Milk paint doesn’t have a lot of body especially the light colors. It acts more like a wash coat. So for the mustard I’m going to need to apply three coats. Being water based it will raise the grain so I did a quick wet sanding to the pieces before painting. To apply the paint use a foam brush or a chip brush. Expensive brushes aren’t needed for milk paint, it won’t make it go on any better. After each coat has dried it should be sanded back to level the paint, remove any hairs that came out of the brush and remove the funny bubble marks. I used a fine grit sanding sponge. This makes a lot of “dust” so have some rags handy or maybe the vacuum cleaner and brush attachment.

And don’t panic about the color yet…

To get the final color and protect the paint from water (milk paint is durable but will spot easily) it is necessary to apply linseed oil (BLO). This brings the paint to its final color and helps seal it against water stains. After the oil soaks in you can apply a top coat of wax, shellac or varnish. I chose to apply wax.

Once I got the door re-installed it is back to not fitting the opening. However this time it is just a hair too wide. Since the milk paint is water based, it swelled the wood just a little more than the thickness of a dime! Dang it! But I’m not going to do anything about the fit for at least a week. My bet is that the wood will go back and the door will fit again. If after a week it won’t fit then I’ll take it off and re-plane the door.

All that is left is to make the knob and latch.

The back of the cupboard will be 1/2” boards of slightly varying widths. The boards will be ship-lapped and rest in a rabbet on the back of the case. Cut nails along the top, bottom and the fixed shelf will hold them in place.

Step one, mill down the board to 1/2”. Decided to just use the power planer for this. Noisy but pretty quick work. Didn’t bother with pictures here.

Next the boards are ripped to width. The widths are all between 4” and 5-1/2” and side by each it is about 28-1/2”. In the end I’ll need 24” to fill the back. Good to go.

Time to get out the rabbeting plane. I could have done this with the table saw and a dado head, or with my router handheld or in the table and a straight bit. But I recently picked up a Chinese copy of a Stanley 78 ($0.99 on eBay but about $12 in shipping, argh!). I’ve honed the blade and done some preliminary fettling of the plane. It needs a little more work but its OK for now.

Here’s the plane showing the depth stop, nicker (retracted) and fence:

And here we see the other side with the fence support rod and depth adjuster (such as it is).

And from the top you can see the cap iron and screw and both the “regular” frog and the bull-nose frog at the front of the plane.

For $13 its a pretty decent worker. I decided to take a risk on what might be a bad casting (it isn’t terrible but does need work) because all the other 78’s on eBay seem to be missing the fence, depthstop or both!

Once again, a three finger grip on the right hand, index finger pointing along the plane for balance. My left hand is riding with my thumb and index finger riding on the nose of the plane and the rest of my fingers are wrapped around the fence post. The depth is set kinda deep but since this is a long grain cut it works fine.

Here are the ends of finished boards.

You can see a little bit about what I mean when I say the plane needs a little more fettling. The fence is not 100% perpendicular to the body. This combined with running along the narrow edge and my limited experience with the plane produces some out of square rabbets. However since this is for ship lap and the joints will be left a little slack anyway, no big deal.

And the coolest part of using a rabbeting plane, the shavings! (Size US11 for scale…)

Once the boards are assembled side by side, the final width is about 26”. Since I needed 24” this is great. I’ll just rip about 1” off each of the two outside boards and I’m good to go. But that will wait until after a quick dry-fit to confirm.

Playing with the rabbet plane took most of the afternoon and now its dinner time. Perhaps some hossenfeffer…

OK, so the door is hanging temporarily. Since I don’t have back boards on the cupboard yet, this is a pretty easy scribe. Zip, zip, mark, mark.

I don’t have a terribly fancy bench so a little bit of creative clamping is in order. People make fun of Harbor Freight for a lot of stuff but their wooden handscrew clamps are a darn good deal. Can’t have enough of them! Use one at each end of the workpiece and then clamp them to the overhang on the bench. Works pretty well, if a bit fussy sometimes.

So here you see the scribed line at the “fat” end of the board. The white block in the foreground is paraffin wax for the plane sole. The scribe line tapers away to nothing at the other end.

