When I was just starting out, I joined the “Poly-everything” crowd. It was a long time ago. It seemed to me that polyurethane was all anyone was using for a clear topcoat. I loved the durability. I still do for projects where polyurethane-level protection is warranted. Unfortunately, poly is a pain. For one thing, it dries SO SLOWLY! For another, it doesn’t play well with others, or even with itself. For yet another, repairs are difficult. But if you need the protection, there is nothing better.

Polyurethane’s downsides are on full display if you are seeking a glassy surface on open-grain wood. I made three walnut doors for our master bedroom. They are sliders (not swingers), so structurally speaking, I was free to make them light. I went with a wink toward shoji screens with rice paper panels that appear to be suspended in their walnut frames. I wanted a pretty formal looking frame around a more organic, natural looking screen. To me, that meant a glassy finish on the frame with barely-finished screen dividers. This article is about getting a hard-as-nails, glass-like finish on walnut. I’ll cover some of the tricks I picked up while finishing the frames, including the use of foam brushes.

Finishing options to get the result I wanted (of which I am capable) are spray lacquer, French polish, or polyurethane. Full disclosure, this is the only project I’ve ever started in my shop that I set aside for any length of time before completing. This door is for my wife’s bathroom. When the old swinging door came down during a remodel, she made this crazy demand that I produce a new door for her bathroom reasonably quickly. I made all the doors at same time but only finished and hung the one on Her Majesty’s bathroom. The other two are for closets, which don’t actually need doors at all. So I set my two unfinished doors aside. Eight years passed.

Why the backstory? Because 8 years ago I was not able to spray lacquer and I didn’t know how to apply French polish. I finished that first bathroom door in polyurethane. All three doors will hang virtually next to one another. So now, like it or not, I have to do the other two the same as the first.

This process is not original, but here’s how I did it. First, I waited until I had several weeks when there were no other projects happening or planned for the shop. This is a big reason for the 8-year time lag. This is critical, because once all the prep is complete and you’re ready to apply a coat of finish, the routine becomes a once-a-day affair. A quick scuff-sanding, let the dust settle, a coat of poly, and walk away. And you can’t do anything else in the shop because you want to work as cleanly as possible. One of poly’s greatest challenges is its lust for dust. It takes a lot of patience and discipline to deny it one of its favorite things.

I considered setting up the shop so I could finish both sides of both doors at once. Perhaps a flippy-type thing where the doors were suspended between a couple step ladders so I could walk around them and flip them over and keep on working. That would have taken up a huge amount of shop space, introduced more dust management issues, sag management issues, and more risk to the finish I’d just applied during the flip. I decided instead to embrace this as an exercise in patience and finish one side of each door at a time. Turns out, each door took about 8 days per side, multiplied by 4 sides, so about a month altogether, typically working about 30 minutes per day every day (plus time for dust to settle). Patience, indeed.

I’m a fan of foam brushes for brushing varnishes. Thomas R. Shrunk wrote a great article on how to apply fine finishes with foam brushes (FWW#116, pg.116). Being a supreme hater of cleaning brushes with solvents, I was an immediate disciple. With a little experience, I have found the lowly foam brush equal to any other brush I own for a finish like this. I would much rather waste a stack of single-use brushes than mess around with solvents every day for a month.

In broad strokes, the process is to finish sand the project to at least 220, then fill the grain. Sand and reapply grain filler as necessary to fill the wood pores as well as you can. Apply coats of poly and sand them down until you establish a perfectly level surface. Apply one perfect show-coat. Rub it out with steel wool and wax.

As with any finish, prep is key. I’ve never regretted spending too much time prepping for a finish, but I’ve experienced lots of regret from prepping too little. Sanding is not fun to me, but it is very efficient compared to stripping finish. Spend the time to sand to perfection. Spend regret on something else.

Cleaning the environment is an important part of the prep process for a finish like polyurethane because it stays wet so long. Clean the shop. Keep it closed up if you’re finishing during the flying-critter months. If fumes are an issue, it is easier to wear a respirator in a closed space than it is to sand out a bug in an open one. The less activity there is in the shop, the less dust will mysteriously appear in your varnish.

Grain filling is messy and tedious. You can skip this step and simply apply more coats of poly until you build up a smooth base, but grain fillers will speed the process. Behlen makes a good grain filler. I was recently turned onto Aquacoat and there are things about it I like and things I don’t. I hope soon to try old-fashioned Plaster of Paris and write an article just on filling grain. Regardless the product, I’ve learned that getting a fully filled surface for top coating isn’t easy. I’ve accepted that all I can expect from Behlen and Aquacoat is to get a nearly level surface that can be leveled by building up the top coat. Maybe you’re better at filling grain. Who knows? My advice is to do your best and move on. Two passes with grain filler is about all I have patience for. Aquacoat is hard as stone, making sanding a bitch.

These photos show me troweling on Aquacoat with a credit card at an angle to the grain. After about half an hour, I follow the first coat up with a ragged-on second coat 90 degrees to the grain. When the grain filler is fully cured, I sand it down to the most level surface I can get at this point.

One tip I’ll share is to resist the temptation to level your grain filler with a random orbital sander. It is simply too easy to grind through the filler and re-open the grain. Hand sanding is real work on a product as hard as Aquacoat but sanding through the filler is defeatist.

With grain filled and everything re-sanded, I apply my first coat of poly. The urethane products I use are the venerable and ubiquitous Minwax Fast-drying (hahaha) Polyurethane and Minwax Wipe-on Poly (both in clear satin). I follow Shrunk’s advice, laying down “puddles” of product to be absorbed and distributed by the foam brush as it passes over the surface. Using overlapping strokes, I make one single pass at a very slow speed- I’d say about 1 second per inch or a tad slower- trying my best to keep the brush moving smoothly. Visualize the foam drawing in and letting out the varnish as it passes over the puddles and you’ll get a feel for how slow to go. Try not to move in a jittery stop-and-go fashion. When advancing from a finished area to an unfinished area, come down on the work into the wet section with a gentle “airplane landing” technique.

For horizontal surfaces, I lay it on thick during the building stage. For vertical surfaces that cannot be rendered horizontal, the process is the same but with much less product to avoid sagging. Less product equals more coats, which equals more days… you get the picture.

The next step is to walk away. Secure the space to reduce dust being kicked up. Plan ahead- my shop is also my garage. I back cars out before I apply a coat of finish if I know we’re going somewhere. I don’t want the garage door going up and down over wet varnish.

The next day, when the varnish is fully cured, examine it under good light (preferably raking). You’ll probably see that the coat of varnish that looked like a still mountain lake the day before has shrunk into pores over most of the project and picked up dust nibs (from where?!?). That’s a normal part of the building process. Give the whole project a light scuff-sanding with 220.

I like to vacuum my project and then walk away for a couple hours to let any dust I cast into the air settle. When I return, I wipe the surface down with my hand, cleaning my hand on my shirt often. The oils on my hands do a great job of picking up dust. After several passes, and when I can’t feel anything foreign on the project, I lay down another heavy coat of varnish, pitch the brush and walk away. Being disposable makes the foam brushes really shine.

The next day you’ll probably see more shrinking, but less than yesterday. Depending on how nearly-level your surface is, you may want to see whether two heavy coats is enough build to sand them down to a perfect surface. I use my random orbital sander with 220 for this, paying close attention to how much sanding it takes to get a perfectly even scratch pattern. Shiny low spots stand out like Bic lighters at a Dead concert. If you’re sanding more than just a little to get rid of the shiny specs, abort. Scuff the project and apply another build coat. Protect yourself from sanding through to raw, open-pored wood.

You’re done when you’ve built up enough varnish to sand it to a perfectly even surface, free of shiny spots and without getting into the wood. At this point, I like to ramp up the grit size. From 220, I might jump to 400, then 600, then 1200 or even higher.

The finer grits help with my final application- a heavy, brushed-on coat of Wipe-on Poly applied in exactly the same way. Wipe-on Poly lays down faster because it contains a lot more solvent. It also dries faster because of the higher solvent content. In a clean shop, I can brush on a coat of wiping varnish and return the next day to a surface ready for 0000 steel wool and wax.

If the final coat isn’t ready for waxing due to some defect or bug, scuff it again and brush on another coat of wiping varnish. The waxing step is more for feel than appearance, so you want the appearance to be pretty much perfect. You can’t apply another coat of poly over wax, so waxing is a one-way gate. Be sure you’re ready to pass through.

This is not a hard finish to apply. It is a tedious finish to apply. It is an exercise in patience, and it renders the shop useless from start to finish (pun!). If you have the patience and you want the appearance and durability of glass, the results are undeniably spectacular. I sincerely hope you don’t wait 8 years from the time you build your project until the time you apply the finish but if it takes that long to find a month to apply this finish, you’ll be forgiven.

I don’t trust people who like sanding. To me, that’s a sure sign of a serious character flaw. I’d rather clean a public toilet than spend a day sanding. That’s the truth, and I’ll bet there are lots of woodworkers out there who would agree. It pays to sand as little as possible then, right?

Sanding is almost universally viewed as a necessary evil, so it makes sense to sand as little as possible. Any bonobo can rub rough paper on wood, but there is actually quite a lot going on between the sandpaper and the wood. If you become knowledgeable about sanding, you will find it is less painful and your results will be much better. Win-win.

