FLW Cabinet - #26 Final Preparation & Finishing

With all the staining done it’s time to start finishing, well almost time.  Before starting to spray the clear topcoat there is still a little work to do.  Since the sides, fronts and backs will be finished prior to assembly I need to put a temporary filler where the pegs and faux tenons go.  That’s to assure the inside of the hole does not get a finish on it which could prevent a good glue joint when the actual pieces are glued in place once everything is all together.  This shows the pieces to be installed after assembly.

The fillers are made out of some leftover cedar which works out just fine as it is an easily machined nice soft wood.  I started by ripping strips slightly oversize then using the thickness sander take them down to a slight friction fit.  Once that’s done a small bevel is added to one end by making a single pass on each edge with 60 grit paper as shown below.

The fillers are cut to length on the table saw using a stop block mounted on the rip fence.

The faux tenon fillers follow a similar process except that because the legs are tapered the outer face of the filler needs match that taper.  Matching the taper angle is easy enough since I have the cutoff from when the legs were originally tapered.  That goes up against the cross-cut fence then the end of the blank is trimmed.  After that the tapered spacer is removed and the faux filler is cut to length using the stop block mounted on the rip fence.  The photo below shows the setup with the tapered spacer up against the cross-cut fence.

This is what the peg and faux fillers set in place look like.

The topcoat I will use is nitrocellulose lacquer sprayed on with an HVLP (High Volume Low Pressure) spray gun.  The last time I needed to get lacquer they had quit making that particular type so I bought all 6 gallons they had.  Now all that’s left of that is just under two gallons which is not enough to do both cabinets.  A visit to the local supplier provided me information on a compatible item so I got a gallon to do a test to see how it matched up.  Below is the old can on the left and the new one on the right.

To do the finish test I took two different solid oak pieces plus a plywood scrap then sanded and stained them the same way as the rest of the cabinet.  They are cut in half, labeled and one half is given two coats of the old lacquer (left) while the other half gets two coats of the new lacquer (right).  After an overnight cure I examined them and could not tell the difference.

Below is a photo after one coat of lacquer has been applied.  I started with the non-exposed unstained faces of the bottoms and sub-tops.  In the photo below they are the four light colored pieces in the center.  Next was the underside of the six shelves at the top and left followed by the tops themselves at the bottom.  After letting them cure they were sanded with 400 grit paper.  Another coat of lacquer followed by an overnight cure then wet sanding using 1800 grit paper.  The third coat is the final one which the next day gets wet sanded or really polished with an 8,000-grit pad.  I could do two coats in one day but have too many things going on to do that so it takes the better part of 6 days to do all sides of the pieces.   More if the humidity gets too high which rarely happens here in the desert but it did during the finishing process so I lost a few days.  First time in 10 years here that it has been a concern.  The problem with high humidity is moisture can get trapped in the finish then the lacquer forms a skin before the water vapor can escape.  That causes the lacquer to turn a milky white or blush.  Not something you want to see happening.  I did have one piece with a little bit of blush to it.  That was fixed by spraying another coat with a much higher percentage of thinner which I think dissolved the surface of the previous coat and let the moisture escape. 

The faux tenons are done concurrently with the first batch of pieces.  To keep unfinished wood at the glue joint when they are installed, I use painter’s tape to mask off the glue joint areas.  To make spraying easier they are clamped between scraps of wood held in place with more tape. If you look closely you can see the marking knife lines made to define how far they needed to go into the mortise.

Once the tenons were done the handles are next.  They are finished all at once just like the faux tenons.  A couple of screws run up into the mounting holes provide a handhold while they are being sprayed and a piece of plywood with a hole for the screw to set in holds them upright until they dry.

After nearly two weeks of working around the weather the first batch of pieces were done and set aside.  The next group is the four sides.  Since there are fewer of them and the weather still was not cooperating I set up an area in the shop to spray.  With this setup the only thing I had to worry about was the humidity level.  In the upper right of the photo you can see the other three sides drying.

Next is the backs and front.  The back is the biggest, heaviest piece and while the front is the same size it is open so it weighs less.  However, both of these pieces are just too big to spray and then try to move without messing up the just applied finish.  Given the space constraints that means just a single front/back set can be sprayed at once.  Here you can see how the two pieces are set up for spraying.

Once sprayed the temporary plugs for the faux tenons and the pegs are removed along with the tape that has kept the area where the sides will be glued and screwed on clean. 

The last or almost last of the pieces to be sprayed with lacquer is the doors and I am glad this phase is over.  I still have to do all the pegs but that will come a little later.  It seems like I have been working on staining and lacquering the parts forever!  In reality it took just over 5 weeks using more than a quart of oil stain, half a quart of gel stain, over three gallons of lacquer and almost 2 gallons of thinner to get it all done.  Now at long last I can start putting everything together.