Here are the weapons of massless destruction… The #6 is set for a pretty decent shaving, maybe about 0.005” to 0.008” thick. Runs stuff down reasonably quick. The #4 is set for the wispy thin 0.001” shaving so I can fine tune things and deal with any tearout. FYI, a good idea to pre-set the irons on some scrap before starting. Otherwise consider having the blade pulled in and slowly drop it back down on some test strokes on the workpiece. Better to take a few thin shavings at the beginning than gouge things terribly!

Chicken wing activate!

Most of the work is done with the #6. A few shorter strokes at the heavy end of the scribe and slowly work backward to the other end of the door. In these two pictures, besides seeing my scrawny arms, you can see that I have the plane a little bit skewed as some insurance against tear out. In the second picture my left hand is not grasping the tote. Instead it is wrapped around the sole of the plane just in front of the mouth and riding against the far side of the door. This helps keep the plane level on a narrow board. It doesn’t hurt to check with a square once in a while too.

The other planing tidbit I want to point out is on my right hand I have 3 fingers on the tote with my index finger pointing along the blade/breaker. This does a few things for me. First I can’t fit all four fingers on the tote and second it balances my hand strength a little bit. Sort of like when you grip a golf club you typically have the little finger of your dominant hand off the grip. Same idea here.

For the #4 plane, I have two fingers off the tote and pointing down the blade. Mostly because on the Millers Falls smoother, the tote is a bit on the short side (for me anyway).

And here we see the scribe line is just gone.

When I scribed, I ran the pencil to the inside edge of the stile so I wanted to plane just to the “far” side of the line. Now I can re-hang and double check and if necessary take a few more swipes to even up the gap.

And that is exactly what was needed. The door fit but scraped in a bit. Back off and a couple of heavy cuts down the length followed by a few smoothing passes and here we are! (Excuse the grubby sneaker, I haven’t made a catch for the door yet so I’m holding it shut for the photo…)

A pretty decent fit. I may take one or two more thin passes next time the door is off but for now I’m leaving it alone.

Time to enjoy a beverage from my favorite mug. :D

1) Compute final panel size based on door. I know that my rail and stile set creates a 3/8” deep panel groove.
Width:
2 x StileWidth + RailWidth@Shoulder + 3/4” = 13-7/8”
Height:
StileLength – 2 x RailWidth + 3/4” = 26-3/4”

2) Allow for panel expansion because it is a solid wood panel. Most of the movement will be in the width so allow 1/8” to have a finished panel width of 13-3/4”. In length, only about 1/16” is necessary so the length will be 26-11/16”

3) Cut stock, mill to 1/2” thick and glue up for the panel, over sized in length and width of course. Note this is 1/2” thick while the door frame is 3/4”. So the panel will be more or less flush on the front but recessed about 1/8” on the back side. I’m using 1/2” because I wanted to minimize the door’s weight. Using 5/8” I could have the front and back flush but I’m too big to fit inside the cupboard and see the back of the door so…

4) Cut out panel blank.

5) Setup up router with panel raising bit. This is a bevel profile.

6) Cut a test piece. Also seen in this test piece is an experiment at cutting a back rabbet so that the panel will can seat all the way into the groove. In this test piece, the rabbet is about 2/3 too deep. The finished panel has just a whisper of a rabbet cut in the back.

If you use a cutter like the one I show above that does not have a second part to make the back rabbet, it is a simple matter to use a dado bit or even the stack-dado cutter on a table saw or a rabbeting plane. I opted for a 3/4” bit in the router table with a little more than 3/8” cutting width exposed and maybe a shy 1/16” cutting depth.

So here is the finished door front and back.

Some other notes about the door. Yes, I realize the grain is going two different directions on the boards. Ideally, it all goes the same way but at the last minute during glue-up I decided to flip boards to make sure the end grain is alternating. I should have also turned them end-for-end. Painted panel, so no big deal. A giant no-no for a clear finish!

The last step is to fit the door. I mentioned before that my case had racked. I checked it again today and it seems better. I think it is a humidity issue as my AC was down for a few days last week so the house got pretty funky (96F + 40% humidity inside!). I still need to do a little trimming on the door and maybe a little in the face frame. There is maybe 1/16” of material that needs to disappear for the door to slip in. So if do the dime-trick to space it I need to loose maybe 1/8” of material. I’ll split that up between the two sides of the door and the two face frame stiles. That should hide any unmatched dimensions pretty well.