Equipment

First, buy good sandpaper. Reasonable people don’t expect a $20 plane from Lowes to perform like a $300 plane from Lie-Nielson. Same for sandpaper. I like 3M Pro Grade No-Slip Grip. This paper cuts fast and lasts a long time. The backing is almost as tough as fabric. The no-slip grip has pros and cons. It’s really nice if you do a lot of fingertip sanding. If you do a lot of profile sanding with folded sandpaper, the heat generated from sanding will glue the paper to itself. When attaching this paper to a block, you must be careful that the tacky backing doesn’t pick up any debris that will create a high spot on the block (and a low spot on your project). These characteristics are easy to get used to. Bottom line- find a high quality sandpaper you like and lay in a lot of it.

https://www.amazon.com/gp/product/B00M12MTGI/ref=oh_aui_search_detailpage?ie=UTF8&psc=1

Next, get the right sanding tools. Make or buy good sanding blocks, good profile sanders, and make or buy a good sandpaper cutter. Invest time and effort in eliminating fuss and drama around sanding and it pays big dividends when you’re sanding your way down the home stretch of an exhausting project. My favorite sanding block is the Preppin’ Weapon, and I have four of all different colors. Yes, they’re expensive, and yes, they’re worth it. I load each of my weapons with 4 layers of sandpaper (a full sheet cut into 4 strips) so I can tear off a dull sheet and keep on sanding 4 times as long. Each of my 4 weapons has a different grit- 100/120/180/220- which is my preferred sanding sequence for most projects. Having all of my typical grits available at once adds a level of speed to the sanding process that rivals a power tool. In the photo you can see my four Preppin’ Weapons. There is also a shop made Melamine sanding block with cork on the bottom, a pile of rubber profiles sanders from Rockler, a couple pieces of PVC covered with self-adhesive sandpaper, and a strip of sandpaper with painter’s tape backing and a couple matchsticks to be used like a shoeshine rag. The tape mitigates tearing and the matchsticks enhance grip.

My sandpaper cutting station is as simple as it gets. I have a piece of ¼” MDF about 18”X18” to back up my cuts. I have another piece of ¼” MDF the exact width of a ¼ sheet of sandpaper and about 12” long I use as a template. I have a utility knife dedicated to cutting sandpaper (because it goes through blades fast). I align my template piece on the edge of a new sheet of sandpaper, cut it on my backer board, repeat thrice, and I’m ready to load 4 fresh layers of sandpaper onto a block. Simple, fast, accurate, dirt cheap and easy to store.

The Physics of Sanding

Like most woodworking, effective sanding requires a combination of knowledge and skill. Let’s begin with a little information about how sandpaper works.

Sandpaper is a cutter. It is made by gluing an abrasive material to paper. Each grain of abrasive on a given sheet of sandpaper is similar in size, and the number of pieces of abrasive in a particular area (like a square inch) establishes the “grit”.  The lower the grit number, the coarser the paper and the more aggressive the cut.

There are two different grit scales- CAMI and FEPA. CAMI paper is marked with a “C” and FEPA paper is marked with a “P”. While it may be helpful to understand there is a difference, I am giving you permission not to care. Take a look at this chart and you’ll see what I mean. For woodworkers, there is very little difference between CAMI and FEPA at the grit grades we use.  I think of CAMI like miles-per-hour and FEPA as kilometers-per-hour. The difference between the two is marginal at slow speeds. The faster you travel, the more disparate the scales become. The difference in grit size between C and P is marginal up to about 220, then things begin to diverge. The difference between 2000C and P2000 is almost 3X (P2000 is roughly equal to 700C). Now you know far more than you ever needed about CAMI and FEPA.

Sandpaper is self-sharpening. Each grain of abrasive on the paper is friable, and the grains are constantly breaking and revealing new cutting edges as the paper grinds against the work. Eventually, you exhaust the paper of cutting material and the paper stops cutting.

Don’t be fooled into thinking that sandpaper becomes smoother with use. Worn-out 100 grit paper does not effectively become 120 grit paper (or higher). You do not get smoother results from well-worn paper as if you changed to a higher grit. If you’ve done this (and I think we can all admit that we have), it is time to acknowledge that you were fooling yourself, like cheating at Solitaire. If the paper stops cutting efficiently, it is worn out. It has not become something different.

Take it easy. Using moderate pressure and a moderate pace will help get the longest life and best result from your sandpaper.  The adhesive that holds the abrasive to the paper does not like heat. Heat will build up faster if you push hard, or use dull sandpaper, or sand at the pace of a twitterpated bunny rabbit. As with any cutter, let the cutters do their work and temperatures will stay low, grains will fracture, dust will clear, and you will sand away more stock in less time and with less effort.

The Sanding Process

There isn’t a great deal of skill involved in sanding, but mastering the basics takes much of the drudgery out of it. Sure, you need be able to hold a tool level and keep pressure even over the length of a stroke, but sanding is really more of a process than a skill. Like any good process, it needs to be followed. You can’t cheat the process without introducing inefficiency, frustration, and poor results. Honor the process and sanding becomes easier and the results are better. I’m not saying you’re going to love sanding. I wouldn’t trust you if you did.

In broad strokes, sanding is the process of removing mill marks and scratches on wood through the use of progressively finer grits of sandpaper until the project is smooth enough to finish. An efficient sanding process begins with the most aggressive grit necessary to remove mill marks reasonably quickly. Note I said mill marks and not defects. Chips, tear-out, dents and dings are better addressed before sanding. Sanding is rarely the best option for stock removal. Prep projects properly for sanding and the sanding process will go more smoothly (ha!).

Step One is to mark up your project with a pencil. I use a wide construction pencil and make a tight sine wave wherever I’m sanding. Starting with my coarsest grit, I sand until all of the pencil has been erased. Check for mill marks and if any remain, mark up the project and repeat Step One. I know I’m done with the initial grit when all mill marks and pencil lines are gone. Mark up the project with a pencil between every grit to manage your progress and minimize time and effort spent on each grit. If this is new to you, try it. I think you’ll be astounded at how quickly you can proceed.

My milling machines consistently deliver a surface I can quickly sand flat with 100 grit paper. Lots of folks start with 80 grit. The big grit paper is the workhorse. It is a fool’s errand to use finer grit papers to remove mill marks. Let the big workhorse do the heavy pulling and the journey through the higher grits to your final surface can be reduced, literally, to a few strokes of each grit.

Your project is probably at its least smooth after milling and jointing, so the bigger grit paper has to remove more stock before it reaches the point where it is reproducing the same surface as you sand deeper and deeper. Stop sanding as soon as the mill marks are gone (along with your pencil lines), joint seams are level, etc. The only work you should leave for the next grit is to remove the scratches left by the bigger, previous grit. The goal is the same for every grit- stop as soon as you have removed the scratches from the grit before, as indicated by the removal of your pencil marks. Eventually, the routine emerges: Mark the work with a pencil. Sand away the marks and stop. Clean the surface. Change to a higher grit. Repeat. Don’t skip grits. I’ll bet you’ll be surprised how little time and effort it takes to advance through the grits if you’ve done a good job at 80 or 100.

I usually take each surface, like the sides of a box for example, through all the grits I’m going to use before re-positioning the work and starting again. On some projects I sand the entire thing with 100, then 120 and so on. When that’s the approach, I always mark the next grit to be used on each surface so I don’t lose my place.

The work needs to be cleaned between each sanding grit. As you sand, the broken bits of abrasive get intermingled with wood dust. Realize that 100 grit abrasive is still 100 grit when it is fractured from the paper and mingled with wood dust. If you don’t clean all that abrasive off the work, then you will still be sanding with 100 grit abrasive bits when you change to 120, 180, 220, and so on. Put simply, if you don’t clean between grits, you aren’t changing grits. I use my shop vac with the brush attachment and thoroughly vacuum all the dust between grits.

I rarely need to start rougher than 100, and rarely exceed 220 on the smooth end, often stopping at 180 depending on the finish. It is a lot of work to advance through 4 grits of sanding, but each grit goes very quickly. For example, one side of a square table leg shouldn’t take more than 20-30 seconds of sanding at the 120, 180, or 220 stage. It takes longer to secure the work, mark it and clean it for each grit than it takes to sand it.

Sanding hard-to-reach surfaces is inevitable sometimes, but good planning goes a long way. Finish sand everything you can before you assemble parts that will render some areas hard to reach. Take care sanding around joinery before assembly so you don’t queer the fit or open a gap. One change I’ve noticed as I’ve gained experience in woodworking is that I do a lot of finish sanding just after joinery and before assembly, and oftentimes a lot of pre-finishing, too. Over time it has become intuitive and nowadays I rarely find myself trying to sand or finish inside small spaces. I’m also getting more intuitive about sanding some components before joinery, as some joints are more apt to get spoiled by sanding than others. A perfect dado joint can be rendered useless by a few strokes of the sanding block. Rabbets are more forgiving.

Sanding fancy profiles, edges and corners takes care so that you don’t break an edge you want to remain crisp. Even a fine grit can remove enough stock on an edge profile to make a once-stunning shadow line disappear. This type of sanding work is best done with profile sanders to back up your paper. Sanding blocks and holders are available in typical molding profiles, but it is also easy to make your own. Shop-made blocks work best with a semi-rigid backing like cork or leather. I keep a roll of real cork shelf liner in the shop for this.

Power Sanders

Sanding techniques change a little if you introduce a power sander. Power sanders include dedicated spindle sanding machines, drum sanders, drill press-mounted sanding drums and random-orbital sanders (ROS). I may get into sanding with other power sanders in future posts, but I’m only going to cover the ROS here.

An ROS is likely to be the first power sander most woodworkers acquire and for many, it is also the last. An ROS is very good at what it does and most woodworkers simply don’t need anything more. ROS’s are not the “jitterbug” sanders of old. Jitterbug sanders simply aren’t capable of producing the result you need for a fine finish. If you’re in the market, check the multitude of tool reviews available for ROS’s and get a good one. Quality counts extra when you’re doing finish work.

As the name implies, an ROS sands in random circular patterns. Sanding in circles obviously means sanding cross-grain half the time. ROS’s solve the problem of cross-grain scratches by making them too random and too fine for the eye to perceive. The randomness scatters light and tricks the eye into perceiving a smooth surface, but make no mistake, ROS’s leave circles and swirls and sometimes you can’t tell until you apply finish. Surprise!

Many woodworkers will use their ROS to work up to their last grit, then sand the last grit by hand to sand away the ROS’s circles and swirls and align final sanding scratches with the grain. This is a fair way to get the best of both worlds if you love your ROS.

As far as the physics go, sandpaper works pretty much the same whether you rub it against the wood by hand or by machine. Grains fracture and material is removed with successive grits until the desired surface is created for finishing. Paper wears out and must be changed. One advantage over hand sanding is that if you are using good dust collection, the ROS cleans as it goes. You can forego cleaning between grits, although a quick wipe of the hand is never wasted.