Next up – Assembly

FLW Cabinet - #25 Sanding & Staining

Final sanding is next, everybody’s favorite part of a project – or maybe not.  To make it less of a chore I sand to a nearly final surface as the project goes along thus breaking it up into smaller bites.  Now before the staining can get started every surface needs to be checked for chalk or pencil marks, dents, smudges, dirt, etc.  Really anything that could detract from the final appearance or hinder the wood from taking the stain and finish.  The solid oak will get sanded to 220 grit and the plywood to 320 grit.  Any exposed end grain will go to 400 grit.  The cabinet is taken apart in pretty much in reverse order it was assembled: shelves, handle, top, door, sub-top, bottom then the front followed by the sides ending with the back.  No photos as the sanded pieces look just like the un-sanded ones plus photos of someone sanding is boring with a capital B.

Way back in post #2 I noted a pair of flaws, one knot and a bark inclusion that needed to be addressed later.  Well, later is now.  Also, while working on the shelves a crack appeared in the back side of one of the rear edging pieces.  Since the process is the same for repair the bark inclusion will be used as an example. Here is what it looks like once all the loose bits removed.

Along the way I have done some experimenting with different materials to use as fill.  After considering and testing both black and clear fill I settled on black.  The product that ended up testing the best is a black epoxy although it was a bit of a problem in finding a true black one.  Most of them cured out toward the gray side.  It took three applications of the epoxy to completely fill everything and here is what the cured second layer looks like.

To flatten the epoxy out a coarse sandpaper is used to knock off the high points roughly smoothing things out.  That’s followed by a scraper to gradually bring the filled areas down flat ending up sanding with 320 grit sandpaper. 

Once the flaws have been patched the finish sanding can be completed.  After the sanding is done all the pieces get the sawdust cleaned off them and stacked up in the storeroom as shown in the photo below.  The shop is then cleaned up really, really well.  That includes blowing off or vacuuming the sawdust from all the equipment and flat surfaces so I can have a clean minimally dusty area for finishing. 

Finishing starts by putting the oil-based stain on the back side or less visible sides of the pieces.  Once that dries overnight, they will be flipped and the show side stained.  There is not enough room to spread everything out in the shop so the first batch consists of everything except the door frames, front and back assemblies, handles and the faux tenons.  

Before applying the stain to the front/back frames I need to mask off the areas where the sides will join.  That’s because they will be glued together there and I need bare unfinished wood for the glue to work.  A strip of blue painter’s tape takes care of that.  To apply I use the marking gauge to draw a pencil line down the legs at the edge of the tape where the side will be attached.  The tape is then carefully applied right at the pencil line.

As the plywood back has already been stained, I needed to keep the stain off it while staining the legs.  A wide putty knife held tight at the leg/plywood joint protecting the already stained back while allowing me to apply stain to the leg.

Here what the stained back looks like with the tape in place.

As a progress shot this is what the bark inclusion flaw looks like stained.

Before applying the gel stain I decided to apply tape to the edges of the sides.   This is where the sides and the front/back will be glued and screwed in place.  I could have used the ½” wide blue tape but wanted a bit more width unfinished in case the taped areas did not match up perfectly.  The next wider tape I have is 24mm wide or just under 1” which is too wide as 5/8” is what is needed.  To get that the first step is to apply the 24mm tape holding it back from the edge about 1/16”.  This is what that looks like.

The next step is to carefully roll the tape over and onto the side as shown below.  Now it is ready to be cut to the appropriate width.

Cutting the tape is done with a marking gauge except this one does not have a pencil in it but a curved razor-sharp knife blade.  In this case the blade started out life as an X-acto knife blade that I reground into the curved shape then sharpened.  In case you are wondering the gauge is made out of scraps of Leopardwood left over from the Snare Drum Project.

Once cut the excess tape is peeled off and I am left with the edge covered except for a narrow band on both sides.

One item I forgot to mention is on removing the tape.  To help prevent pulling any of the thin oak veneer off the tape should be pulled up and away like this.  If you were to pull the tape down and the veneer on the top edge is loose the tape could pull a piece off.  I know the tape here is blue.  That’s because I ran out of the green tape and did not think about taking this photo until I had pulled all of it off.  Once that's done I can start the second phase of staining which is the application of the gel stain to all the surfaces. 

The sides are the last four pieces to be stained and I am glad it’s done.  Putting the two different stains on all sides of all the pieces took a couple of weeks.  It did not help that I could only get about a third of them spread out in the shop at once, could only do one side at a time and needed to let them set overnight before moving on. 

Next up – Final Preparation & Finishing

FLW Cabinet - #24 The Handles

Making the cabinet handles is next.  The design is from an Arts & Crafts table I made some years back.  The form fit the piece and I thought they would look right on these cabinets.  Here is what they look like on the original piece.

