Router Plane

Monday, October 23, 2017

Frank Lloyd Wright Desk Lamp - #12 Wiring, Assembly & Final Photos

Because I could apply the wipe-on poly finish to all sides of the central column at once I was done with it before I completed all the different faces of the boxes and baffles.  With the column finish completed I could epoxy in the threaded pipe that the light fixtures will get attached to.


After letting the epoxy cure overnight, I took the light brackets off to make pulling the wire easier.  I started by tying a 3/8” X #4 screw to a pull cord and stuffing it down through the epoxied pipe.  To pull the cord I attached a magnet to the end of a dowel then fed that up the chase in the column until the magnet grabbed the screw.  Gently pulling the dowel out gave me my pull cord.  The photo below shows the magnet taped to the dowel, the screw and the pull cord running into the column.


Here the pull cord is taped to the electrical wire that will power the lights.  If the wire looks small that’s because it is.  Since I will be using a 6.5-watt LED bulb which only draws .05 Amps a smaller wire works just fine.  In addition, I can get a light output that matches an old 60-watt incandescent bulb without the heat. 


I digress, sorry back to pulling the wire.  With a gentle pull on the cord while pushing the wire I had no problem getting the wires pulled through the chase and out the bottom of the column.


With the wire in place it’s a simple matter to put the fixture brackets back on or at least I thought it was going to be.  The first bracket I put on did not line up the way it should have.  What I found out was that the threads starting points in the brackets vary so in order to get them lined up square I had to find the one that was used to set the threaded pipe.  Not hard it just took a bit of time before I realized what was going on and then more time playing musical chairs with the brackets.   Once the brackets were matched up with their mating pipes it was a simple matter to make the electrical connections to the fixtures and screw them to the brackets. 

The finishing touch to the fixtures is a decorative cover that I had to cut to length.  The covers are fairly flexible plastic which makes getting a nice square smooth cut a problem.  As far as I could tell just about everything I have has some sort of problem making the cut.  The solution ended up fairly easy although a little odd.  I turned a piece of cedar on the lath to slightly tapered cylinder matching the interior diameter of the plastic cover.  The cover is slid cover over tapered cylinder until it fits snugly and a skew chisel is used to make the cut.  This gives me a smooth square cut which fit the fixture perfectly.


Next is to glue and screw the spacers to the boxes.  To make sure everything went together well I test fitted each spacer/box connection.  Getting the screws in was harder than I thought.  Since there is such a narrow space between the sides of the box I had to use a right-angle attachment on the drill with a very short driver to even get in there.  





In fact, to make the last half a turn on the screw I ended up using a ¼” drive socket with the driver in it.  The photo below shows what I mean.

In the last post, I talked about a slight relief cut on the underside of the spacer because of the little mound created when a screw enters a piece of wood which can keep the piece from screwing down tight.  In the photo below you can see an example of what I mean.


Once the test fit of the spacer to the box was done I disassembled it, added some glue and screwed the two pieces together.  Nice thing about using the screws is that no clamps are needed to hold the pieces in place while the glue cures.  At last all the pieces that go into the box assembly are done and put together.  The last step is to glue them to the column.  This is what the assembly looks like glued and clamped in place. 

There is a bit of setup that goes into locating the box assembly.  You may have noticed the spacer is not flush with the face of the column but sits back about 1/16” from the edge.  You can see this in the left-hand drawing below.  So that I could get a nice consistent setback of the spacer from the face of the column I made a little jig by carefully milling a couple of pieces the same thickness as the setback and gluing them to a plywood base.  You can see the jig cradling the column with the spacer setting on it.  The piece of green tape on the column acts as my locating mark for the top of the spacer.  One down and two to go.


Once all the boxes are glued to the column I added the baffles.  They slide into a slot that runs around three sides of the column and are glued in place.  It’s a pretty snug fit and no clamps are required.  After the glue cures holding the baffles in place I glued and screwed the column to the base completing all the wood working for the piece.  That left just the wiring to finish.  Below is a photo once I had soldered the wires together, encased them in heat shrink tubing and added a clamp to act as strain relief.


