, and provides great contrast with the pale maple. On the other hand, one of the shapes I chose was too intricate for my chisels, so it’s a little gappy in its setting. I also cut on the wrong side of my line in the headstock, blowing the entire design. These are both errors that could be fixed by chopping the inlay out and relaying something else. For now, it’s just reminders of the need to practice and pay attention.

The body was a fun experiment. Neither my leftover cherry, nor my leftover mahogany was thick enough to form a guitar body on its own. So, what I did instead was sandwich a mahogany core between two layers of cherry. The cherry also had some nice figuring in it, so I resawed my planks, and bookmatched them. I’m very happy with how the cherry looks, and also with the contrasting strip of mahogany down the middle. From the side, it almost looks like I put edge-banding around the front and back.

Somewhat accidentally, a component of the Les Paul that I copied unintentionally is its weight. This guitar comes in at almost exactly eight pounds. I thought about cutting hollows in the mahogany before gluing up the sandwich to reduce that weight. But the potential complication of needing to remember exactly where those hollows were, when I wasn’t yet sure of exactly where the neck would sit, was a hinderance to just finally getting the project started. Eight pounds is the heaviest guitar I own, but it doesn’t feel too bad on my shoulder.

If you’d like to see some more pictures of the build process, please follow me on Instagram @willthatwork. Instagram is something new for me. I’m still pretty uncomfortable with its features. I was hoping I might be able to use it to find more of an art network, to contrast my mostly tech twitter network. We’ll see if I stick with it. For now, it is also where you will find a demo of this guitar:

I have to give one final shoutout to Pete and Andrew. I’m between shops at the moment, and they offered me time in each of theirs to complete this project. It’s awkward to work in someone else’s shop. Even if the tools are great, learning which are available and how they like to be used takes time. Pete and Andrew were each extremely helpful and tolerant of me adapting to and adapter their workspaces.

Dresser: Finished

I’ve pushed off writing about progress in the past few weeks, for the practical reason of spending that time in the shop, and for the vain reason of keeping secrets before a big reveal. Last night it became possible to dispense with both reasons at once.

It’s finished! It’s in place. Drawers are filled. The shop can move on to its next project. But before that happens, let’s catch you all up on the intermediate progress.


If you missed the first several steps, these posts will catch you up:


Picking up where I left off, I mounted the drawers on ball-bearing slides. The drawers are just short of 18 inches deep, so I used 16 inch full-extension slides. The small overhang of the drawer above feels natural, like the bit of drawer left inside in a traditional design.

I used Rockler’s slide-mounting jig to ease installation. The only complication I had was matching the flush alignment on the internal dividers to the set-back on the external walls. The top four drawers are essentially inset on the left and overlay on the right. To manage this, I cut a small block to the depth of the inset, and then placed that in the jig ahead of the slide.



With every project, there are intermediate points where things actually look pretty good, and I wonder if maybe I shouldn’t continue with the rest of the plan. At this point, I kind of liked the highlight of the birch next to the cherry. Maybe I didn’t need to continue making faces. A friend even remarked that the blue tape temporary pulls matched well. Continue, I did, though.

The simple face solution would have been to run a plank horizontally across each drawer. I though this would break up the appearance too much, though. It wouldn’t be the simple solution for the cabinet door, either. Instead, I chose to run planks top-to-bottom.

My initial plan had been to glue up this whole panel, and then cut each drawer and cabinet face from it. But, with the assembled dresser taking up space, and the need to keep a path clear for laundry, there just wasn’t room. Instead, I very carefully labeled and cut each piece for each section, and glued each drawer face together individually.

This required some extra attention to alignment along multiple axes while clamping, but I think you’ll see that it all worked out.


I delayed any choice on handles for a long time. Cherry or an accent wood dovetailed into the edge? Leather loops, especially after seeing how the temporary tape pulls fit? I ultimately fell back on my second favorite project material: slate.

Four-inch long, near-square rectangular prisms: one-half inch top and back, Five-eighths inch front, and an approximate ten degree bevel connecting front to back on the underside. I drilled holes an inch to either side of center, into which I gorilla-glued insert nuts.

I used a flagstone sealer to give them a richer tone, and a smoother feel. It’s the same sealer that I used on my coffee table a few years ago, and it has held up well there.


