FLW Cabinet - #5 Bandsawing, a Revision & Work on Doors

With the bandsaw now in better shape than it ever was I was ready to re-saw the 16 side spacers.  First step is to use my marking gauge set at 9/16” or 1/16” over my final finished dimension of ½”.  Here you can just see the pencil mark on the board edge.  The knot is not a problem as it will be cut out before I am finished.

This is my standard re-sawing setup; a pivot point so I can correct the blade path when it starts to drift one way or another and my magnetic featherboard set to apply a little pressure against the piece to hold it tight to the pivot point.

Here is the cut piece right off the saw and I am really happy with the result.  Too bad it only took me 18 years to correct the original problem. 

Once all 16 pieces were re-sawn I ran them through the thickness sander to clean up what little roughness the bandsaw left then set them aside.  These pieces are all flatsawn so I think that they will cup at least a little as they adjust to their new thickness.  The flat stack on the right are the ripped cutoffs I will save for another project.

Next the ¾” thick parts for the doors were cut to rough width and set aside to rest.  While doing this it occurred to me that I had a problem with my original door corner joint.  Because the glass panel in the door is large and fairly heavy, I wanted as much strength at the corner as possible.  To get the maximum glue surface I will use a bridal joint as shown on the left.  The problem with that is to cut the center out I need to use a jig on the table saw which requires the piece to stand on end.  Not a problem with a normal cabinet door but this one is almost 78” tall which is 1” more than the distance between the top of my table saw and the ceiling.  My next version was to go to a regular mortise and tenon joint shown in the middle.  However, with that cut the available glue area is reduced by almost 50% so it’s not a good option.  Last and what I plan on doing is reversing the bridal joint so the piece I need to sand on end to cut the tenon is not the upright but the cross piece.  That piece is only about 31” well within what’s possible to cut. 

The only real difference is the how the stiles and rails intersect.  The left image shows the original plan and the right shows the revised plan.

While making those changes, I also made some minor detail changes to the case construction and now hope the design is pretty much set although from past experience I know that’s not the case as things will change.  Below is a revised cut list based on the updated plans.

Because the existing leaded glass window is the key component and has a fixed size that means the actual construction of the piece starts with fitting the door to the glass then working my way out.  It’s like building a room to fit a piece of furniture instead of finding a piece of furniture to fit in the room.   The drawing of the horizontal rails is below.  The left side show the back with the recess at the bottom for the window while the one on the right is the front side.  The small squares are where wood pegs will go.  More on that later.

This is the drawing for the back of the vertical stiles.  The recess shown on the bottom is where the sides of the window will go.  The tenon on the end gets cut and fitted after the mortise in the rail since its easier to fit the tenon to the mortise than the other way around.  More on that later too.

As the door stiles and rails have been sitting for a few days it’s time to check them out to see if they are still true or if they have moved.  As I left the rough-cut pieces their thickness was roughly 13/16” thick.  Fortunately, they are stable so the work could progress without any gymnastics to get them back into shape.  What was needed was to get rid of the planer scallops and get all the pieces to a constant thickness.  As they sit now the thickness varies a little and to make the corner bridal joints being off a little will make the difference between a great fit and a sloppy one.  The photo on the top shows the before view and you can see the board is far from being smooth.  After running it through the thickness sander loaded with 220 grit sandpaper the bottom photo is the result, nice and smooth.  Checking with a dial caliper the thickness of all the pieces varied by only .0025” or 25 ten-thousandths of an inch, about like a sheet of typing paper.  They are a little over ¾” but in this case that’s not a problem.  Getting them the same exact width is the primary goal.

Spending some time deciding which face to present to the outside and which to go inside the cabinet is next.  As the boards were rough-cut about ¼” wide flaws within that range can be cut off.  About half of the boards made for an easy decision with better looking grain on one face or a flaw that needed to be dealt with.  Here is an example of a crack that I want to get rid of.  It will go inside the cabinet and up next to the glass window.  That edge has a ¼” by 5/16” recess cut for the glass to sit in so between trimming the width and cutting the recess the crack will get removed. 

A trip through the table saw gets all the rails and stiles to their appropriate width plus 1/32”.  Just a little insurance in case I get some chip out while making cuts.  Here are the pieces ready to have a trim cut on the #1 end to square them up followed but a cut to length on the “X” end.