Sanding with an ROS takes more care than hand-sanding to protect the work from damage. These machines are powerful enough to seriously mess up a project before you can stop them. Here are a few rules that will help you sand effectively and safely with an ROS:

1. Always start the ROS in contact with the work. Never drop a running ROS onto the work.
2. Take care lifting a running ROS from the work. Lift straight up. Better yet, turn it off before lifting.
3. Set the ROS aside on a safe, secure surface, or allow it to stop completely before setting down.
4. Use dust collection with an ROS.

The jury is out whether using an ROS has any real advantage over hand-sanding. I used to believe it was a lot faster to sand by machine, but I have developed doubts after refining my sanding process. I have an ROS, and I use it based on the project. I’ll grab it to sand something big and flat like a dining table except, perhaps, for the final grit, but it isn’t a big deal to sand a large project by hand if you follow the process. I don’t like the noise and I always wear ear protection when machine-sanding. Machines do not perform elegantly like hand tools and I don’t find woodworking by machine to be as soul nourishing. If I sand up to an edge or inside corner with an ROS, I don’t like how much work is left to be done by hand and how difficult it can be to blend the work. On the bright side, I really like how clean it is to sand with an ROS. Bottom line, I judge an ROS to be a useful tool, but not critical. When I bought mine, I thought it was a borderline requirement to produce good results in a reasonable amount of time. Now, if I didn’t already own one I might think hard about making the investment.

Conclusion

I dislike this idiom, but sanding is what it is. Sanding is not sexy but fine finishes are, so maybe sanding is sort of like pregnancy. We aren’t likely ever to free ourselves from having to sand, but that doesn’t mean we are powerless to do anything to make it suck less. That’s why it’s worth it to:

1. Better understand what sanding is and what it does.
2. Refine the process and thereby, sand easier.
3. Learn how to sand less (Yay!).
4. Bring something beautiful into the world by sanding well.

Most of my articles include step-by-step instructions for readers who like what I’ve made and want to make something similar. This article is a little different. I’ll get into some instructions, particularly for procedures I think went really well, but for the most part this article will focus on the big idea of making your own custom molding. If you decide to make your own molding, you’ll figure out the processes and sequences easily enough. The goal of this article is to help make sure you know what you’re getting yourself into.

There’s nothing like the look of well-crafted custom molding. It’s the cummerbund and spats of trim carpentry. Great molding never shouts out for attention. Like a plain girl who transforms into a stunner by letting loose her hair and removing her glasses, great molding is happy to wait to be noticed. And when you do… bow-chicka-chicka-bow-bow!

My house is typical of many median-priced homes of the mid-80’s in my area. Whomever built our house installed above-average trim. The package included fairly substantial solid wood base with base cap, liberal use of heavy chair rail, and decent crown throughout. The custom solid oak kitchen cabinets they installed were okay, but are now in a landfill. All lower-level interior doors are solid wood 6 panel jobbies. The finish crew stained and varnished everything. The trim carpenters were capable and the bits of trim we haven’t replaced still look pretty good after 30 years.

That’s the good news. The bad news is that the whole lower level was trimmed in fir. The varnish work is crude. The stain color is 80’s-orange and splotchy. There is no evidence of any finish prep other than filling nail holes. Mill marks are apparent throughout. The interior doors are solid wood, but the panels are glued up from material with no attention to color or grain pattern. In short, it’s a builder job designed to appeal to people who don’t know or don’t care what wood can do in the hands of someone who knows how to use it.

The opening line to so many woodworking stories.

If you’re a regular reader (thank you!), you may have deduced that I chomped down on a total kitchen redesign in 2015. I did the design work myself with some help from Her Majesty, I made the cabinets, and I did the lion’s share of the work myself (excluding framing, plumbing and HVAC). It was this kitchen project that set me on the journey of making my own molding. I simply couldn’t buy what I needed to create the impact I wanted, which is the opening line to so many woodworking stories.

My design aesthetic is Arts and Crafts. Whatever your architectural or design aesthetic may call for, I recommend a path similar to mine. I feel like the time and effort I put into learning the basic design elements of my aesthetic and how best to integrate them into my project paid off big time.

It’s tough to pull off Arts and Crafts with speed base. The essence of A&C is to show your work, and to show that your work is good. To get that look, I approached my molding like I approach a furniture project. I tried to think of the project as furnishing the walls, and as with that plain girl in the glasses, I knew the work had to be stunning when you noticed it.

I started by pouring over Google Images for A&C Molding. It didn’t take long for themes to emerge. Traditional characteristics include stained hardwood (usually oak), 3-piece headers (a board with a crown above and a bead below), plinth blocks, corner blocks, corbels, and presence. By presence, I mean the molding makes a statement. Most is quite large, but not Georgian-large. It’s interesting but never busy. Its presence is understated- the strong silent type.

Materials selection and acquisition is a big consideration. I am shocked every time I calculate the material needed for a trim project. Attribute that to my cognitive limitations if you want, but be prepared. Trimming a room takes a massive amount of wood. Many of the pieces will be big and unwieldy. Storage can be difficult. The cost can run away from you like Usain Bolt.

I tried to balance materials considerations in the best possible way for my situation. I have the tools to mill lumber, so I saw no need to pay for S4S. I saved money on board feet, but spent it in time and food- it takes a lot of calories to mill a big stack of oak. Storage limitations prevented me from laying in all the materials I needed at one time. That meant buying multiple small loads, which nixed the efficiency of having it delivered, which resulted in multiple trips to my preferred sawyer who is more than an hour away. Together with the time it takes to sort through stacks of boards, it takes me half a day to buy even a small load of lumber. That’s a big commitment and a massive distraction if you have a day job.

You will have to abandon several boards as sociopathic sons of bitches.

Wood being wood, getting long straight stock for things like door casing requires a lot of time and waste. The wood needs to acclimate. Expect curving, cupping, twisting, and all that other stuff that makes wood act like a slow-motion Latin dance champion. The only way I know to combat this is to start with a lot more material than you need in the end. I buy 5/4 rough stock to get a 3/4 board that is straight and stable, and I work the rosary hard in the process. It may sound like a lot, but 1/2″ of waste is quite a small margin over a distance of several feet. There will be failures and betrayals. Plan on making lots of small projects for years to come from whatever material you use to make molding. Every milling operation needs to be followed by a period of acclimation. My long pieces of red oak (~7’) took a couple months to mill. Read that sentence again before you embark on a project like this.

All the standard rules for milling apply. Flatten a face, square a straight edge to the flat face, plane the other face flat and parallel, then rip to width. Try to take equal amounts of wood off all surfaces and sneak up on straightness and flatness. For long boards, creep up on your final size over a period of several weeks, allowing the boards to rest between milling operations. You’ll need lots of extra material to mill off curves, cups and twists that appear between milling sessions. You will have to abandon several boards as sociopathic sons of bitches. Your goal is to end up with a board that is relaxed, straight, flat and stable in your environment.

You could buy S4S material from a variety of sellers, including big box stores. That will require a tremendous investment of time sorting through boards because you have no margin for wood movement. Also, a board that was straight at the store can change so fast when you get it home you’ll swear somebody pulled a switch-a-roo at the checkout counter. Then there is the cost… O holy night, the cost. In my world, S4S is for small projects and trust-funders. Mere mortals who work for a living cannot succumb to that siren’s song.

With my design work done and materials procured and milled, it was time to make some molding. Here are a few tips from my school of hard knocks for assembling trim work.

If you rely on metal fasteners injudiciously, you may regret it. Tacking together built-up molding is undeniably fast and efficient- which is why pro trim carpenters use nail guns- but it can be costly on the back end. I’ve learned there is no apparent end to the need for a little taper here, a little hollow there, to get everything to come together in a house that left the concept of “square” on the front lawn. Adjustments are quick and easy with a hand plane or a pass over the jointer so long as your work isn’t full of nails. Just sayin’.

I glued up 3-piece head casing without fasteners so that I could drill and plane with impunity. I also like having no holes to fill- not even the nearly-invisible blemishes left by a pin nailer. To reiterate- the ethos of A&C is to show off good work. Details matter, and the absence of fasteners makes a strong impression.

I have only two hands and I always work alone.

My side casing is fixed to plinth blocks at the bottom, and to the head casing at the top, using pegs and pocket screws. The pegs keep the components aligned and the screws keep them tight together. I hardened the screw pockets with CY glue. The joint can be disassembled and reassembled for transport from shop to site. This was one of the most valuable ideas I came up with on this project. I have only two hands and I always work alone. I cannot move a complete door casing through doorways, up staircases, and so on without damaging something. This simple, repeatable joint allows me to make the millwork in the shop, take it apart, put it together at the worksite, take measurements and make marks, take it apart and back out to the shop, as many times as necessary to get everything ready for installation. It takes a few minutes to make the joint, but the investment comes back with an exponent.

Another thing I’m glad I did was to install my trim with adhesives instead of metal fasteners. This takes a little courage, because one great advantage of nailing up trim is that it can be removed with minimal damage. I prefer to use adhesives to eliminate nail holes, but I know that whomever is tasked with replacing this trim in the future (including, possibly, future me) will cuss with vigor. That being said, most of the trim I have removed in my lifetime was caulked to the wall, which acts like an adhesive, often tearing drywall paper upon removal. Same problem, still with nail holes.

I use a couple different adhesives depending on the installation. Construction adhesive like Liquid Nails works great. I use it sparingly, just in case a piece needs to be pulled off. Another adhesive I like is VHB tape. I ran a few tests. VHB tape will tear the paper off drywall before the bond fails. That’s plenty strong for trim work, and taping up molding is fast, easy and clean.

You may be picking up on a theme here- that I don’t like metal fasteners showing on my work. If so, you’re right. Here are a couple other tricks I figured out to install trim without nails.

I assembled several runs of trim as units, fitted to the location where they would be installed, but unattached. This allows me to do finish work in the shop, and when it comes time to install, I can use fasteners that will be hidden behind base shoe, or I can use adhesives, or whatever. I have fairly long, complex runs of base molding that can be removed for shop work.

Another big bonus of this approach is that everything can be pre-finished in the shop. Whether varnishing or spraying lacquer, there is simply no way to achieve a finish on installed woodwork that will equal what you can do in the shop.