Here is the finished lamp and the prototype setting on the workbench along with the design drawing.  Pretty similar except for the base, which although revised in the drawing and final product was not rebuilt from the first prototype version.

This is the last pair of photos for this project.  The one on the left is the lamp alone and the one on the right is the lamp at my desk with the lights on.  I am happy with the final product in it does just what I wanted.  It provides good light on the desk without any visible bulbs, glare off the computer screen and does not put out a lot of heat.



I do have to say that there is a lot more to this than first meets the eye.  It’s simple looking but sure was not simple to build.  That's due to having all the joints exposed with no trim or anything similar.  Any error however slight would really show up.

Monday, October 16, 2017

Frank Lloyd Wright Desk Lamp - #11 Dye Testing, Box Pre-Finishing & Sub-assembly, Finishing

Now it’s back to the last (I hope) dye test piece.  With the last coat of finish on the sample I took it into the room where the lamp will sit to see if the color went well with everything in the room and it looked good.

With that decision made I mixed up ¼ of a cup of the dye and applied it to the spacers and all the surfaces of the boxes except for the outside face plus the top and bottom edges.  
After the dye dried I gave the inside face of the boxes 2 coats of finish.  That should be enough to seal the wood and allow for easy glue cleanup.

Next is to start assembling the boxes.  To make sure everything was going to go together correctly I did a dry run with the biscuits and clamps.  Everything went together just fine so I applied the glue and clamped everything in place.  To make sure the parts were going together square I set the clamped assembly on a flat surface and made any slight adjustments to get all the pieces aligned in the same plane.  The prefinished surfaces worked out great all I had to do was use a damp rag and the glue came right off the finished surface.  Here are all three boxes glued and clamped.


After the glue had set up as I was removing the clamps I noticed that one was hanging down a bit below the edge of the box.  Normally this would not be a problem but in this case the clamp was at the joint where I was checking to see if the pieces lined up.  That meant that when the piece laid flat on my reference surface the clamp was holding the one corner up a little.  End result is that this assembly has a twist to it.  In the photo below you can see how the left corner is not setting flat.  It’s not twisted a huge amount but enough that I was unhappy.





Fortunately, when I looked closely at the box it ended up not being as bad as I thought.  The corner at the other end of the board that looked twisted was setting a little high compared to the end piece.  This meant that by using a full sanding sheet on granite tile to flatten the joint I got rid of most of the twist.  This is that setup.

With that out of the way I finish sanded the outside of the boxes and applied some tape to the areas that will get glued to keep the finish off.  It’s easy to do, I just marked the tape length on a scrap, then cut the end square with a small square and knife.


There are three locations that get tape:  the inside of the boxes where the spacers go, that I had just done. the corresponding place on the spacers and last where the spacers get glued to the central column.  Below are the last two.

I can now get on with dying the rest of the box faces and the baffles to complete that work.  Once the dye has dried I can start with the finishing in earnest.  It’s going to take a while because I will be using a wipe-on poly that I have to let dry overnight between coats.  Well, really it only takes about 3 or 4 hours to dry here but then it’s the afternoon and that’s when I do other things, like work on this blog.  Besides, I am retired and don’t have to rush to make a deadline!

After the dye had dried I notices some scattered places where the dye had taken either blotchy or just did not look right.  You can just see it in the lower right of the baffle.

Looking closer at those areas I noticed it looked like the dye had soaked into the maple then later oozed back out and left tiny rings of concentrated dye on the surface.  

When I felt the surface the blotchy areas also felt rougher.  I tried using just some alcohol to remove the excess dye and blend it but all that did was to lighten the color, soak in then ooze out again which left me with two problems a light area that was also splotched.  After lightly applying the full strength dye I was back at the beginning. 

Trying a different approach, I took some 400-grit sandpaper and very lightly sanded the affected areas feathering it out into the surrounding.  That worked and I got a uniform unblotched surface ready to apply the wipe-on polyurethane finish.

