Worklog: Tetris Shelves – Sanding, Priming, More Sanding, and Painting

Sorry for the break unexpected break gang. My daughter was born the week after the last installment, and just as I was starting to recover, I got sick and was out of commission for a few more weeks. Everything’s groovy now though, and I’m ready to wrap up the build log on my first project of 2013, a set of modular Tetris shelves for my baby’s room. If you’re just joining us now, you should probably check out part 1 and part 2 before you read this.

When we last left off, I’d successfully assembled five Tetrominos, using biscuits, glue, and square clamps to join everything together. At this point, I was pretty sure that the hard part was done, and all I had left was a little bit of painting and sanding. I couldn’t have been more wrong.

The pieces were completely assembled, and I was alternating between sanding, priming, and painting. The pieces were completely assembled, and I was alternating between sanding, priming, and painting.

Sanding Really Sucks

For a project like my shelves, sanding serves two purposes. It was important to smooth over any rough edges where my board cuts were slightly off. Typically these little mistakes amounted to edges that were just fractions of an inch off, but the effect was jarring. Any place that the boards didn’t perfectly line up, I sanded until the joints were smooth. I also made sure to sand down any places where glue had dripped or the board edges had burned during my initial cuts.

This is where an orbital sander is extremely handy. My orbital sander–a cheap Black & Decker model–uses a quarter-sheet of sandpaper and allowed me to remove material much more quickly than I could by hand. I experimented with different grit sandpaper, remembering that lower grit numbers are more coarse and will remove excess material more quickly, and found that 150 grit was great for sanding down edges that didn’t meet, while 220 or 300 were better for sanding the surfaces and edges of boards.

There were also a handful of joints with particularly bad overlap. To take the excess wood off of those joints faster, I used a rasp. A rasp is basically a cheese grater for wood. Once the offending pieces almost lined up, I also needed to sand the surface of each piece before priming to ensure a clean surface for primer adhesion.

This is a rasp. There are many like it, but this one is mine. This is a rasp. There are many like it, but this one is mine.

The entire time I was removing material, I was cognizant of the fact that it’s very easy to remove too much material and very difficult to restore material once sanded. Whether I was rasping or sanding, I was careful to remove a bit, check my seams, and then remove more if needed.

All told, the first sanding pass took between 20 and 30 hours to complete all five pieces. The sanding was absolutely backbreaking work, but the finished shelves would have looked crummy had I not taken the time to sand them properly. Sanding both allowed me to fix any minor flaws and cover any mistakes I had made while assembling the shelves.

As a warning–I did a fair amount of my sanding inside, but I used a respirator mask to block particulate the entire time I was working. I also coated everything in my garage with a fine layer of dust because I was cheap and didn’t buy the Shop-Vac attachment for my sander. As a result I’ll have to spend a weekend in the near future dusting the entire garage–not fun. The difference between a good sander with a vacuum attachment and the one I have is so great that the next time I need to sand anything substantial, I’ll probably pony up for a new sander and the adapter.


After consulting the helpful professional at my local paint shop, I bought a gallon of Zinsser B-I-N Primer from Home Depot for my first coat. Why use primer? Well, there were two reasons. Using a primer designed for use on new wood reduced the number of coats of paint I needed to finish the shelves. It also has a high particulate count, which help fill in the handful of gaps that I found in the side of the thin-ply I used to build the shelves. By using the primer, along with a little wood filler, I was able to avoid having to add veneers to the cut edges of the plywood, which saved me a ton of time. If I was doing this over again, I’d have had a light grey tint added to the primer–in its pure white form, it required two coats of the thinner colors to cover the white. Had I bought the paint from a real paint shop instead of Lowe’s, I would have been able to get a slight tint added the primer, which may have helped get decent coverage with the first coat of paint.

This is the primer I used. It's recommended for new wood, as well as sealing knots and blocking odors. This is the primer I used. It’s recommended for new wood, as well as sealing knots and blocking odors.

Applying the primer is just like painting–except instead of paint, you’re applying a viscous colloidal suspension of particles and a stinky water base. First, I wiped the entire shelf with a damp rag, to remove the last traces of the dust from the first sanding pass. After the shelves dried, I used a 2.5-inch brush to apply a liberal coat of primer to every side of the shelves, starting with the top piece. It seems like the best way to avoid drips and weird corners was to skip every other face, then come back and finish the unpainted faces after the rest of the paint had dried. To do the last side, I rested the shelf on two pieces of scrap wood, then let it dry for at least 2 hours, but preferably more. Then, it was time for more sanding.

I used a finer grit sandpaper on the primer than I did when sanding the naked wood. The goal when sanding the primer was to smooth any rough patches I’d inadvertently added when I applied primer. After some experimentation, I found that 220 or 300 grit paper on the orbital sander did a good job taking off rough spots without removing too much of the primer.

Aside from a handful of spots I missed, I was done priming. I wiped off the sanding dust from each piece with a damp rag, and was ready to paint.