I have a number of openings that required jambs and casing. I assembled one side of the casing to the jamb using glue and biscuits. Then I fitted the assembly into the rough opening and shimmed it using glue and clamps, avoiding driving nails through the jamb. I taped the assembly to the wall with VHB tape. Using clamps, glue and biscuits, plus some more VHB tape, I attached the other side of the casing to the jamb and the wall and glued it to the jamb with clamps. The result is a cased opening with no evidence of fasteners. It’s a nice look.

It is a poignant illustration of what it means to be anal-retentive.

I have a number of doors that took special attention. In order for the doors to hang neutral and plumb, the jambs ended up out-of-flush with the adjacent walls. Applying trim to these door jambs would leave noticeable, unacceptable gaps between the molding and the drywall. This would typically be addressed with a caulking gun, but my experience with caulk is that no matter how good you are, that caulk line is going to crack from seasonal changes and the passage of time. To my eye, fractured and torn caulk looks worse than a shadow line. I made tapers and shims to fill the gaps and glued them onto the back of my molding, mating them to the profiles and grain directions of my case stock, resulting in nearly-invisible marriages of trim to walls and jambs. Yes, this took forever. Yes, it is a poignant illustration of what it means to be anal-retentive. No, I’m not sorry I did it. It looks friggin’ awesome. Besides, some of this millwork is stained. How do you caulk that? If you just answered, “they make brown caulk”, you’re reading the wrong blog.

My story is another in the age-old genre of cautionary tales of woodworkers engaging in carpentry. As so many before me have cried out, “There was no other way! It would have cost a fortune to hire this kind of work done, assuming there is even anyone who could have managed it! I had no choice!” Truth be told, it has been overwhelming. It has burned me out. I can’t say I regret taking on the kitchen and all the spinoff projects it produced, including this molding odyssey, but I will readily admit that it was (and continues to be) too much for one guy. I still have a lot of trim work to do, and it is going slowly. It’s slow in part because all the joy has gone out of it, but that’s material for a different blog.

This handy carrying tote is at home just about anywhere. Whether you fill it with gardening tools or bath products, it is a strong, lightweight and classy way to get stuff from one place to another.

A relative I like got married recently. I wanted to make a wedding gift, but knew nothing about their home, their décor, or their needs. I knew they share my passion for coffee, so I assembled a gift of coffee gear. Still wanting to give something personal and hand-made, I settled on making a tote to hold and present the coffee equipment. I hope this tote will find many other uses. I envision a handy tote like this working as a picnic basket, or full of garden tools, or even in the guest bath as a towel basket. Heck, I can easily imagine one of those Whole Foods hippies using a tote like this as a shopping basket.

I faced a couple design challenges. I wanted this project to be sort of primitive, like a peach crate, but I felt that as a wedding gift it needed to be somewhat refined. I stuck with the crate concept and invested time in the joinery to give it some polish. If I’m honest, I also wanted the joinery to show that I have skills.

Had I the time and experience, I would have steam bent the handle for strength and for a traditional look. Alas, I don’t know how to steam bend yet and I don’t have the equipment, so I used what I know and you see the results.

A project like this needs to be light weight. It’s a holder of things, and the user’s effort should be invested in carrying the things, not the holder. Strength is important too, and the larger you make a tote, the stronger it should be (because it can hold more stuff!). High strength and light weight are best achieved by distributing stress thoughtfully and employing joinery effectively. For example, a carefully placed and oriented dovetail can out-hold a lag bolt without adding a single gram of weight.

Choose a light weight wood that is suitable for whatever use you envision. If your tote is going to sit outside in the weather, cedar or cypress or redwood make sense. I chose radiata pine for its balance of strength, weight, stability and workability. I had a nice, clear board on hand that was just the right size to yield all of the parts for this project.

For my needs, I made this about 18 X 12, with sides about 7 1/2″ tall, and about 16″ tall at the top of the handle. This was just big enough to hold the gift and it is a versatile size for re-purposing.

I began by milling the slats for the sides and bottom. I needed about 18 altogether, so of course I made 20, each about 5/16” thick and 2” wide. Lengths varied based on the most efficient use of my board. I was able to resaw my ¾” S4S board from the Big Box store in half, yielding a lot of slat material from a fairly small board.

Next I milled my four corner stiles and the vertical supports for the tote handle. These are all from ¾” material. The corners are ¾” square. The handle supports and the handle itself are ¾” X about 1 ¼”.

Cut the four corner stiles and vertical handle supports to final length in preparation for joinery. The side slats will be let into the six vertical members for strength. The side slats will be mitered for finished, elegant corners. It is critical that all the stock removal for joinery from all six vertical components be aligned.

I kept my joinery aligned by cutting all the components at the table saw, using a crosscut sled and a pair of stop blocks. Nibble out most of the material, then sneak up on a snug fit for a slat by tapping the appropriate stop block until the fit is just right. Tighten the clamps and mill all six components. Note that the corner stiles get milled on two adjacent sides. Stay focused and make sure you mill the correct side! Adjust your stop block setup for each slat, cutting all six vertical components using the same setup.

With the side joinery done, and your components looking like rejected Lincoln logs, you can turn your attention to cutting the slats to final size and mitering the ends. The only tricky part about this is making sure all the slats are the same length for the sides and ends. I shot all my miters by hand on a custom shooting board to achieve razor sharp, square ends. I cut all my components to the same length using a crosscut sled and stop block, then blackened one edge with a carpenter’s pencil so I could see it clearly at the shooting board. If you prefer, you can rough out the miters at the table saw, leaving only a bit to be sweetened with the plane. I hand planed the whole miter. With the pencil line as a visual guide, I was able to miter all my slats to within a micron or two of one another without having to start over!

With my side slats at final length, I was able to determine the final length of the handle. The handle and the slats both extend to the full length of the tote, so it is a simple matter to cut the handle exactly the same length as the side slats.

Next I cut the joinery connecting the handle to the handle supports. This is obviously a stress point, vulnerable to both pulling and racking. I opted for a double dovetail, mostly for the strength and a little bit to show off. I oriented the pins so that the handle would have to split out the supports to fail. I think this joint is overkill, but if you are nervous about your supports splitting you might add a cross-grain spline. Had I opted for a single dovetail, I may have added a spline.

The last bit of joinery is for a slat across the bottom. I was a little insecure about the length of the bottom slats. They had a lot of flex. I added a cross member made from the same material as the slats, jointed into the bottom side slats the same way the joints were cut in the vertical supports and corner stiles.

I finish sanded the inside faces of all my components before assembly, taking great care not to mess up any joinery.

Glue up occurs in stages, starting with the ends. Use whatever combination of clamps and cauls you need to get the six components comprising each end unit fully seated, flat and square. I like to clamp as much as I can directly to my bench. I was careful about shop rash, as pine is very soft and easily dented. It is easy to attach clamping blocks with painter’s tape to save the project from clamp marks.

With the end assemblies cured, you can add the handle and side slats. Start with the handle and get the dovetails pulled together, then begin adding slats. I worked my way up from the bottom, allowing each slat to cure before moving up to the next. This takes time but affords access for clamps in all necessary directions. It is a simple matter to pull the miters together because each joint has a flat section on the back against which you can clamp without fouling the miter.

There are a number of ways to handle the bottom. One would have been to cut a traditional groove in the bottom slats, but I felt like this material was too thin for that. Also, as a slatted bottom, there is little point in giving the two slats along the sides different support than the middle slats. I opted to support all my slats at each end and across the middle. I glued a 5/16” X 5/16” ledger along the bottom edge of the bottom slat at both ends. Simple and effective.

Installing the bottom is straight-forward except for notching in around the corner stiles and handle supports. Depending on how you laid out your ledger strips, you may want to glue up a bottom slat wide enough to span around the handle support, catching the ledger on both sides.

I used rasps and sandpaper to round the bottom of the handle for a comfortable grip, and rounded off the jointed ends which resulted in some extra pizazz on the dovetail pins.

I finish sanded the outside faces and used a burnishing tool to close up any gaps in my corner miters before giving the whole project a quick coat of wipe-on varnish. I kept the finish light to preserve the peach crate feel, while also adding some contrast to the dovetails.

This tote should provide years and years of use doing who-knows-what. It is a versatile project that I hope will provide the happy couple with lots of uses. Maybe they’ll even be sighted at the Whole Foods!

Gentleman’s Valet

                          Problem:                                                           Solution:

File this project under the heading, “Because You Can”.

I grew weary of my personal grooming tools taking up a whole shelf in my medicine cabinet in a state of
disarray. Disarray is not my friend, and I am deeply grateful for the myriad ways in which I can bring order to chaos with my woodworking skills.

As for as I know, there is no name for a piece like this. Jewelry box is good, but for the fact it was never intended for jewelry. I’m calling it a Gentleman’s Valet. I fancy myself a gentleman and Webster says a valet is “a device for holding… personal effects”. With Webster on my side, I feel unassailable.

Here are the vitals: It’s made of walnut and pine and is about 7” tall, 9” wide and 5” deep. The construction and joinery are traditional and the finish is French Polish.

Design

There are a couple design details worth mentioning. Wood first- I had some walnut scrap in the bin with a little figure on one end. I wasn’t able to mill it to have consistent figure across my drawer fronts, but I was able to create an asymmetrical effect by carefully choosing and orienting the available figure to create the illusion that the drawers are from one board. Don’t look too close.

I stole an idea from Mike Pekovich’s “Small Cabinet Is Big on Details”  (Fine Woodworking #254), where he set the door and drawer of a shaker-inspired cabinet back from the stiles, rails and divider. It’s a subtle effect that creates great light lines without shouting about it. It also breaks the reveal so that inconsistencies between the moving bits and the case bits disappear. Brilliant. See the effect on my little valet. The case is about 1/16” proud of the drawer fronts when the drawers are closed.

Another subtle design feature is that the top two drawers are the same size and the bottom drawer is a bit deeper (or taller, or bigger… not sure how to say this… see photos). The larger bottom drawer gives the project a sense of being grounded and stable. Even on something this small.

I opted for finger holes instead of hardware pulls simply because there isn’t room between the valet and the cabinet door for a pull when the cabinet is closed. I was loath to drill out so much figured material, but alas, form follows function.