As I have said on this project there is nothing on this project to hide any errors or gaps in the joints.  So far, all the joints have come out good and tight with few exceptions.  Inside the boxes there are a couple miter joints that did not exactly match up.  Yes, they are on the inside where you would really have to work to even see them but I know they would be there.  I tried some different ways to fill the gaps and here is a sample piece with two methods.  They both start with the sample piece stained and with one coat of the wipe-on ploy finish.  The dark line is where I mixed fine sanding dust with stain then after it dried mixed it with the poly finish and packed it in a thin saw kerf.  The method I used is in a similar saw kerf that runs parallel to the first line.  The pencil it pointed right where I made that patch.  For it I mixed the fine sanding dust with the poly finish to a thick paste and packed it in the cut.  I am really happy with the match.

Here are all the pieces except the base with the wipe-on poly finish in progress.  From here I would apply a coat of finish, let it dry overnight, sand lightly with 400 grit sandpaper then repeat.  Once I had three coats on I wet sanded the last coat with a 3,200-grit pad just to get the dust nibs off.

Last bit of work on the spacers is to set up a couple of stop blocks on the drill press so I could add the pilot holes and countersink for the screws that will attach the spacers to the boxes.  Here is the pilot hole setup.
This is for the countersink, same setup except for the bit.

Once the holes had been drilled and countersunk I put the bit in a hand drill to cut a slight relief cut on the underside.  That is to allow a little relief when I drive the screw in.  As a screw exits the back side of a piece and enters the face of the next one it often creates a little mound that will keep the piece from screwing down tight.  I could have drilled a larger pilot hole but did not want any clearance around the screw.  That’s because the screw needs to hold the spacer tight to the box in two directions.  Below on the left shows the little relief cut while the red arrows in the right drawing shows the two directions the piece needs to be constrained to.



Next Up – Wiring, Assembly & Final Photos

Monday, October 9, 2017

Frank Lloyd Wright Desk Lamp - #10 Dye Testing, Sanding, Completing the Base & Central Column plus Fixture Preparation

Now it is time to start nailing the finish down.  I took some scraps, sanded them to the final 320 grit then mixed up a set of 5 samples using alcohol as a carrier for the dye, applied it and after they had dried put on three coats of wipe-on satin poly.  The good news is that I am mixing up the samples using only a tablespoon of alcohol and a few drops of dye so there is not much waste.

While the finish test pieces dried I started doing the final sanding on the baffles.  I began by working on the end grain.  The blade I use to make cross cuts is a high quality one specifically designed for this type of work and it does a really good job.  However, in this case I wanted the end grain to look every bit as good as the faces.  That means sanding it through the grits to 400.  This is one grit finer than the faces and that’s because when applying dye to end grain it takes darker than face gain.  Sanding to a finer grit will lessen the difference.  Here is a photo of the left and right sides of the notched baffle end.   On the left is the saw cut edge.  It’s good, no blade marks or burning but it is not up the smoothness needed.  The right shows the quality of finish I want.


Now that the glue has cured holding the plywood top to the base I can take the clamps off and start to bring the base flush with the plywood.  In post 5 I cut the rabbit a little deeper than the thickness of the plywood.  That was so I could use a scraper to bring the top of base piece down to match the plywood exactly.   Since plywood varies in thickness that gives me a little buffer to work with.  I can always shave a little off the top of the perimeter pieces but because the veneer on the plywood is only about .025” thick there little to no margin of error there. 

A scraper is simply a hardened piece of steel.  The one shown is about 2¼” X 6” and only about 1/32” thick.  It cuts with a razor-sharp burr on the edge so small you can barely see it.  With a good burr, it will take off shavings so thin you can see through them which gives great control in shaping a surface.

The photo below shows the three stages of bringing the solid wood base perimeter pieces flush. 