Next up – Door Joints Part 1

FLW Cabinet - #4 Rough Cutting 5/4 & Adjusting the Bandsaw

With the leg blanks set aside next is to rough out the upper and lower rails that tie into the legs plus the top.  The legs are made from 6/4 material while the rails and the top are made from 5/4 material.  That’s because the rails are set in a bit from the front of the legs to provide a shadow line.  The drawings below shows the parts and what I mean.

The 5/4 material is rough-sawn just like the 6/4 so I will need to go through the same process to get flat, smooth and square boards.  Here are the three boards I started with, the good news being that they are not as curved as the 6/4 planks.

While they may not be as curved and not as bowed, they are still not flat.  The process starts with cutting the planks into rough lengths of 36” and 42”.  Here you can see how much bow the have when clamped down so they can be ripped lengthwise. 

Once ripped the apparent bow is greatly reduced which means less material to be removed before I can get flat boards.  The photo below really shows the improvement in apparent flatness.

Now it’s time to set the planer up again and start running the pieces through it.  I start by taking off 1/16” per pass but after a couple passes as the cut gets wider since the crown is being removed the cut is reduced to 1/32” per pass.  Once the first side is fairly flat the piece can be flipped and work on the second side can begin.  I emptied the shop vac before starting and may need to again before all is done.

Several passes later the boards looked pretty good that is until I checked for flatness as four of them had a little twist.  I hate that, of all the problems it is the hardest to remove.  Cup or curve are not too hard to correct but twist is the one that gives me the most fits.  A jointer makes the correction easier but I don’t have one so end up using either a big hand plane or a power plane.  Oh well, using the 18” long, 7 pound hand plane gives me my cardo workout for the day.

Anyway, after flattening the offending boards and squaring up the edges on the table saw I had a nice stack of square, flat smooth material.

Since some of the boards had moved around after being cut and surfaced I thought it’s a good idea to set the just milled pieces off the side for a bit to see if they had stabilized or wanted to move around some more.  Here are all the 6/4 and 5/4 pieces set on slats on the floor with spaces between them so the air can freely circulate.

While waiting to see if the thicker material was stable, I moved on to roughing out material for door frames and spacers on the sides.  The drawing shows these parts.  The door frame material starts out at 13/16” thick and will probably end up between that and ¾”.  It all depends on how much I have to remove to get rid of the surfacing scallops.  The spacers also start out at 13/16” thick but will end up at about ½” thick.  There are a couple of ways to get there but I will go through that later.

Below are the boards used and how the pieces get laid out on them.  There were a few knot holes, cracks and the like to work around but all in all after some time measuring and thinking a pretty efficient layout emerged.  I just hate to waste good material.

After cutting the short spacers that go on the side, I needed to cut them down to ½” thick.  There are a couple of ways to do this.  One is to run them through the planer and turn the ¼” of oak that needs to be removed into sawdust.  The second way and what I am going to do is cut a thin slice off the piece using the bandsaw.  

First, a little back story on the bandsaw.  When I got it in 2000 it had some significant alignment problems that really impacted the quality of the cut.  After spending some time on the phone with the manufacture we got it so the cuts were satisfactory.  There was still a problem with the blade fluttering while running which gave rougher cuts than what I thought it should.  At the time I could not make any finer adjustments so that’s the way I have been using it for the past 18 years.   Now before starting to rip the spacer pieces I wanted to have a go at refining the alignment.   Back in 2000 my only really good straight edge was a 4’ level which was too short to align the wheels.  However, not so long ago I broke down and got an 8’ level.  Not for the leveling function but as a precision straight edge and that’s what I will use on the bandsaw.

What I believe needed checking and adjustment is the alignment of the wheels.  They need to set along the same vertical axis, neither canted or offset.  In other words, both of the wheels need to be in the same plane.  Here I have clamped a couple of spacers exactly the same thickness to the level then held them tight to the outer rim of the top wheel. 

Below the left photo shows a closer view of the top wheel where you can see the spacer is touching both the top and bottom of the wheel.  The right photo shows the bottom of the bottom wheel with almost a ¼” wide gap.  The gap at the top of the bottom wheel was only slightly less than that so I had two problems, the wheels were both canted and offset. 

By adjusting the top and bottom wheels I was able to correct both problems and align all four contact points so they just touched the spacers.  Below is the bottom of the bottom wheel in proper alignment where it should be.

When I restarted the bandsaw the blade fluttering was gone and a test cut proved to be way smoother than anything I had ever gotten before.