As an added plus I already had 3D drawings of them so thought I could print out full size plans, attach them to the oak blanks and cut away. 

Unfortunately, after taking a look at the printed out full-size drawing and holding it up to the cabinet the scale did not look quite right.  The form was good but the size was too small so I made a test piece scaled up to what I thought was about the right size.  After clamping it to the cabinet, looking at it and using it to open the door I found that it needed to be upsized just a little more.  Below is the original paper handle, the first test and an in-progress final version.  As I am still unsure about what the final length needs to be so it’s a couple inches longer than what I think it should be.  The process for making it was similar to the bottom rail work in post #12 Details on Legs, ProfilingRails & Starting the Top so I won’t go through it again.

Once the handles were cut out, I softened the edges using a router with a ¼” radius bit.  The rounding sure does make the handle a lot more friendly to the touch.    The top handle is before routing and the bottom is when completed.  The “NO” is where the handle gets mounted to the cabinet which needs to be flat so that reminds me not to route that edge. 

Most of the time my deigns start out with a rough sketch then I go the computer and do a 3-D model.  However, sometimes old school hand done drafting is the way to go.  Coming up with the support for the handle is a case-in-point.  Here are the tools and the drawing done in the shop for that part.  Look to the right above the left clamp at the bottom and there it is drawn full size with dimensions.  Well, almost all the dimensions as the final length is still a little bit up in the air.  The mechanical pencil next to the circle template is the first one I bought while in high school so it’s well over 50 years old and still gets used on a regular basis.

Next is to work on the support block that goes near the end of the handle based on the drawing I just finished.  It starts by making a block matching the height and width shown in the drawing and about 7” long.  That’s more than twice as long as what’s needed for both pieces but is safer as it keeps my fingers farther away from the cutters.  Below you can see about where the block will go and the layout on the end of the block.  Tape is used to do the layout on to make it easier to see.  The “U” shaped opening at the top gets cut away so the handle fits in it.  Very, very little room for error here as the radius of the handle and the radius of the cut in the block have to match as does its width. 

A cove bit will be used to route the arc and make the transition to the flat bottom.  However, to reduce the amount of material the bit has to hog out most of it is cut away on the table saw.  Here you can see the setup and a closer look at how the waste is cut away.

With most of the oak removed I switched to the router table and the cove bit and started to remove material a little at a time sneaking up on a fit by moving the fence a little then making a pass along each side of the opening.  Because the cove bit will not cut deep enough in a single pass, I had to work in the width first.

Once the width is right the bit is raised a little bit at a time until the proper depth is achieved.  Because the cove bit is ½” wide and the cut opening is ¾” the bottom of the opening is not flat but has a couple of ridges.  To get rid of them the piece needs to be moved over and routed again.  Since I did not want to move the fence a narrow shim is clamped to the fence and the piece run through then flipped 180 degrees to cut the opposite side which almost completely flattens the bottom. 

At this point there is just one little bit of routing to do and that to route a 1/8” radius along the top edge of the piece. 

Now I can move to the chop saw and cut the handles to their final length since the blanks were intentionally cut long.  To get the length the handles were clamped on the cabinet door and marked at what looked right.  Same process is used to size the support block length, which comes from the just completed blank.  With all the handle parts cut to length those exposed sharp sawn edges need to be softened.  This is way too small to be done with a router or power tools and will be done with sandpaper by eye.  To help my eyes out came the 4x headband magnifier.  I know it may look a little funny but they sure make doing detailed work a lot easier.

This view of the handle shows the connection between the support block and the handle itself.  As you can see the two pieces have to fit together perfectly, there is no room for error.  The routed radius and cove have to match up perfectly as does the width of the slot and the width of the handle.  The degree of accuracy, the design decisions and the prototypes that had to be made are the reasons it took a week to get them from the initial drawing to completed handles.

After all the sanding is completed here is what the two pieces of the handle look like put together.  I can tell you that the mounting height involved quite a bit of thought clamping them in place, standing back and looking at them then adjusting up and down until a balance was established between the hinges, pegs and visual vertical location. 

The mounting to the door is via a couple of long wood screws.  The process I used to mount the hinges is:

1. Mark on the door where the screws go into the handle.
2. Drill a pilot hole through the door from the outside face.
3. Clamp the handle in place then drill the pilot hole into the handle from the door’s inner face.  The top hole is drilled first with the drill left in place to keep the handle from shifting while drilling the lower hole.  The photo shows what that looks like.
4. The clamps are removed, the handle set aside and the hole in the door drilled larger to provide clearance for the screw.
5. The handle is then screwed onto the door.

  

The handles complete all the wood working (I think) except for the 6 million or so pegs that need to be made.  Well, maybe not quite that many but it’s a bunch.