With a few test cuts and a little tweaking you can get a very good fit. :)

First step is to attach the 1/2” thick piece with the single ogee:

Next we add the larger built-up molding:

And here we can see another interesting point. I went ahead and routed the roundover on the upper face frame before installing it and routed it all the way to the end of the board. What this means is there will be a small gap between the corner of the face frame and the corner of the molding. Not a big deal and I knew this would happen. What I could have done is stop the roundover on the face frame short of where the molding would be, then I could chisel in a roundover after the molding was in place, or some other simple transition. I opted for the “easy way” because the gap is small and from a distance of about two or three feet you have to know its there to really see it.

There is a little bit more fine trimming to do on the back side but I can get to that later. I thought I trimmed the last side piece short enough but it is hanging over about 3/32”. At least it wasn’t too short! This is pretty easy to deal with even with it attached to the case. I have a pull-saw that will track just fine and zip the overhang right off next time I have the case down on its back on the bench.

This took me a while to get the molding attached for two reasons. First, I left the case standing, didn’t feel like dragging it back out to the garage and laying it the bench. Second, it takes several test cuts to set up the saw and then I’m probably over cautious and take too many “sneak-up” cuts.

By the way, blue tape works great to keep the saw from tearing out the profiled edge.

The next step is to work on the door’s rails and stiles. I measured the opening and to my surprise it is no longer quite square. It has racked about 1/16”. I’m not sure why, could be the addition of the upper face frame and clamping operation caused the racking. Could also just be the change in humidity. Standing back and eyeballing things, you can’t see the difference. But it is important for setting in the door. I’ll take it into account and trim the door a little bit when the time comes. I can also trim the case stiles a tad by aggressively sanding with a block. Doing both the door and the frame will split the difference in the un-squareness and might be the best way to camouflage things.

(The sides are straight, the photo shows them bowed because I didn’t bother taking the camera out of its wide angle view and it has barrel distortion as you move off the center axis. This is common on most point-and-shoot cameras, the solution is to zoom just a tiny bit. But I didn’t, just lazy today.)

It is interesting to me about the top molding. From some angles, it looks great and it good proportion. From others it looks a bit heavy. If I were to make another piece, there are two changes. First I’d extend the overhang of the center shelf about 1/2” more to add a little more balance and I might use a 1/2” or 5/8” round over (which I don’t have) instead of the 7/8” to shrink the molding size a bit.

But I must reserve the final judgment until the case is painted and some dishes are setting inside. With the paint, dishes and door fitted, I can imagine the molding might be brought a little more into proportion. Just closing in the bottom of the case will add some visual weight.

Back to the rails and stiles. The Mike Dunbar article goes into making the door using through mortises and pinning it along with wedging the mortise. I’m going with a cope and stick router bit set instead.

To set up for the coped portion, I use a simple coping sled at the router table. Just a piece of 3/4” MDF and some 1/8” hardboard. Good ole’ Harbor Freight supplied the toggle clamps.

A few passes and the coping is done in the ends of the rails. Next I switch to the stile bit and run the rails and stiles both.
The bits:

Here is the stile cut on the test piece:

And a finished rail piece:

I suppose one could argue that the Shakers wouldn’t have used such a joint. But I wonder about that. They (1810, Sarah Babbet) invented the circular saw blade, they invented machines for cutting tongue-and-groove joints and wainscoting beads. They seem to have been quite “modern” in their construction methods, I see no reason that some of them would not have embraced shaper/router tables and bits.

So, next up is to glue up a panel for the door and cut a bevel profile. I’m going to make the door only about 5/8” to 1/2”.

What was hard though is I decided it had to look nice AND could only be made with bits I already had. The original article I’m taking my inspiration from has some sample moldings and while I don’t have all the same profiles, I can get pretty close. But his big case molding was another story.

Step one was to get out the graph paper and sketch the basic profiles I have at hand then start putzing around until I got something with enough weight to look right.