  1. At the start make a squiggly pencil line across the plywood and the solid wood perimeter piece.
  2. Using the scraper start cutting away the outer piece which removes the pencil line from the solid wood piece.
  3. When the pencil line on the plywood is just removed the joint is flush.

Last bit of work on the base is to finish sand to 320 and “break” the sharp corners.  This is done by making a few passes of 220 grit sandpaper held at a 45-degree angle to put a tiny chamfer on the edge.

Now I had the initial dye samples finished and in evaluating them two were pretty good and the rest were OK except for one, it just did not work.  Looking at the couple I liked I mixed up one more sample and stained a scrap piece.  I will go ahead and start giving it the three coats of finish then take it in the room I will use to see how it looks.

While I was going through the finishing process I went back and worked on the actual light fixtures.  They are simple candelabra bases that get screwed onto a 1” long section of threaded pipe.  The pipe will get glued in place with an epoxy glue.  Below is a cutaway section showing the components.


The pipe comes in longer sections that get cut to length with a hack saw.  The ends are filed smooth and a bevel is added with a grinder on the outside ends to get rid of sharp edges.  Last, I mount the pipe on my lathe, wrap some sand paper around a small round file and smooth the inside edges.  I don’t want any sharp edges that could abrade the wire and cause a short.


I also glued the top onto the central column.  It is pretty straightforward glue the tenon into the central column making sure it’s aligned right then glue and clamp the cap on.  There are a few process notes though.  First is to apply the glue to the hole the tenon goes in.  That way when you put the tenon in the excess glue is forced further in the joint.  If I put the glue on the tenon the excess glue would end up squeezing out the joint where the cap and the column meet making a mess to clean up.  Second, is to use a pair of clamps to make sure the cap is on square and not cocked due to uneven pressure applied by a single clamp. 

Last, I added a backer board clamped to both the column and cap.  This made sure the faces of the cap and the column were in the same plane. 

I have been giving the assembly sequence quite a bit of thought and as of now this is my plan:
  1. Dye the central column.
  2. Dye all the exposed surfaces of the spacers.
  3. Dye all the surfaces of the boxes except for the outside face plus the top and bottom edges.
  4. Finish just the inside face of the boxes.
  5. Assemble the three sides of the boxes.
  6. Sand the outside faces and edges of the boxes.
  7. Dye the remaining box face and edges.
  8. Mask off the areas of the central column, spacers and boxes that will get glued.
  9. Finish the central column, spacers and boxes with 3 coats of satin wipe-on poly.
  10. Epoxy in the threaded pipe for the light fixtures.
  11. Wire fixtures.
  12. Glue spacers to boxes.
  13. Glue the spacer/box assembly to the central column.
  14. Glue the baffles in place.
  15. Screw the column assembly to the base.
  16. Complete wiring.
I think this will work.  By finishing the central column and the spacer/box assembly when I glue them together any excess glue will be on an already finished surface.  Clean-up will be easy as the water-soluble glue should just wipe off.   Also, because the pieces are already finished the water will not raise the grain nor will the excess glue leave a film that would inhibit the absorption of the dye.

At least that’s the plan, for now.  Time will tell if it works.


Next Up – Dye Testing, Pre-Finishing, Box Sub-assembly, Finishing

Monday, October 2, 2017

Frank Lloyd Wright Desk Lamp - #9 Baffle Fitting, The Boxes Continued & Gluing in the Plywood Top

I had planned to continue working on the boxes but am still wrestling with the assembly order.  While I think about that I will finish fitting the baffles.

The start is to size the thickness of the baffles so they fit in the dados cut on three sides of the column.  That’s done with the thickness sander shaving off a little at a time until they fit.  That was the easy part. 

Cutting a notch in the baffle so it fits the three-sided notch in the column is a little more fun.  After some careful measuring, I had the notch centered in the baffle.  To cut I put together a simple jig. The backer will minimize any split out as the blade exits the back side of the baffle and a stop locates the cut.  After the first pass through the saw I flip the piece 180 degrees and made the second cut.  A little work with the coping saw and I had the notch close.  Another set of passes and it was just a tiny bit too small.  Adding a playing card spacer did the trick. 
 