Next up – Bandsawing, a Revision & Work on Doors

FLW Cabinet - #3 Making & Using the Taper Jig

To make the jig I keep a few good looking 8’ long 2x4’s on hand for just the purpose of making jigs and such.  The one I am using has drying out for probably a year so if it was going to warp it’s done by now.  Sure enough when I set the long level on it there was a bit of a bow so I used the same straight edge and plane procedure on it that I had used to straighten the leg blanks.  Once that was done a couple passes through the thickness sander gave me nice flat faces and a good reference plane. 

One block at each end of the leg blank is needed to hold it in place while being cut.  They start by taking a short length of 2x4 and cutting one edge to match the taper of the leg.  Rather than adjusting the chop saw to cut a .8 degree angle a small spacer is set between the stop block and the saw’s fence.

The plan is for the stop block to get screwed to the reference 2x4 and to the leg blank.  Since the screws will run into the bottom of the project’s leg, I don’t have to worry about them being visible in the finished project.  The leg blank is longer than needed with extra length at the top so the screw holes there can be cut off.  Here is the marked-up bottom block that shows a whole lot of layout lines.  Some explanation of them is in order.  Because the screws are not long enough to go through the block they will need to be countersunk.  The light solid lines on the face of the block show the counter sunk depth.  The dotted lines on the face show where I need to drill a pilot hole for the threaded part of the screw to clear.    

Now the drilling can commence, the left photo shows where I have drilled a shallow 3/8” counter sink hole and am getting ready to drill the next one.  It will be drilled full depth later but for now its only purpose is to allow me to center the pilot hole bit (right photo) and drill it to the bit’s full depth.  It’s too short to go all the way through but it almost makes it.   The screws that will be used are self-tapping so they will finish it out.  Main reason for drilling the pilot hole here is to help get the screw started straight and square.

Now I can go back and drill the countersink holes to their full depth.  The photo on the left shows green tape that acts as my depth gauge.   The bit being used is a Brad point drill that makes a flat bottom hole.  That’s to match the screw I am using as it has a flat bottom.  In the photo on the right the screw driver is chucked up in the drill press and is being used to help keep the screw square to the block.

Here in the left photo you can see the screws I used and where one is run in until it almost comes through the bottom.  I need to stop short so the piece does not get screwed down to the drill press table.  The right photo shows a couple of things.  First, you can see how the holes are offset from each other.  If they were all centered in the block, they would cross each other and having two screws try and occupy the same plane just doesn’t work.  I know I accidently tried it once.  Second, is that no matter how you try sometimes the drill or screw wanders off course and they pop out the other side not exactly where you planned.  Here on the left side of the block one is right in line and the other is well let’s just say it’s a little bit of a free spirit.

With the two end blocks done just one last piece needed to be made, it’s about an 8” long wedge whose taper to matched the leg taper.  The taper is laid out and the disk sander is used to grind the angle then it is screwed in place.

All the leg blanks need to be the same exact length to work in the jig.  By using a stop block clamped to the work bench and the chop saw I made quick work of that.  Next, one of the blanks is set against the right stop block which has already been screwed to the reference 2x4.  The blank is then clamped tight to the 2x4.  The other end is clamped against the recently installed wedge and the left stop block butted up against the leg blank.  With all the pieces held pretty close to where they will end up a band clamp is used to pull the left block, the blank and the right stop block tight together.  Two clamps are used to make sure the left block is flush with the 2x4.  The photo bottom photo shows how all that comes together except the two clamps holding the blank tight to the 2x4 have been removed for clarity.

Once the stop blocks have been screwed in place the band clamp is removed and the clamps that temporarily held the leg blank in place are put back on as shown below.  You can see the screws that will hold the blank in place ready to run in.  Once that’s done the clamps removed and the first leg is ready to be cut.

This is the setup for making the tapered leg cuts.  Out of frame is a roller stand to support the back end of the jig as the cut is started.  The 8’ level is clamped to the rip fence and a test cut about ½” wide is set up.  The jig with the securely attached blank is then run through the saw.  The test cut is just in case anything went wrong so I could adjust and recut without irreparably damaging a blank.  Everything worked perfect with a nice clean straight taper that is repeatable so I reset the fence to the final width and cut all 8 legs.