Materials

As I mentioned, this is made mostly from walnut scrap from the crap bin, which is one rung above the kindling box. Translation- I built this for free. The walnut case sides, top and bottom are about 5/16” thick, as are the drawer fronts. The dividers are about 1/8” thick (to be exact, they are equal in thickness to the width of a thin-kerf table saw blade, which made the dados into with the dividers are fit). The case back is also made from 1/8” thick walnut strips to keep milling to a minimum. Each horizontal back strip is roughly the width of the drawer behind which it is nailed.

The drawer sides and backs are ¼” clear pine. I love the pop in contrast between the nearly-white pine and the chocolate end grain of the walnut at the dovetailed drawer corners. The drawer bottoms are ¼” MDF, chosen because there was a small chunk in the bin and because I knew from the get-go I would be lining the drawer bottoms with felt. No need to put anything pretty down there, and no need to worry about wood movement.

Construction

Guilty. Of total overkill. Surely you are thinking that dovetailing the case of a box like this is ridiculous; the behavior of a hopeless show-off at best and a person of questionable character at worst. But it looks cool as hell, right? Truth be told, until I decided to display this project to the 14 people who might read this article, I didn’t think anyone else on earth would ever see it. It’s in my bathroom medicine cabinet, for Christ’s sake! The main reason I dovetailed the case was for practice. That’s a recurring theme in my shop. Practice the tricky bits on projects where mistakes matter little. While we’re on the subject, half-blind dovetails in ¼” material on a drawer that will never hold more than about 6 oz? Cuckoo’s nest. I know.

Actually, on the subject of dovetailing these drawers, I have wisdom to share. Chopping pin sockets in highly figured walnut is hard. Now I know, and so do you. The wood is unpredictable and brittle. True story- of the 6 dovetailed drawer corners on this project, I cut 5 of them without incident. It was on the 6^th and final joint that the wood started popping apart with every chisel tap. On the bright side, highly figured wood is great for hiding repairs. Multiple repairs. One night while glue was curing, it occurred to me to drill out all the socket material I could on the drill press. I was able to pare what remained instead of chopping. The wood (and glue) held together in the absence of mallet blows. Whew. Always on the last joint of a piece of wood you can’t replace, right?

The last note on construction is in regards to the case back. Fitting a miniature chest of drawers inside a cabinet is a game of millimeters. Setting a back panel into a groove cut ¼” from the case back would have had a dramatic impact on the depth of the drawers. To minimize loss of drawer depth, I milled a rabbet around the perimeter of the case and set in the thinnest possible strips of wood to constitute a case back. Also, because the back presents a cross-grain problem, I brad-nailed the back strips into the rabbet with the expectation that the brads will allow the wood to move without buckling the back panels. I used brad nails in predrilled holes with the heads nipped off for pins, which were eventually sanded down flush to the case back.

Finish

I have envied French Polish forever. This project was virtually screaming for it. I’ve tried it before… you can guess how it went. I took a deep breath and decided I couldn’t get too badly hurt. A project as small and simple as this can be sanded down to new wood in half an hour. It seemed like a perfect learning opportunity. Lots of Youtube videos were watched. The exact materials recommended by the experts were gathered. A test board was prepared. And the rubbing began.

I’m pleased, especially with the drawer fronts (where it matters most). I learned tons about French Polishing. I got some “roping” on the sides, but I challenge you to find them from where you’re sitting. All in all, it’s not too tough to do. It takes some skill and experience. I’m not saying I’m ready to French Polish a priceless museum piece, but I’m eager to try it again on another project where the stakes are fairly low.

I’m not going to try to teach the technique here, but I’ll tell you what I think I did that resulted in a fairly successful finish. Watch several videos. Different finishers have their own approaches and it is valuable to understand the underpinning constants and not get too hung up on one person’s technique. Use the right materials- linen wrapped around cotton or wool for the rubber. Shellac in a 2 lb cut (which is easy because Zinsser’s ubiquitous SealCoat is a 2 lb cut of dewaxed shellac). Denatured alcohol. Non-curing oil (I had flax oil on hand with no earthly need for it, so why not?). Be patient, keep moving, and build thin coats. Thin is good. My biggest challenge was charging the rubber without making it too wet. My second biggest challenge was never stopping with the rubber on the project. You can’t change direction without lifting- EVER. This gets trickier as the surface being polished gets smaller and as surface irregularities are introduced. I’m trying to figure out how one would do French Polish on something like a thin table leg with a carved surface. Shudder.

In Conclusion…

Nobody needs a doll-house-sized chest of drawers for their nail clippers. Projects like this are born in minds that cannot help looking at their surroundings through a lens of possible improvements. My little valet has been in use for a couple months now and I love it. I love taking a drawer out as a tool tote to perform a little grooming in front of the TV. I love not looking at a chaotic pile of stainless steel this-n-that in my medicine cabinet. I still pull the drawers out from time to time to look at the joinery. Need has nothing to do with this. I made this because I wanted it and because I could. If you have a better reason for any project, I’m all ears!

Serving Trays

(Or any box-shaped project with mitered corners and miter keys)

As projects go, this one is so simple to build that detailed instructions may be overkill. There are techniques here, though, that are fundamental to any box-shaped project with mitered corners. Projects like this build knowledge and skill, and there is wonderfulness in the ability to make custom accessories that fit your personal needs. I have found that the simplest projects can have a profound impact on my woodworking and on daily conveniences and efficiencies, and these serving trays fall into both categories.

The impetus: We have three inexpensive trays that we use for dinners in front of the TV and for coffee/tea/snacks around the house. These are the type of things you get at Target or Pier One. Two are too large; more like the “breakfast in bed” affair you see in old movies with fold out legs. One is a good size, but as a single, it is in too much demand (translation: it is never available for my use). All are built cheaply and are out-of-context among our other furnishings and accessories. I decided it was time to take advantage of my ability to make a set of trays that suits us and how we use them.

My design decisions started with where these things were going to be stored. The trays we have are a pain to store. They’re always in the way and don’t stack nicely. I have a lower shelf on a serving bar with space just right for a set of nesting trays, so I designed my trays to fit on that shelf. The largest is about 15 ¼ X 15 ¼. That also happens to be a great size for a lap tray that will hold a dinner plate, beverage, tableware and maybe a side dish.

The next decision was based on use. There are two of us, so we need two lap trays we can use for dinner. But we also use the smaller of the trays we already own for drinks and snacks, and we seem to compete for it. It’s also a little large and cumbersome for a drink and a snack, so I thought maybe we would get more use out of a couple half-sized trays. I settled on a set of four trays. The largest is 15 ¼ X 15 ¼, the medium tray fits inside the large tray and is 14 ¼ X 14 ¼. The two half-sized trays fit inside the medium tray, with exterior dimensions of 13 ¼ X 6 each. They nest, but loosely so that they can be pulled from the group at odd angles or with one hand. This is not a situation that warrants a piston fit.

I made my tray sides from 5/16” oak, because I had oak scraps on hand from another project. I made the tray bottoms from 3/32” Luaun, with cork shelf-liner attached with spray adhesive. The combined thickness of the Luaun and cork is exactly ¼”, which is convenient for milling. I avoided solid wood bottoms because they must be left loose in the frame for expansion. Ply material like Luaun doesn’t move, so it allows the bottom to be glued to the sides. This provides additional strength and eliminates rattles which is one thing I hate about my store-bought trays. Rattles sound cheap.

My thoughts on using cork- First off, I already owned it. I originally thought I would laminate the tray bottoms with some kind of grippy rubber sheeting, but then remembered I over-bought cork for a different project. Second, I checked it for its ability to prevent glasses and plates from sliding around and it is excellent. I also know cork is extremely durable and can be washed. The biggest advantage, though, is that it can take stain like any other wood product, and polyurethane, too. If you don’t overdo it, you can poly cork without making it slick. Cork’s a winner.

I started by laminating my cork shelf-liner onto a Luaun substrate and setting it aside. Then I milled my tray sides in the usual way. My sides are 2 ½ inches tall, which leaves plenty of room to cut in handles without weakening the sides. It is also helpful to leave enough material between the hand holes and the tray bottom so small things can’t fall out the sides of your trays.

I milled my tray sides from boards long enough so that all four sides could be cut from a single board. This allows the grain to be “wrapped” around the corners (three out of four, anyway), which is a nice touch on an item like this that will be handled up close. This is simple enough to do; just mark your boards as you mill them so you can match up your corners when it comes time to assemble the trays.

Next, I cut a shallow groove in all my sides that will fit my tray bottom laminate precisely. Then I cut the tray sides to their final length.

I opted for keyed miters on the corners. I tend to dovetail corners like these, but I think keyed miters have a more elegant appearance. Cutting miters can be tricky, so some care is warranted here. If you intend your joints to be glue-ready off the table saw, you’ll need to make sure your saw is tuned up, that your miter gauge is set square, and that your blade is tilted accurately. On tune-up, keep in mind that tilting a table saw blade introduces a third dimension (or plane) that must be in perfect adjustment for miters to come out accurately. Very little is said about shimming table saw tables front-to-back for this additional dimension, and if you have no idea what I’m talking about, do yourself a favor and find out. Roland Johnson describes how in Fine Woodworking, Issue 179, Page 46. http://www.finewoodworking.com/how-to/article/how-to-tune-up-your-tablesaw.aspx

Alternatively, you can rough-cut your miters by hand or machine and tune them up on a shooting board for miters. Regardless how you proceed, make sure every opposite side of each tray is the same length. Variations in length will mess up your miters just as badly as being out of square or off 45 degrees. Tight miters require all angles to come together harmoniously.

Next I cut the handles into the sides. I put handles on all four sides so that we never have to look for a handle. I cut my handles quite small, because a good grip equals fewer accidents and fewer carpet stains. After some trial and error, I settled on handle holes that were ¾” by 3”, a secure fit for my Sasquatch-sized hands, without being too sloppy for my wife’s hands-of-a-twelve-year-old. Using a Forstner bit, I drilled ¾” holes at the drill press, centered on each tray side from end to end and from top to top-of groove, 2 ¼” apart. I sawed out most of the waste between the holes with a dovetail saw, then routed out the remaining material at the router table using a straight fence. I finished the operation at the router table using an 1/8” radius bit with a guide bearing to ease all the edges of the hand holes for a comfortable grip.