So roughly, the profile I’ve drawn in the lower right is where I want to go. But this still isn’t quite large enough. Mike Dunbar’s article has a piece behind this with a smaller ogee. I have a 5/32 ogee bit so I’ll start with that piece. Here is one end of a test piece – excuse the bobble at the end, it will get cut off.

Next I need to make the lower roundover. The block will be 1” thick so I milled down some of my remaining 5/4 and then chucked up a 7/8” roundover bit. This is a bit of a monster as roundovers go, good for a table edge.

So two things I need to do in preparation for this, one is set up some feather boards on the fence. The second is to set up some stop blocks behind the fence so that once adjusted to just ride the bearing, I can swing the fence forward and make my cut in two passes.


One of these days I’ll make a fence to match the size of this table, this fence I made for the previous table and it is about 2” too short to span side to side. But this one still works OK.

So, two passes to get the full profile, then rip it off the test board (SNIPE! Doh!).

Here’s a better piece.

Next up is the cove portion. I’m going to use a 1/2” cove. But after cutting a small test piece I realized it will be a nightmare to get a smooth transition from roundover to cove. The solution, make a small step-back shoulder and accentuate the size difference instead of trying to hide it. The small shoulder is made by first cutting a groove with a 1/4” spiral bit about 5/8” in from the edge of the board, then running the roundover leaving the little stepped shoulder. Forgot to take a picture of the groove cutting. Sorry.

This has also been ripped to width for the final stackup. And here it is, just a little double sided tape holding things so I could see it.

So all that is left is to glue together the roundover and cove pieces and do a little detail sanding. A few boogers to clean up but I have more than enough length that I should be able to cut around the snipe from my planer. And the 5/32 ogee board will be ripped to the correct width too. It gets applied to the cupboard first, mitering the corners. Then the larger glued-up ogee (I think technically, this is more of a reverse ogee but I’m not 100% on that) goes on. Once painted, the glue line should disapear all together! :)

Next I want to get the molding applied then work on the door.

Notice the glue run near the center of the side. Not a big deal, I caught as much as possible. The rest gets scraped and sanded when the clamps are off. Easy peasy.

And here it is standing up again plus a small closeup.

So, three parts left before painting begins. I need to make the trim for the top, make the door and cut boards for the back. I won’t install the back other than to lay in on for size. The back will be tongue-and-groove 1/2” boards. I’ll leave it off until the painting is done to make things easier.

I glued up the two stiles and rail for the upper face frame. Something to keep in mind when doing any sort of work is the order of operations. Round over before glue up? After? The reason I mention this is because I made a strategic choice about when to chamfer the inside edge of the frame. I could have done the chamfer on the pieces before glue-up and also cope the shoulders of the rail to fit or miter the inside corners. If this cupboard was to have a clear finish, the coped joint or mitering the inside corner of the chamfer is the way to go. But as I’ve said, this is going to be a painted piece. So I opted to do the chamfer AFTER the glue up. This means the “funny” of the chamfer into the corner will be camouflaged. The “funny” will be quite clear in some pictures below.

What are the implications of doing things this way? Well here you see the rough routed chamfer. The burning is because I had to slow my movement of the router (working handheld this time) as I turned the corner. Slow feed == burning.

No big deal, just a fact of life. But what is more important, you see that a router cannot make a square inside corner. Just can’t, the geometry of the bit rules.

This is an easy situation to rectify, I just strike a pencil line to extend the two chamfers into the corner.

Next I push my chisel in on that line, using a block cut to 45 so I’ll match the router cut chamfer. Not going in far, just enough to create a little bit of a guide for the paring cuts to follow.

FYI, a handy jig to have for the table saw in addition to a miter sled and a panel sled is a 45 miter sled.

Makes things like 45 corners or in this case a small paring block a snap.

Just pare out the rest of the waste, which also gets rid of most of the burn mark.

Just a touch with a sanding block and it is done.
And now you can see the “funny” quite clearly. If this was not going to be a painted piece, it would look very strange in the corner. This is why you cope a joint when putting a profile on an inside corner of a piece that will receive a clear finish. The milk paint will cover this and it won’t be visible later. And it only took about 5 minutes to pare the two corners versus the cussing and fussing of coping the shoulder of the tenon.