Next up – Sanding & Staining

FLW Cabinet - #23 Building Tall Fence Sled plus Finishing & Installing the Tops

First is to build the tall fence assembly that gets attached to the base.  Here is what it looks like mostly done.  Because of the angled braces the clamping is a little different.  To allow for some adjustment when finished all the pieces are screwed together without any glue.  That way if needed shims can be added to square the face of the fence with the table saw top.

Once all the parts are screwed together the piece is bolted onto the sliding base.  The top photo shows the back with its bracing and five of the six bolts holding it to the sliding base.  The bottom photo shows the face of the fence where the top will be clamped on.  Checking the fence for squareness against the table saw top gave good news as it is so close to being dead on that no adjustments are needed.  The vertical piece is an old plastic laminate counter top and as you can see the plastic laminate is facing in.  That’s because I wanted a surface with some tooth so when the top is clamped in place it would not slip. 

Next is to clamp the top in place to see how the whole assembly is going to work.  Here is how it looks when ready to make a horizontal cut.  No problem here, the top is held securely in place with four clamps and is very stable.

The vertical clamp did give me a little problem as only three clamps can be used and I really want to use four – one at each corner.  If you look at the lower left you can see I ended up making the fourth clamp to hold that corner tight against the tall fence.  The clamp is made from a short length of 3/8” rod, 2 nuts and washers plus a spacer matching the thickness of the top all mounted through a hole drilled in the tall fence.

The photo below shows my first test cut with the blade set at 29 degrees taking only a partial cut.  The final pass will make the bevel much wider.  Everything worked well in making the cut, the sled ran smooth while the tall fence held the top solidly in place.  You can also see a closer view of the fourth clamp described above.

Next is to remove the top, rotate it 90 degrees then clamp in place.  In this photo you can see the second test cut down the long edge of the top.  This cut came out better than the first.  I would like to say it’s because I learned from the first one but in reality, it’s mostly because the first cut was across the grain and this one is with the grain.

A closer look at the corner where the two cuts come together gives a preview of what the finished geometry will look like.  On the vertical end the pencil line shows how much more I have to take off.  At this point I am happy with the results and can proceed with making the finish beveled cuts.

Because I did not know for sure if any problems were going to show up in the test bevel cuts the top blanks were left long just in case, I needed to start over in making the cuts.  With the successful tests the blanks can now be cut to their final length using the panel jig.  As the back of the top is a square cut the final width will be cut after the bevels.  Just in case something goes wrong.

Final bevel is cut in two passes, first one takes off almost all of the material but can leave some burn marks as in the top photo.  The bottom photo shows how making a second pass taking off less than the saw blade’s width gives a much cleaner surface. 

That does not mean the bevel cuts are of a finished quality.  I still need to sand them starting with 60 grit and finishing up with 220 on the long edge.  The crosscut ends are taken to 400 grit.  That’s because at the same smoothness end grain takes up more stain resulting in a darker finish.  Going a step or two finer results in a surface that absorbs less stain making the end grain blend in better with the long grain. 

Because I want to keep the edges of the bevels crisp and the bevel itself flat a pad sander is not such a good choice as the pad can flex rounding things out a bit.  In cases like this a piece of MDF or smooth plywood is used faced with hook material which the loop backed sand paper is mounted to it.  That way I end up with the paper being backed by a hard material with very little give.  Here is what I mean: on the left is the 220-grit loop backed sandpaper, the disk it goes on is next with the hook backing and the next two disks have the sandpaper applied.

With all the bevels cut and sanded I can rip the tops to their final width.  This photo shows one set in place on top of the cabinet.

Mounting the top to the sub-base is next.  The attachment is not as simple as either gluing the top down or running some screws up from inside the cabinet through the sub-base into the top.  That’s because the sub base is ¾” plywood and the top is solid 1¼” oak.  When exposed to different humidity levels the plywood is stable but the solid oak will expand or contract across the grain in response to the humidity change.  If the top is solidly attached to the plywood the forces can be enough for it to crack and that’s not a good plan.  To get around this the top is screwed to the plywood with 8 screws around the perimeter.  The two at the center of the ends are run up through a small hole in the plywood to hold the top in place.  To allow for the top to expand and contract about its center line the holes down the long edges are really slots with the screws just pulled down snug.  This leaves the top free to expand or contract and not crack as the humidity changes. 

The slots require the most work.  Once laid out the center point is marked with a punch then a hole is drilled at each end and one between them.  Here you can see the alignment hole in the center, punch marks for the slot at the lower right of the photo, and the three slot holes drilled at the upper left.

After the holes are drilled the interior edges are roughed out using a Dremel tool with a burr then cleaned up with a flat and round riffler.  This is what the slots and the center hole look like completed.

Next up – The Handles