With all tray sides milled square, grooved for the tray bottom, cut to length, pierced for holding and sporting perfect miters on every end, it’s time to do some pre-finishing. Avoid trying to finish sand and stain the inside of these trays after glue-up. The inside face of each side, plus both sides of the tray bottoms need to be completely finished (save perhaps for a coat or two of wipe-on poly) before glue-up. Be careful sanding. Use a block to keep material removal level and consistent, and don’t over-do it! Sanding too aggressively (and unnecessarily) can throw your corners out of alignment on final assembly.

It’s time to assemble the trays. Referring to your marks for grain wrap, line up the four sides of a tray with the finished, insides facing down. Using a straight edge for registration, push two tray sides together and observe your beautiful gap-free joint. Stretch blue painters tape across the joint and repeat for 3 of your 4 corners. Flip the assembly, finished side up.

Using an acid brush, I applied glue to the tray groove and all the faces of the miter joints. Next, I inserted a tray bottom into one of the sides in the middle, then wrapped it up with the other three sides. The blue tape “clamps” should bring your miters tightly together on 3 sides, and the final side can be adjusted with a few light taps and taped shut. If everything was cut square and true, there should be no need to torture the tray into alignment, no ghastly clamp-a-palooza, and no gaps in your joints. If you Dagwood this glue-up, relax. There are probably tricks remaining to sweeten up your joints, and if they can’t be made to look perfect, at least they represent your own work and not the work of a child laborer in Bangladesh. Remember the point of this project is to have something functional and customized just for you.

One common problem with miters is a gap along the length of the joint, either because your miters are not exactly 45 degrees, or the 45 isn’t square to the stock, or the stock isn’t straight, or some other issue. Take heart. You can probably burnish your corners to close the gap. Those wood fibers at the sharp end of your miter cut are very thin and very flexible (and fragile). Small gaps along the joint’s length can be closed by bending those thin fibers toward one another with a burnishing tool or the smooth part of a 3/8” drill bit. Gently and patiently close the corner by drawing your burnisher out and across the joint. Try to move each edge equally toward the other. When your joint is nicely closed, secure and harden it with CY glue (a/k/a/ CYA glue).

The next step is to spline or “key” the corners. End-grain to end-grain joints are not strong enough for the daily use these trays will see. The mitered corners must be reinforced, and corner splines or keys will do the job with elegance.

If you don’t already have a jig to cut corner keys, it is time to make one. My jig fits the table saw. I’ve seen some guys do this with a slot cutter on a router table. My design may take a bit more time to make, but I can add sacrificial surfaces to the inside faces and use this for a lifetime, whereas a jig for a slot cutter, with its limited depth of cut, will chew up a jig over time and the whole jig will need replacement. Another option is to create some sort of attachment to a biscuit joiner. Choose your method.

I started by deciding how many keys would be needed to have strong, beautiful corners. I decided on 5, favoring an odd number so that one key would be dead center on the corner joint. Next, I worked out a distance between the keys that left me with enough material at the top and bottom of the corner to be strong and look good. I didn’t bother trying to get the same space between the keys and the top or bottom of the tray as between each key, but you can fuss around with the math until all keys are equally spaced if you wish. I ended up with 3/8” between each key, which left about 1/8” at the top and bottom.

Using a 1/8” dado cutter on my table saw adjusted to a depth well shy of cutting through my mitered corners and into the inside of the tray, I centered my first cut on each corner of each tray. I clamped a stop block to my jig to ensure consistency. After all my centered key grooves were cut, I marked the location of the stop block on the jig and moved the stop block ½” away from the blade, representing the 3/8” space between the keys plus the 1/8” thickness of the cutter. Then I cut all four corners of each tray twice, once with the tray bottom to the left of the stop block and once with the tray bottom to the right of the stop block. This added two grooves to every corner, each 3/8” from the center groove. Again, I marked the location of the stop block on the jig and moved it another ½” away from the cutter and repeated the operation, finishing the process of cutting 5 key slots on each corner.

Next, I milled the key material. I used some walnut scrap for contrast and started at the band saw, slicing off strips roughly ¾” wide and about 3/16” thick, jointing the material before each cut so that one side of the key material was always straight and smooth. Make lots. You don’t want to have to abandon a glue-up to make more keys. I used my planer to sneak up on a fit that was too tight by a hair. My machines are too violent to produce consistent thicknesses in material this light and thin. I left enough material so that every key would require a few strokes on sandpaper to achieve that lovely piston fit.

I used my crosscut sled to cut dozens of keys about an inch long, using a pencil with an eraser to hold the small pieces and clear them from the table saw blade. This operation takes care and attention to perform safely. If you aren’t comfortable doing this at the table saw, cut them by hand with a miter box or at the band saw.

I cleared space on my bench to glue keys (I had 80 to do all together), which included a piece of granite with adhesive sandpaper on it to “joint” the sides of my keys to a perfect fit. One by one, I sanded the sides of my keys to fit each groove, and I slide them into place with a drop of liquid hide glue- just enough to squeeze a bit out of each end of the key. I used hide glue for its long open time, which really helps in a future step, and I made sure every key was centered on the joint and proud of the tray side at each end.

Keys were curing as I was still gluing, and as a box would cure to a fairly firm tackiness, I would break from gluing to saw off the exposed keys. A very timely “Methods of Work” article was printed in Fine Woodworking #255, in which John Bates from Tampa described a method of trimming miter keys that worked great on this project. Basically, a guide board is clamped to the band saw, and the saw blade is recessed into the guide board. The blade is recessed deep enough into the guide board so that the project can be passed through the saw against the guide, trimming the keys without marring the side of the work. The guide board must be thin enough for the lowest key to pass over it. For these trays that required a guide board 1/8” thick. It worked great! I was able to quickly and accurately trim all 80 of my miter keys to within about 1/32”, leaving just a few strokes with a sanding block to level them.

Here’s that future step I referred to where using hide glue helps. Trim your miter keys while the glue is set but still gummy. Then use some cheap, coarse (80 or 100 grit) sandpaper to level your miter keys. You’ll have to change paper often due to glue build-up, but your sanding dust will be forced into any gaps and will be fused with the gummy hide glue. You’ll have perfect, gap free joints that match the surrounding wood. This works with PVA glue too, but PVA isn’t as forgiving as hide glue.

Here’s a trick to fix miters that didn’t come together at the top or bottom. Force some hide glue into the gap and let it get tacky, just as with the miter keys, then sand the corners to force dust into the glue. Unless your gap is as wide as the space between students at a middle-school dance, this should close the gaps.

After the trays cured fully, I finish-sanded the top, bottom and sides and applied the same finish I used for the inside. For projects like this one, I really like General Finishes gel stain, which is really a tinted gel polyurethane. I know these will be wiped down frequently with a damp cloth and will be exposed to other fluids from time to time, and so poly is the only finish I would consider. I like the satin finish of the stain by itself, but it is simple enough to add a few coats of wipe-on poly if shiny is your thing.

Viola. A wonderful set of functional, easy-to-store serving trays with elegant, keyed miter corners and grippy cork bottoms. These will provide us with a lifetime of use, and I can use everything I learned about making keyed miters for almost that long, too!

One day I will share all my theories for great kitchen design, but for brevity, here are a few distilled thoughts. Cooking is like woodworking. There is a routine to the production of work, whether it is omelets or ogees. A kitchen, like a woodshop, should be arranged, equipped and organized to perform the tasks essential to produce the desired work.

This article is about organization- specifically the efficient and effective organization of kitchen drawers by making custom drawer dividers. My kitchen drawers are well-organized now, including that one drawer that plays the role of the schizophrenic dump site for all odds and ends including a tape measure, box cutter, screwdriver, keys and more. It may seem silly, but getting my drawers sorted out has had a big impact on how much I enjoy working in the kitchen.

My drawer dividers are special. The plexiglass drawer bottom affords the extra benefit of protecting my painted wood drawer bottoms from scratches and moisture (like when my wife gets a little eager to put things away before they are completely dry). But I’ve gotten ahead of myself…

Step One is to plan carefully. Determine what goes in which drawer. Don’t under-think this as so many do. Be logical. Spatulas and other cooking tools go near the stove. Silverware and serving utensils go near the dishes, which ought to be near the dishwasher. Cutlery, if stored in a drawer, is near where food is prepped for cooking.

You may find, as I did, that you have more crap than you need to do the job of preparing food. You may own 4 ladles, like me, but you shouldn’t let what you own determine how your precious drawer space is used. Ladles, for example, are space vampires. I kept my two best ladles in the kitchen and put the others in a box, with dozens of other duplicate and misfit tools and accessories that were rarely, if ever, used. What’s left has been thoughtfully vetted, resulting in a food preparation tool kit that has everything I need and nothing I don’t. It’s like a big, spread out Anarchist’s Toolchest.

Once the inventory of each drawer is determined, you can experiment with divider designs. Here’s where the fun begins. You can do ANYTHING! You are not constrained by the limitations of commercially available divider systems. You can make an individual compartment for every device in a drawer if you want, although that’s not a great idea.

Design in some flexibility. You will acquire new things and replace old ones over time and you’ll be unhappy if you discover that your awesome dividers preclude the easy storage of some new or replacement tool. Drawers that contain a wildly diverse inventory are best divided into large, undefined spaces. Sort contents based on criteria like “more frequently used” and “less frequently used” and situate them front to back in the drawer accordingly. The best solution is to simply keep the less-used items from fraternizing with the oft-used items.

Silverware is another matter. In a silverware drawer, there should be a space for each type of frequently-used utensil. Depending on your drawer, you may combine silver with serving ware and other devices closely related thereto. I ended up with an empty compartment in my silverware drawer, which suits me just fine.

My design process involved cutting a piece of poster board or heavy paper the size of each drawer and arranging the drawer contents on it until I was happy. I then noted on the cardboard the location of every divider component, using a scrap of wood the width I intended to make my dividers (5/16”). Keep joinery in mind. You can’t have two dados directly across from one another in stock this thin. You can use bridal joints as an alternative if you absolutely must have two dividers cross one another perpendicularly.