The last router operation on the face frame is to add a round over and lip to the two outside edges of the rails. Again, handheld routing. But here is another common problem. Lots of routers use what is sometimes called a PC or PorterCable base. Mine does. The good news is that lots of accessories and bushings fit the PC bases. The bad news is that you can’t pass a bit larger than about 1” diameter through the stock sub-base of the router. Since I have two bases for my router, plunge for most hand-held work and fixed that lives in the router table there is a very easy solution to this problem. Take the fixed base and attached router plate from the router table and use it as a handheld base. No where is is written in stone that you must have a round base on a handheld router. My router plate has one of those snap-apart rings that lets you pass larger bits by removing inner rings. Problem solved.


One other benefit, this plate is quite wide and so I can have plate in contact with the entire 3” width of the stiles. This makes for a more stable handheld router pass and thus safer.

I’ve mentioned before about changing the bearing size to expose different parts of the roundover bit. This closeup really helps to illustrate that point.

The small bearing is inside the straight cutting portion of the carbide. If I put back the 1/2” bearing then I would be back to the smooth transition roundover.

The tenons will be nominally 1/4” thick, 1” long and 3-1/2” wide (the top rail is 5-1/2” wide). Nothing special about those numbers, just nice easy numbers to work with while laying out the parts and I have a 1/4” spiral bit.

Lots of different ways to cut mortises, all chisel work, brace and bit + chisel, drill press + chisel, fancy loose tenon tools, power mortiser and of course router. I made a simple mortising base for my router some time back. This one is not the self-centering type because when I made it, I did not need a centered mortise.

Here you can see from the underside the two sliding fences that will ride against my work piece. Red oak for the fences and 1/4” hardboard for the base plate.

The stock is only 3/4” wide so I’m going to clamp the two stiles together to get 1-1/2” surface to ride. Layout the start and stop lines and set up a couple of blocks as stops to keep the router bit inside the lines. I’ll square up and clean up with a couple quick taps of the chisel.



Set the two sliding fences so the bit is centered over one workpiece. To route the second piece, you just flip the jig around.

The mortise is going to be 1” deep so don’t take all that in one pass. I’m using a plunge base so the maximum depth is set to 1-1/16” (leaves a little room for glue later) and I use the turret stop to make the mortise in a few passes.

Done with one, now flip around and cut the second.

It took longer to write this than it did to cut them. The longest part of the set up was unscrewing the plastic sub-base from the plunge base and then screwing on the mortising jig base.

I don’t have a power mortiser, or true mortising chisels (only bench chisels) and my drill press is pretty darn small. So to make a mortise near the end of a relatively long piece, I need to pile up blocks to support the far end. With this rig, I could clamp everything to the workbench so it was very well supported over its full length. And poplar cuts so nicely with a spiral bit.

Tomorrow I’ll clean up the corners of the mortise (notice I did slip a little over the end line on the first one, no big deal). Then I’ll get the tenons roughed at the table saw and pare them down to fit. After that, there is a detail to route on the outside edge of the stiles and the whole thing can be attached to the carcass. After that, all that is left is to make the door and molding. :)

Anyway, to the cupboard! The plan had been to get the lower case stiles attached as well as the upper face frame. Then I could attack the door and top molding (that will be interesting). As it was, all I got done was to mill the boards for the lower case stiles and the stiles and rails for the door so they are all the same thickness.

Here you can see the two lower case stiles attached (glued). I’ll run about three of the cut nails through each to give some mechanical fastening too as the door will hang from these.

As you can see by comparing to earlier photos, having the lower case stiles in place hides the cleats.

And the curve at the bottom of the stiles completes the front leg shape.

The shape on the outside face corner is a simple roundover but with the smaller bearing as used for jazzing up the cleats (in case you do see them) so you get a little shoulder on both edges of the roundover.

I’m not sure what those two spots are on the foot detail photo. In the picture they look like glue smudges but when I look at the piece itself, I can’t see them. I’ve seen this sort of thing before where the spectral response of the camera is wider than our eyes. It may very well be a glue smudge. Not a big deal in a painted piece but I will need to do a little scraping and sanding.

So now some evening this week I’ll attack the upper frame. It will have the same edge detail as the lower frame. Notice that the lower face frame is only two stiles, no rails. The door when closed will provide the horizontal members.