Because these shelves were going in my kid’s room and I don’t have a good place to let them air out, I wanted to stay away from paint that would need a long time to dry away from people. Even though it isn’t the best paint for furniture, I ended up going with a glossy latex paint. I was able to put it in the baby’s room after a day or two of drying, and if she scrapes chunks of paint off it isn’t the end of the world.

As I wasn’t exactly sure how well the latex paint would work on furniture, I didn’t want to spend a ton of money on it on the chance it didn’t work. With that in mind, I bought quarts of Valspar from Lowes and of each color I needed. I matched the colors by bringing stuff from home that was the colors I wanted and comparing it to the many thousands of color swatches at the store.

I used a couple of nylon brushes to do the actual painting, both 2-inch and 3-inch versions, making sure to avoid drips. After a fair amount of experimentation, I found that painting adjoining surfaces was a surefire way to mar fresh paint, so I switched to painting alternate edges.

Unfortunately, because I was using bright colors of paint, I needed to apply at least two coats of paint to every shelf. That did give me the opportunity to fix a few rough patches and drips that happened while I was painting. A few moments with some fine grit sandpaper were enough to fix the mistakes.

The Finished Product!

After a few days in the garage to give the paint plenty of time to dry, it was time to set the shelves up. I originally thought I was going to have to add a sealant to them in order to prevent sticking when they were attached to each other, but when the paint was fully dry, I didn’t have any serious sticking problems. The arrangement my wife and I chose for the layout gives us plenty of shelf space and leaves each piece both well supported and locked into place. I do need to tether the shelf to the wall, but I can do that with a pair of screws and some inexpensive nylon webbing.

All told, I spent about $700 on the shelves–$220 on wood, $200 on clamps, $60 on tool rentals, $50 on saw blades, $60 on paint and primer, $30 on sandpaper, and another $50 on assorted small tools and consumables (mostly glue and biscuits). About half of that was on tools, tool rentals and other stuff I’ll be able to use again in the future–at least if I ever have a project that requires 16 corner clamps again. It took me somewhere between 60 and 80 hours of work to complete the shelves, spread out over a month or so. I guess it’s kind of a cliche, but I was surprised that finishing the shelves too much longer than actually assembling them.

The shelves turned out much better than I thought was possible. They actually look like they were made by someone who knows what they were doing. The individual piecesjoin up properly at the seams much better than I anticipated, so the individual pieces look like one unified piece of shelving when they’re assembled. My careful measurements and all those corner clamps really paid off as the good fit saved me from having to make either of the mechanisms I designed to affix the shelves to each other.

I wish I built enough pieces to take the shelves all the way to the ceiling. Knowing what I do now, I’d pay a wood shop to break down the plywood sheets for me and rent a sprayer to prime and paint the shelves, which would save me a ton of time. I’m carpentered out for the time being though. Maybe I’ll build the rest when she’s old enough to help, if she’s into them.

Worklog: Tetris Shelves – Cutting and Assembly

Last week I talked about research, planning, and rounding up supplies and equipment for my Tetris shelves. After a month or so of careful planning, I had all the stuff I needed to get started, you know, actually building the shelves.

This week, we’re going to talk about ripping big sheets of plywood down into the appropriate sizes, I’m going to learn to use a plate joiner and biscuits, and we’ll explore the importance of clamps.

Breaking Down the Plywood Sheets

I didn’t fully appreciate the challenge of breaking down big sheets of plywood without a large table surrounding a table saw. I was able to get my 4’x8′ sheets cut down to 4’x4′ sheets that (more or less) fit in my car at the lumber yard, but I ultimately needed boards that are ten inches wide for the shelves. I sketched out each 4’x4′ sheet to make sure I’d actually get all the boards I needed, but I wasn’t sure how I was going to be able to make those cuts safely and accurately.

Breaking down the plywood was a three-step process–first, I had to make some rough cuts using a circular saw to make the sheets manageable on the small table saw I rented. This was actually the scariest part of the whole process, I’m not good enough with a handheld saw to cut a straight enough line on a four foot sheet. To keep my saw on track, I made an improvised guide using a couple of clamps and some scrap lumber I had laying around. By aligning the right edge of the saw with the guide, I was able to make my circular saw cuts much straighter than I would have been able to otherwise. Starting out, I wasn’t sure how accurate and clean these cuts were going to be, so I made sure to leave enough extra wood that I could fix any mistakes later on.

As I was making these cuts, I learned the incredible value of leaving notes to yourself on the materials. After each cut, I labeled each board with it’s original position on my sketches, as well as the boards I expected to get out of the piece and any notes specific to that piece. This helped me avoid making mistakes, and also took the guesswork out of my cuts. To know what I expected to get out of a board, all I had to do was look at my markings and consult my sketches.