Eventually I ended up with a template for each drawer. I was able to build my dividers with confidence that when they were fitted into the drawer, all the contents would fit too. The more accurate you make the poster board template, including accurately fitting into the drawer, the less drama you will have when it’s time to put dividers into drawers. The drawer may not be perfectly square, so mark your template’s orientation to the drawer!

It’s finally time to do some woodworking! Mill lots of straight stock the width and thickness you want for your dividers. My stock is 5/16” thick, which is plenty strong when secured to the plexiglass base, and I varied the width (which ends up being the height) based on what the divider was dividing. I made my dividers 2” tall in drawers with large compartments and 1 ½” tall in the silverware drawer because the silverware compartments are small and would be hard to get down into if the dividers were taller. I finish-sanded all my stock before cutting joinery so that I could be assured tight joints, as they are highly visible.

Set up a dado stack and test it on scrap stock to sneak up on snug dados about 1/3 the depth of the thickness of your stock. Use stop blocks to keep dados aligned where your design calls for it. Note that you needn’t bother to cut the dividers to their final length yet, which may make cutting the joinery easier. I left my dividers unfinished, but if you intend to apply finish you should do that now, before you do any assembly. Dry fit your dividers together per your design, which may involve lots of little sub-assemblies that won’t come together until they are screwed to the plexiglass base. Set them aside.

Use your poster board templates to cut 1/8” thick plexiglass to fit your drawer bottoms as accurately as you can. I do not recommend testing the fit in the drawer before you have the dividers in place because you can lose the plastic in the bottom of the drawer with no easy way to pull it back out. If you cannot resist test fitting the plexiglass, string some dental floss across the drawer before letting the plastic drop into place. Hopefully you will be able to use the floss to lift out the panel.

Carefully set out your dry-fit drawer dividers on the poster board template, then situate the plexiglass on top of the dividers. If your dividers have a top and bottom, they should be top down. Mark the ends of your dividers at the edge of the plexiglass with a knife tick, then disassemble your divider assemblies and cut the divider components to final length. Reassemble the components and glue them together this time.

Reposition your glued divider assemblies on your templates and position the plexiglass drawer bottom over the dividers. To attach the plastic bottoms to the divider assemblies, drill and countersink holes for #4 X ½” wood screws centered on the drawer dividers. Make sure to secure the unsupported ends of any divider assembly and to secure any jointed intersections to the plastic. I didn’t use any adhesive, but if you want to add a super thin smear of silicon between the plastic and wood for good measure, feel free!

The penultimate step is to check the drawer divider assembly for fit. If things ended up a little tight, a few strokes along on edge with a sanding block should to the trick.

Lastly, fill up your newly organized drawer with kitchen tools and enjoy years of efficient, convenient access to all the knicks, knacks and paddy whacks that make up your kitchen tool kit.

Postscript- I bought a few extra sheets of plexiglass to line drawers that are subject to special punishment. I opted for large drawers instead of cabinets in my kitchen, so pots and pans live in drawers. My drawer bottoms are painted, and the thought of cast iron skillets and stock pots marring my beautiful paint work was too much to bear. I dropped a snug-fitting layer of plexiglass into the bottom of these punishment-prone drawers. I have to look closely to see the plastic down there, and it takes a tremendous amount of abuse while keeping the insides of my drawers looking like new!

Typical cabinet construction involves building plywood cabinet boxes and attaching face frames to the front. Methods for attaching the face frames to the boxes vary. None that I found met all my needs, so I developed my own way of building and attaching face frames to cabinet boxes. My way of making and attaching face frames ensures a perfect fit, installs without nails or screws through the show face of the frames, and get this– they’re removable!

Years ago I read an article by Norm Abrams where he described how he uses biscuits to attach face frames to cabinet boxes, virtually eliminating the need for hardware and allowing face frames to be finished before installation. I really liked Norm’s approach, primarily because it keeps the finishing process out of the house. He rips a slot in the front edges of his cabinet boxes, then uses his biscuit joiner to cut slots in the backs of his face frame components. He glues the frames to the boxes with biscuits so there are no nail holes to fill and finish. It was easy to rip a groove in the front edges of all the box components, but when it came to cutting biscuit slots in the back of the face frames I was vexed. On my cabinets, the face frame edges are not flush with the edge of the plywood boxes so I have no reference edge from which to set the biscuit joiner. I was unable to divine a method of setting up the slot cutter to mill the stiles that was accurate and reliable, and Norm’s article is surprisingly silent on this crucial point. I was inspired but not converted.

I knew I could improve on Norm’s technique. My solution does everything Norm’s approach does, but is easier to do and has the added bonus of making the frames removable. My solution is to use 1/4″ MDF as a spacer between the cabinet boxes (creating a “female” void between every cabinet box) and as a spline running down the back of the frame stiles (creating the “male” element of the joint that fits into the void instead of a biscuit). The result is that all my cabinet boxes are 1/4″ apart, which makes the total width where two cabinets meet = 1 3/4″. My 2″ stiles cover beautifully with 1/8″ overlap on each side. I ripped a 1/4″ groove down the center of the stiles. The MDF spline was glued into the stile’s groove when it was time to install the face frames, and the MDF spline slid nicely between the boxes and were easily secured with a couple screws located discreetly inside the cabinet box. I used screws instead of glue for easy removal and repairs if needed.

I had a couple long frame rails in the kitchen and was nervous about leaving them unattached. One option was pocket screws. Another was an attachment block glued to the cabinet box through which screws could be driven to secure the frame rails. For cabinets up to about 18″ wide, I didn’t see any need to attach the frame rails to the boxes. My frames are all 1″ X 2″, so they’re pretty stiff and strong.

For end panels like the one shown in the featured photo, it too is attached to the cabinet box with 1/4″ MDF spacers to create the void for the spline on the face frame. Careful fitting of molding hides the reality that the face frame is not attached to the end panel save for the spline joint, although these corners can be glued if you have no interest in making the removable.

For cabinets that terminate against a wall, a scrap of plywood attaches to the cabinet box separated by 1/4″ MDF, as shown in the drawing.

This method checked multiple boxes for me. First, it was a foolproof way to align face frames with cabinet boxes. Rails were custom cut with the stiles dry-fit in their grooves- no measuring ensured a perfect fit despite variations in plywood thickness and flatness (Grrr). Second, the face frames were made to be removable… forever. I’m jazzed about the ability to pull all my frames out when the hardwood floors need to be refinished, saving my beautiful legs from the floor sanders.

Of all the benefits of this technique, my favorite has to do with finishing. I love being able to finish all my frames, doors and drawers in the shop and then install them without ANY nail holes or any other breach of my finish coat. I like to spray opaque lacquer, so it was especially nice to keep that stuff out of the house.

I have need from time to time for a long miter like those on the vertical piece of the crown molding below, or on a case side. These joints can be vexing to make on a table saw or miter saw because of blade flex, miter slot slop, etc. Also, miter saws have limited capacity to cut wide stock held vertically, and I can’t ever seem to get an accurate 45 degree cut with the saw head leaned over. I wanted to rough out long miters on the table saw and shoot them to perfection with a plane just as I do with square ends and flat miters, so I spent a few sleepless nights working out the details of a shooting board for long miters.

Turns out the nice folks at Fine Woodworking find cutting these joints to be challenging, too, and they liked my solution. They published an article on this project.

A shooting board for case miters

This fixture is my take on the venerable donkey’s ear for shooting long miters. Why not just make a donkey’s ear? Because they are hard to make, hard to store, and hard to adjust. This design solves all those challenges.

This is a typical shooting board made of 3 pieces of baltic birch plywood and two pieces of hardwood. Plywood makes up the base, the stock bearing surface, and a back stop for the plane skate. Hardwood is used for the plane skate and the fence. My plane skate is ash for long wear. My fence is poplar because that’s what was in my scrap bin.

The only hard part about this project is getting a perfect miter along one edge of the plane skate. I roughed out the 45 degree rip cut on my table saw using an oversized piece of ash. My 45 degree rip cut wasn’t close enough to the tolerances required for a shooting board so I spent a long time tuning it with hand planes until it was as close to 45 degrees to the base along its entire length as I am capable of making it. Working from an oversized piece of stock affords you the option of running it through the saw and starting over if you get too far out of whack. After you tune up the 45 degree rip cut, you can rip the skate free of the oversized piece.

All my plywood bits were cut dead square so I could use their sides and corners to register them to one another and to the skate. I glued the skate and backer strip to the base plate. I screwed the stock bearing surface to the base so that I could make adjustments if needed. If you aren’t able to get your 45 just right, you can shim the stock bearing surface to cover your tracks. The relationship between the stock bearing surface and the skate is crucial. It must be 45 degrees.

With everything assembled but the fence, run your favorite shooting plane on the track until the stock bearing surface is chamfered and the plane is riding on the edge of the sole where there is no iron protruding.

I milled my fence straight and square and roughed 45 degrees off one end. At the drill press, I drilled two 1/4 holes through the fence (aligned with the back edge of the board to get near square) and 3/4″ into the shooting board. I removed the fence and enlarged the holes in the shooting board to 3/8″ to accept threaded inserts. I attached the fence with 1/4-20 bolts and fender washers, put a backer on the fence (also cut to rough 45) to prevent blowout, and shot the fence end to 45 degrees. At this point the end of the fence and the skate should have a perfect 90 degree relationship.

The final crucial adjustment is squaring the fence to the shooting track. Rough a 45 degree cross cut in a piece of soft stock at least 6 inches wide with parallel sides, making two test pieces. Shoot the ends of both pieces until all your saw marks are gone and the plane is shaving along the whole length of the cut. Put the pieces together to form a 90 degree corner and check for square. If it’s perfect, congratulations! If your corner is out of square, you can shim the stock bearing surface until the miter comes together at 90 degrees.

Next, put your test pieces flat on the board with the miter cuts nested, tight against the fence. If there is no gap at either end regardless which edge is against the fence, congratulations! I was not so lucky, so I drew a pencil line on the stock bearing surface to gauge adjustments, loosened the bolts and nudged the fence as needed to get a perfect 90 degree shot along the full edge of the joint. There was just enough flex in my bolts to square up my fence, but you can open one of the 1/4″ bolt holes through the fence a little to create more adjustability. Once my fence was perfectly square to the track, I screwed it down for all eternity, I hope.