Here are the cut legs and for the most part all the problem areas were removed.  I say almost because there are still some pieces I will have to address.  The bottom piece I knew about as that flaw showed up while I was surfacing the planks.  The flaw will get filled and that face will end up going against the wall.  The cracks on the right ends will almost completely get cut off when the legs are cut to length.  In the end I have two problem children.  First, the third piece from the bottom has three cracks that showed up when I ripped the taper.  The one circled in chalk is the biggest as it wraps around and is visible from both sides.  My second problem is two of the blanks decided to twist some after cutting the taper so that could be the biggest pain to resolve.  For now, I am going to set these pieces aside to see if anything else develops or if I am lucky the twist will reverse itself.  Not a high probability of that but sometimes lady luck smiles.

Next up – Rough Cutting 5/4 & Adjusting the Bandsaw

FLW Cabinet - #2 A Straight Edge, Ripping, Surfacing & a Jig

Deciding how wide to make the 6/4 rough-cut leg blanks is a bit of a balancing act.  Wider gives me more material to work with but increases the cup that has to be removed.  Also, after removing the curve from the planks a few problems reared their ugly heads.  Everything from some cracks and splits that were not visible to some knots that showed up when I had a clean edge.  After a lot of measuring and marking out different widths 4” more or less looks like it will give me enough to get the finished pieces while getting rid of most of the problems.

Here the cutting jig is clamped in place to give a roughly 4” finished width piece.  The plank is not wide enough to clamp the jig in place without the clamps interfering with the cut so I have to put on multiple clamps and relocate them as the cut is made.  The top photo shows the initial setup and the bottom one shows the completed cut.

Converting the rough-sawn faces to nice smooth ones is accomplished by the use of a surfacer.  Its knives spin at about 9,000 RPM’s and smooth the surface out by taking off 1/32” per pass.  I could take more but I don’t want to stress the machine.  Also, shaving off a little at a time gradually reveals what the wood looks like and gives me a chance to decide which face needs the most removed to obtain the best-looking result.  Here it shows what a board looks like pretty much done.  By the time I am finished the red shop vacuum will probably need to be emptied.

With nice flat, smooth faces on the boards I could turn my attention to the edges.  Using the ripping jig and my circular saw gave a more or less straight edge.  However, because the boards were cupped and when clamped in place the jig followed the curve the edges are not 90 degrees to the faces.   Before I could rip a good straight and square edge on the table saw I needed to true up the initial cut edge.  The method for that is to use my 8’ long level to find the high and low spots.  That’s done by using a piece of paper to find the gaps between the level and the board then marking the high areas.  Using either a hand or power plane to knock off the high spots allows me to true up the edge to within about the thickness of the piece of paper.

Ripping a good edge on the table saw is next.  I start by clamping the long level to the table saw’s fence.  A long fence will help average out any high or low points giving a straighter edge.   First is to put the edge that was just cleaned up with the planes against the long level then taking a light cut to get my first true square edge.  Second is to flip the board putting the just cut edge against the level then taking another light cut.  If everything works out right, I now have a board with all four faces straight, square and true.  Now it’s time to wait to see if the boards are stable and stay straight or if with all they have been through some internal stresses have been released.  That can cause any one or all of the following problems to show up; twisting, cupping, bending, getting waves or in general not staying straight and true.

For most of the pieces I was able to work around the flaws to get a clear part.  There were a couple that will probably need some help as the project progresses.  Then there was one that no matter how I laid it out it was being a real problem child.  The photo below shows what I mean.  Each of the chalk circles outline some sort of problem; cracks, splits, worm holes, knots or a bark inclusion.  Most of them will be gone or hidden but one knot and the bark inclusion will need to be addressed later.  The horizontal pencil line is roughly the taper to be cut which is next.

Cutting the taper on the legs is normally done using my tapering jig which clamps the work securely in place and has a variety of clamping options that allows most any long taper to be cut easily.  This is what that jig looks like ready to cut a taper. 

Unfortunately, its maximum part length is about 44” which is way to short for the 78 ½” length of these legs.  If it was just one piece, I probably would cut it with either the bandsaw or use the ripping jig and circular saw combination then clean it up with a router run against a straight edge.  However, I need to make 8 identical legs so that’s not a good idea for either efficiency or consistency.   Because there are that many pieces to cut it makes sense for me to invest a little time in a jig to cut them all the same.  Below is my plan for the jig.  It’s pretty simple and should hold the leg in place to make repeatable cutting.  The light tan is the jig, the brown is the finished leg and the white is what gets cut off the leg blank.  The taper will be cut first then cut to length on the chop saw.

Next up – Making & Using the Taper Jig