Now it was time to work on the brass disk wing disks. Each disk would fit into a shallow (about 0.04” deep) 5/8” diameter well.

Before proceeding, the first decision I had to make was when to drill the wells—before applying finish to the door slabs, or after. I try to avoid handling parts after finishing them. It’s too easy to damage the uncured finish. I also worried about seeing thin halos around the disks if the drill bit exposed any of the underlying wood. I decided to drill first and finish later. (Spoiler alert: that was the wrong decision.)

To position the wells on the doors, I created a template in SketchUp. SketchUp has a nice printing feature that will split a drawing whose dimensions are larger than a physical page into multiple pages. I had never used it before, but it wasn’t difficult to use.

The trickiest part was telling SketchUp what part of the model to print. It will either print the entire model (useless in my case because of how I organize my models) or the contents of a view. If you print a view, and there is a lot of blank space around the template, SketchUp will waste paper.

The solution was resizing SketchUp’s main window so the view closely fit the template.

In the Print/Preview dialog, I just had to select “Current view” as the print area and specify a 1:1 (actual size) print scale. In the bottom left of the dialog box, you can see that SketchUp would use six sheets of paper. (If I hadn’t resized the main window as described above, it may have used many more pages.)

Here are the six pages SketchUp produced.

The printed sheets had blank areas around the template. I trimmed the excess with a paper cutter and taped the sheets together.

To mark the hole centers in the right door slab, I taped the template to the slab, making sure the column of discs on the inside was parallel to the slab’s inner edge.

Then I used a center punch and mallet to mark all the locations (59 in all). Slight taps of the mallet were sufficient.

To do the left door, I flipped the template upside down (making it a mirror image) and taped it to the slab. I used two measurements (one vertical, one horizontal) from the first door to position the template. After 59 more taps, I was ready to start drilling.

I installed a 5/8” Forstner bit in my drill press and set the depth stop to about 0.04”. This depth leaves the top of the disks slightly above the surface.

Each door has a column of disks along the inner edge. I wanted to make sure they were in a line and equidistant from each door’s inner edge, so I used a fence and drilled the wells for both doors.

After that, I removed the fence and drilled the remaining wells. Lining up the bit was easy. I just positioned the bit an inch or so above the slab, tilted the slab so a location divot fit around the bit’s tip, and slowly lowered the slab and bit at the same time. Then I just had to make sure the slab didn’t shift when I started the motor  and drilled the well.

It took a while, but it went well. I had only two problems. The first was that one disk center in each door was in a position that my press couldn’t reach. (It’s a 13” press, meaning 6 1/2” between the post and bit, and the doors are slightly less than 13 3/4” wide.

I drilled those two wells by hand, being careful not to go too deep.

I noticed the other problem when I was about halfway through the second door. My depth stop had drifted somewhere along the way, and many of the more recently drilled wells were deeper than the initial ones. (I should have stopped every 20 holes or so to check the depth setting.) A few wells were deeper than the disks’ thickness. I’d have to deal with that later.

Now I was ready to apply finish to the doors, but I had to do one more thing first. I’ve used wood glue to affix metal to wood before. It’s sufficiently strong as long as the metal sits snugly in a groove (or well in this case). I decided to use white glue (Elmer’s Glue-All), and I wanted to make sure it had bare wood to adhere to. I found 1/2” diameter waterproof plastic stickers at Amazon and placed one in each well, hoping they’d keep dye and poly off the bottom of the wells. (I used 1/2” stickers instead of 5/8” ones so I didn’t have to precisely place each sticker.)

When I applied the finish, I tried not to slop too much into the wells.

I did a much better job on the second door (after I had refined my technique) than the first.

Now it was time to start gluing the disks. But it was also time to deal with the uneven well depths.

I measured the depths of all the wells and saw that most were okay. The ones that were too deep fell into two categories: very slightly too deep, and more seriously too deep. The only solution I could think of was spacers/shims to raise the disks in the deep wells.

I started measuring the thickness of various materials, including card stock, cereal boxes, and other things in our recycling bin. Card stock was 0.008” thick and should work for the shallower wells, and some other cardboard I found was 0.018” thick and should handle the deeper wells. I cut a set of octagonal(ish) spacers in both sizes.

It was finally time to start gluing. I was overly optimistic, and I started by adding a small dab of glue to six wells at once. Then I discovered a problem. The disks didn’t slide neatly into the wells. (In my test wells, the disks fit snugly and sat flat in the wells. But I had drilled those test wells into either finished scraps or raw wood.)

The culprit was the small amount of finish that had spilled into the wells. I also think the water-based dye and poly had slightly shrunk the wells and/or affected their edges so they were no longer perfectly round.

I used a 1/4” chisel to try to scrape square up the wells, and that helped some. Some disks  dropped neatly into place, but others didn’t. I ended up adding a lot more glue to each well and used the back of a screwdriver to apply pressure to the disks to push them into the wells. I couldn’t tap the disks with a mallet to push them into place, because each tap popped other disks out of their wells.

Ultimately, I got all the disks glued into place.

If I were to do it over again, I would drill the well holes after the doors were finished. Applying the finish would be much easier, I wouldn’t need the stickers, the wells would be clean and flat like the test cuts, and it’d feel much more secure about the glue bond between the disks and doors.

At that point, the project was basically complete. I brought the cabinet inside, installed the drawers and doors, and left things alone for a few days to let the wood shrink if it was going to. Then I adjusted the gaps, touched up the finish in the finger pulls, and I was finally finished.

In the end, I consider the project a success, but not a complete one. I’m not happy with the unevenness of the disks, but it’s a forest-and-trees (or painting versus brush strokes) situation. If one looks at the cabinet as a whole (and from a distance, as is typically the case), the problems aren’t visible. It’s only when one focuses on the details with a critical eye that they become apparent.

The best news is that our friend loves the cabinet. If I don’t have to make a bunch of support calls to deal with fallen disks over the next few months, it will get easier for me to consider it a  complete success.

Thanks to anyone who has taken the time to read this blog. I truly appreciate it.