The reason I did not go through the same process of ripping and gluing as I did with the box pieces is because these will fit in a dado and the end cut will be hidden. The top photo shows the baffle almost all the way in and the bottom shows it fully inserted.  You can see how the dado hides the rough edge cut with the coping saw.



Here the three baffles are fitted in place in the central column.


That leaves the boxes to finish up.  I have been procrastinating some because I did not have a good plan on the final assembly sequence.  There are quite a few constraints to take into consideration.  Alignment, time and keeping glue off finished surfaces are the biggest items.  Here are my thoughts so far and the problems I see with each.
  1.  My first thought was to glue the spacers on then assemble the boxes and lastly glue them to the spacers.  The problem there is the fit between the spacers and the horned parts that go on them are really snug plus and as I slide the box onto the spacer I will get glue all over the exposed part of the horns or smeared onto the exposed faces of the spacers.  Also, if the spacers are at all out of square or don’t line up just right then the box won’t slide on. 
  2. Another plan would be to glue everything together at the same time.  Problem there is getting all the pieces assembled, aligned where they go and clamped in place within the 5-minute working time of the glue. 
  3. I could assemble the boxes then add the spacers and finally glue the whole assembly to the central column.  With this method, I am once again concerned about glue getting smeared where I don’t want it and proper alignment. 
  4. Splitting the process into four stages.  First, spacer to box horn pieces.  Second, spacer/horn to end.  Third, spacer/horn/end to column.  Fourth, the remaining spacer/horn to column and previously glued on spacer/horn/end.   The time and glue problems are removed but while the alignment problem is reduced it is still there. 
For now, I will work on cutting the miters and all the biscuit slots.   I left the box parts with only the cutting of the miters and biscuit slots yet to be done.  However, before I start that process take a look at the photo below.  It is the one of the box side pieces (labeled B1) and the end section (labeled B2).  The two pieces meet right below the end of the pencil.  Note how the grain flows from one piece to another.  That’s the result of all the careful tracking I did of the pieces as they went through the machining process.  When the box is finished those grain lines will flow right around the corner.
 
To cut all the similar pieces the same size I used a simple jig with a stop.  I will make a couple of cuts, one rough and a final skim cut set by a spacer.  This shows the final skim cut with the cardboard spacer.

Next is to cut the slots for the biscuits.  They are cut using a biscuit cutter.  I have screwed the tool to a piece of plywood and clamped it to my workbench.  Below is the business end of the tool with the small saw blade extended to show what it looks like.  It cuts a semi-circle in the face of the joint for the biscuit goes into.  To use, I clamp the one of the side pieces to the rectangular piece of plywood that is screwed to the reference face set at 45 degrees.  The edge to be cut is against the face with the blade now retracted, the tool turned on and the blade pushed into the piece cutting the slot.


Here is what the pieces look like after using the biscuit cutter along with the biscuit that goes into the cut slot.  Nice thing about these is that once glued in place they are invisible.

A test and clamping fit is done with each of the boxes to make sure the corners on the miters close, there are no gaps and all the pieces go together square.  Fortunately, there were no problems everything fit together.

That finishes up all the major woodworking.  I still have final sanding and softening of the sharp corners to do.  Here are all the non-electrical parts and pieces that go into the lamp.

First part of assembly is to glue the plywood top to the base.  When I cut the rabbit in the base for the plywood to sit down into I made the vertical cut a little deep to create a pocket.  That pocket will act as a catch for any excess glue to keep it from oozing out at the top when glue it in.


When gluing the top on I do not put any glue on the interior structure.  This allows for it to move independently from the plywood top due to expansion and contraction caused by humidity changes.  I just run a small bead of glue around the perimeter of the base, set the top in and clamp it down.  To get even pressure on the joint I use cauls and a bunch of clamps.



Next Up – Dye Testing, Sanding, Completing the Base & Central Column plus Fixture Preparation