If you get the skate right at 45 degrees to the stock bearing plate and get the fence set at 90 degrees to the track, you can make long, tight miters on moldings, case sides, boxes, or anything else that presents an opportunity to use this beautiful, classic, and challenging joint.

This fancy little bowsaw is an elegant replacement for my hardware-store coping saw. It may be more accurately called a turning saw. As I understand it, turning saws are slightly larger than coping saws. This little saw takes standard coping saw blades, so I’m sticking with “coping saw”. Whatever you call it, make one. It is an awesome little saw and I never tire of using it!

Several years ago I took a class at the Woodwright’s School in Pittsboro, North Carolina and learned how to make a bow saw. The class was delivered by Bill Anderson, a recognized authority on bow saws among other things. I came home with a lovely bow saw, with a blade roughly 18” in length that functions nicely for what it is. Someday I will put a finer blade on it because the blade we received in class is too rough for my taste. I enjoyed making the saw and I like having a fine, handmade bow saw in my shop, but if I’m honest, I never use it. The blade is too rough and I have a bandsaw that cuts better and is easier to use.

Nevertheless, taking the class turned me into a competent maker of bow saws, and so it should not come as a surprise that when I was cussing my coping saw for being too flexible, it occurred to me to make a coping-sized bow saw.  It’s the perfect type of project for when one finds oneself blessed with a weekend without anything in particular to do.

Even though I had the skills to make a saw, and the hardware (harvested from a humble Robert Larson coping saw), it still occurred to me that buying a fancy coping or fret saw might be the smarter option. If you haven’t lusted after those spacy looking Knew Concepts red anodized aluminum hotrod saws that cost $100, you must not have the same tool DNA as me. But I decided to invest time rather than money, and the die was set. The challenge would be to have a saw that worked as well as the finest saw money can buy, for free.

Making a bow saw of any size or shape is pretty much the same. Four pieces of wood, a blade, some twine and a few pieces of hardware is all that is involved. Here is my entire materials list:

Wood scrap, roughly 1X2X20, of a dense, hard wood like oak, ash, hickory etc.

Blade holding pins harvested from Robert Larson saw (now discarded)

2 bolts, 3” long, with smooth shanks for at least 1 1/2” (I used ¼-20 threaded bolts)

Coping saw blade (I like Craftsman fine blades- they’re cheap and cut like a furious beaver)

About 2’ of jute twine

The process is basically this. You make two main bits that form the vertical ends. Then you make a cross member that acts as a fulcrum at both ends. Keep a little scrap about the size of a tongue depressor to make a windlass. Get everything rough cut and squared up. Cut the joinery. Make the hardware. Cut the cross member to length and make the tenons. Shape the parts. Assemble. Apply finish if desired. Use.

First, determine the size of the saw and rough out the ends and cross member. My finished ends and my cross member are both about 8” long, but you’ll want to start with longer pieces. If I make this saw again, I’ll make the end pieces an inch longer so I can get all 4 fingers between the blade and cross member.

Joinery on the end pieces is next and should  precede any shaping. You’ll want to cut joinery on square stock. There are two important joinery tasks. The most intensive is the mortises where the tenons on the cross piece fit into the ends. These joints must act as a hinge so winding the windlass tensions the blade. Great care is warranted here so that your end pieces are held in parallel when the saw is in use. The tenons need to move up and down in their mortises, but not side to side. Furthermore, the mortises need to be centered and vertically oriented on the end pieces. This is a good place to spend some extra time to get good, tight hinge action.

Placement of the mortises relative to the top and bottom of the saw is important. I position the horizontal piece on the end pieces by eye and instinct. I want the horizontal piece to be somewhat closer to the winding cord than the blade to moderate tension on the blade. The end pieces are levers, hinged on the cross member. Where you place the horizontal member relative to the blade and the winding cord determines the leverage available to tension the blade when winding the cord. A saw with long ends and a cross member positioned close to the blade will be difficult to tension without breaking the blade due to excessive leverage. If there is a science to this, I don’t know it. Based on the antique bow saws I examined and on my own experience, I set my horizontal cross member closer than midway to the cord by a distance sufficient to moderate the leverage of the windlass. For example, if I were to make a saw with 12” from the blade to the cord, I’d cut my mortises about 7” from the blade (5” from the cord). I eyeball that ratio on saws of different sizes. The center of my mortises is 3 ½” from the blade and 2 ¾” from the cord. Another consideration when determining the distance from the blade to the cross member is clearance of stock when cutting. The closer the cross member to the blade, the less cutting capacity you have.

My mortises are 3/16” wide, 1/2” long and 1/2” deep, created with a 3/16” brad point bit on a drill press. Multiple stabs were needed to clean out the mortises until the bit would slide from end to end, and a fence kept all the holes in a row. No chiseling necessary. The finished dimension of  my cross member is 5/8” by 3/8”, so a tenon cut to 3/16” by 1/2” leaves four nice shoulders

The next bit of joinery is the holes where the blade hardware attaches. I deviated from tradition here, eliminating the tapered pins and handles. For a saw this small, I think it’s easier to hold it by the frame. I kept this in mind when determining the distance from the blade hardware to the fulcrum mortises, making sure there was room for my hand. I also opted for a non-traditional attachment method. I simply drilled and tapped the wood to accept a ¼-20 bolt. I cut threads in the wood for about an inch, and they are more than amply strong to take the stresses they will endure. The blade will snap before the threads strip. Drill a 7/32” hole (preferable on a drill press for accuracy) as deep as you want your threads to go, then tap the hole with a ¼-20 tap. If you’ve never tapped wood before, you will be delighted to learn how easy it is and how strong the threads are. Go slow and reverse often to clear your tap’s flutes. If you’re like me, you’ll find many opportunities to use this technique on other projects.

There is one more bit of joinery that isn’t really joinery at all, and which is undertaken later at the shaping stage. The top of your end pieces need some kind of groove to hold the jute twine. A simple notch is more than sufficient. I like to think of my handmade tools as heirlooms, so I go a little overboard. I also find it inspiring to use excellent tools, so whatever time and effort I put into making my tools special comes back to me in pleasure of use. The end pieces of a bow saw are a wonderful opportunity to off-gas some creative energy, and the tops are veritable magnets for special treatment. Knock yourself out. Or not. A shallow nock will do.

Make your hardware next. There are two bits of hardware I didn’t feel comfortable making, and they are the two slotted pins that hold the blade. The slots are too narrow for anything I have that will cut metal, and they need to be flared or tapered to hold the pins on the blade. I took a short cut and harvested these two small pins from the coping saw I replaced. They were held in place by friction and it was a simple matter to tap them free.

The goal is to drill holes in the shanks of two ¼-20 bolts for the blade-holding pins, then cut off the heads and screw the bolts into the threaded holes you cut in the saw ends. Here’s how I did it. First, don’t cut off the bolt heads until you’ve drilled the holes. Second, mic out the pins to determine the size hole to drill- you want it to fit tight. Mine were 1/8”, so I drilled 1/8” holes. To keep the drill bit from wandering, I filed a small flat spot on the bolt. Then I took a small block of wood, maybe 3” square and ¾” thick, and cut a groove across it just shy of ¼” wide. I set my bolt in the groove with the flat spot on top and held it steady with a box-end wrench. Eyeballing the drill bit to the center of the bolt and using a bit of oil, I was able to drill holes dead center. I wasted two bolts before I produced two centered holes. Bolts are cheap. Keep trying until you get it right. You’ll find you can sort of steer the drill bit from side to side with the wrench, but take care not to wallow out the hole. One of my holes was a little loose, but a quick peening with a punch secured the pin in the hole and saved having to drill another bolt.

When you have two bolts with holes centered and tight-fitting pins, cut off the bolt heads a short distance from the holes. Mine are about ¼” and that looks good to my eye. I chucked the bolts in a drill and used a series of files and sandpaper to make the bolts look like specialty hardware milled specifically for a coping saw. You may prefer to knock off the burrs and move on.

Now it is time to cut the cross member to length. Insert the pins into the end pieces. Laying everything flat on your bench, install a coping saw blade and adjust the end pieces so they are roughly square to the blade. This reveals the distance between the end pieces with a blade in the saw. Set your cross member on top of the end pieces, and with enough material on each end for tenons, mark the distance between the ends. No measuring needed.

I cut my tenons using a stop block on my crosscut sled, nibbling across the shoulders and cheeks. Choose your preferred method. The goal is a sweet fitting tenon that doesn’t leave any side-to-side waggle or slop.

Shaping is next, and here is where you are let go to run free. You can leave your saw square, make it round, or any combination thereof. My shape is fairly traditional while also pretty fancy. I’ve never carved scrolls, so I tried my hand at it and I’m happy with the outcome. All that really matters in the shaping is that the ends have a place to hold the winding cord, that the saw feels good in your hand, and that you like it.

The last bit to make is the windlass. I like a smooth, straight windlass so I can slide it through the wound cord to catch the cross member rather than twisting it to one side to clear and then catch the cross member.

For the cord, jute is typical, simple, effective and cheap. I am hard pressed to deviate from tradition because it cannot be topped. There is no glue or other anchor on the cord. For a saw this small, two loops around the ends is sufficient. Cut the cord to length by laying the saw flat on the bench. Put the cut end of the cord near one end. Loop the cord around the ends and cut the cord near the opposite end from where you began. You’ll have 6 cords between the ends with two cords going around each of the saw ends (4 cords form loops and two cords are loose ends). Lace the loose lengths of cords in and out of the 2 cords forming the loop on the same side as the loose length. Tension will lock the loose ends into the looped cords and form a strong, clean, adjustable winding cord. Insert the windlass between the looped cords and wind it to tension the blade.

Experiment with blade tension, taking care not to break the saw ends. It doesn’t take long to get a feel for how much tension is too much. I over-tensioned my saw to see what would happen, and it was the jute cord that gave way first. If you need more tension than the cord provides, just add another loop. Whatever you do, you’ll want the cord or the blade to give up before the saw breaks (with your heart).

Good luck with this project, and enjoy one of the finest coping saws available at any price.