I chose “dark antique” brass butt hinges for the cabinet door. Only a small portion of the hinge is visible, but the dark finish matches the slate well.


Mortising hinges is the second technique I practiced with my box project last spring. The process here was the same: use a marking gauge to layout the cut, cross-chop and clear waste, and lay in the hinge.

To keep the door closed, I embedded a magnet in the door, and a matching one in a small block installed in the case. A one-half inch forstner bit made a perfect hole for neodymium magnets, tacked in place with a dab of Old Brown Glue.



I’ve learned that for eight years, the note about the simple mineral oil and beeswax finish that I used on my bed has had the ratio reversed. If the way is was written, one part oil to four or five parts wax, is correct, then I didn’t pack the wax into the tablespoon at all. I think it’s more likely that the correct ratio is four parts oil to one part wax. That mixup is likely what caused so much trouble finishing those toy blocks last year.

Since I didn’t realize the error until mixing a large batch at the wrong ratio, I had limited options to recover. So, this project’s finish is a two-to-one ratio of oil-to-wax. It took a little more elbow grease to smear on again, but it did smear this time.

In preparation for wax, I first sanded everything to 220. After that, I ran a damp cloth over the wood to raise the grain. When it dried out, I lightly sanded to 320 grit. At this point, I applied a light coat of straight mineral oil. My thinking here was to get the saturation started in the wood, so that fewer coats of oil with wax would be needed. When the oil had soaked in, I lightly sanded to 400 grit. Finally, I applied two coats of oil and wax finish over an eight to twelve hour period, and then buffed off the excess with clean microfiber towels.

I had been a little worried about these dovetails. They’re good, but not perfect. With the wood dry and pale, the gaps were kind of obvious. Oil and wax swelled, darkened, and filled everything. I’m quite happy with them.

In place

We moved it in and transferred my clothes as some final touches were curing. It felt enormous in my garage, and it feels large in comparison to the small dresser it’s replacing. But, I think it does fit the space.

I’m not a great photographer, but I think that some of the curl can be seen catching the sunlight from the window here.

Many of the edges also have a beautiful ray flake that gleams as you pass by.

My sweaters now have a home, instead of piling up on an ottoman nearby. I worked in two small drawers for accessories, including one protected by a lock. These were my chance to include classic techniques as well: horizontal grain orientation (still matched across faces), and wood-on-wood slides.

My first picture in this project’s album, of the wood loaded in a trailer ready to take home, was taken on April 5, 2019. I officially said there was nothing left to do on March 11, 2020. Just over eleven months to complete this project broke down as roughly four hours per week (half of one weekend day) for the first nine months, followed by six hours a day, five days a week for the last two months. That comes out to nearly 400 hours of work. I think it was worth it.


Dresser: Drawers

It turns out that there’s a reason everyone recommends building a crosscut sled for your table saw: they really are very nice. I spent the past several days making a few, and it made building my drawer boxes a breeze.

“A few” table saw sleds? Yes. I wanted a full-width sled mostly to fully support some of the larger pieces I would be working with. It also made for a good solution to the problem I have with my table saw: the inset for the throat plate is so shallow that it’s difficult to make a zero-clearance insert for it. I also made two half-table sleds! One for the left side, and the other for the right. The left sled allowed me to use my dado stack without ruining the zero-clearance kerf of my full-width sled. The right-side sled I haven’t used yet, but I imagine it being the small-piece tool after I’m eventually forced to use the left side with the blade leaned over in an acute miter.

Pictured above is the extension I made to cut box joints using the left-side sled. This is another YouTube gem that I now understand why people love. Once tuned, it cuts very even, easy box joints. A testament to this fact is the first picture above, of my top drawer. If you look closely, you’ll notice the corner in the front is labeled C2 on one edge, and C1 on the other. I assembled the side of one drawer with the back of another without noticing – it’s a perfect fit.

I also have to praise the five-cut method. I used it to square my fences, and didn’t appreciate how good it really was until it was proved to me how much more accurate my cuts were that the factory cuts on the ends of these pre-finished baltic ply drawer panels. The box corners that I cut fit like a glove, but the ones where I trusted the factory were off just enough to be annoyingly tight.

In short: sometimes the internet is right. Who knew?