Next, I needed to rip the smaller sheets into 10-inch wide boards. I set up the table saw and did a couple of test cuts to ensure that its guide was accurate (it wasn’t). After I started cutting the larger sheets, I quickly realized that the portable table saw wasn’t large enough to handle sheets as large as I was cutting without some extra support. I enlisted a helper to hold the boards level as they came off the saw and from that point on, the rips went very quickly.

I did the final cuts on the miter saw. While I could have used the table saw to do this, I was more comfortable making the precise cuts on the miter saw than the table saw. More importantly, because I’d labeled each piece as I cut it, I was able to cut each piece of shelf as quickly as I could measure and mark my boards.

As planned, I about a quarter sheet of plywood left over after I was done cutting. Before I packed up the tools, I inventoried all my pieces to ensure I had everything I’d need for a complete set of shelves. I also had a couple of pieces that were either marred or a bit short, so I used the leftovers to make replacements for them.

Learning to Use a Plate Joiner

Next up was using the plate joiner to cut slots for biscuits at each joint. I planned on using edge joints with biscuits to add stability and strength to the shelves. From what everyone told me and what I read online, a well constructed, glued joint would actually be stronger than screwing the shelves together or using some other mechanical attachment. I was skeptical, but sometimes you have to trust that the common wisdom is right.

In an edge joint, the narrow edge of the board butts up against the face of the board it will be attached to. To use biscuits, you cut small slots into the facing boards and then insert a small wooden puck into the slot–that puck is the biscuit. Biscuits make it easier to align the boards properly and add more surface area for glue to hold the final joint together. The biscuits also absorb some of the moisture from the glue and swell up, which also helps to make a more secure joint.

Cutting the slots for biscuits can be tricky–they have to line up perfectly and remain tight in order to work well. Luckily, I was able to borrow a plate joiner to help cut the biscuit slots. A plate joiner is a specialized saw that makes it easy to cut the oval-shaped slot for the biscuits into each board so that they’ll align perfectly. With the joiner I borrowed, you can do a variety of different types of cuts to join boards in many different ways.

Once I set up the plate joiner’s guides set up for the type of cuts I needed, cutting slots for biscuits was straightforward. All I had to do was line up the boards where they met, make a line across both pieces of wood, and then line up the joiner’s guide with the line on each piece of wood before cutting. Once everything was marked, it took longer to clamp and unclamp each piece of wood than it did to make the cuts. This sounds more complex than it actually is.

The joiner makes it easy to make these cuts reliably, but there are a few tricks worth noting. I found that the saw made a much better cut if I gave it a moment to get up to speed before plunging the blade into the wood. It’s also super important that you the blade only moves in or out of the wood on a single axis. If the joiner tilts while the saw is running, the slot for the biscuit will be too big, and your joint won’t be as secure. Finally, a large bench vise would have been handy for cutting biscuit slots in the faces of boards. I have a small vice, but it isn’t big enough to hold a 20-inch board without wobbling. I was able to make the necessary cuts by contorting myself, but this definitely wasn’t good for my back.

Gluing and Clamps (Bonus: Learning From Mistakes)

After months of planning and a couple of days of hard work, it was time to finally assemble my first Tetromino. I started with what seemed like the simplest piece–an L. My initial plan was to glue a few corners and clamp them long enough for the glue to set (about 30 minutes, according to the instructions), then move on to the next corner. I bought four corner clamps and figured I had two days of work in front of me.

Using this theory, I assembled the first L. I was extra careful, giving each joint a full hour for the glue to set. The first couple of pieces went together perfectly, but then disaster struck when I tried glue the last piece. I applied the slightest bit of pressure and…

… the whole thing collapsed. I said a naughty word and got back to work.

Luckily, this failure taught me lessons that saved me from even bigger mistakes later on. I immediately realized I needed more clamps–unless I wanted to take a month to finish gluing the shelves, I needed enough clamps to glue an entire piece at a time. With that in mind, I trotted down to my local Home Depot to stock up on corner clamps.

While there, I realized my second mistake. I didn’t actually have any clamps that could hold the joint together while I was securing the corners. Instead of using a tool, I was relying on mere human strength. In this case, the tool I needed was a bar clamp. Since I was already in for $150 worth of corner clamps, I went big on the bar clamp, and got a 33-inch Irwin bar clamp that you can cinch down one handed. Much later, I realized that some of the trickier clamping would have been easier with two of these, but I was able to make do with one.

By buying a big bar clamp and increasing the total number of clamps at my disposal to 16, I was able to glue and assemble one Tetromino every 12-24 hours, all while maximizing the strength of my joints. Yes, I had to cleanup the surfaces where the glue didn’t hold and cut the broken, glued biscuits out of their slots, but if I hadn’t realized that I needed more clamps, it’s likely the shelves would collapsed the first time I put them together.

Seven days later, I had a more or less complete set of Tetromino shelves, and I thought the hard part was done. After all, all that was left was sanding them and cleaning them up so they’d be ready for a quick prime and paint. More on that next week.