The last installment of this series was originally titled Milling the stiles and rails and described prepping the blanks for the panels. Sigh. Sorry about that. I’ve fixed that entry title.
This door would be for the passage between my foyer and formal living room, so I thought the best side should face the foyer. I inspected each blank for the stiles and rails and picked out the best side as the “foyer” side, marking each part with chalk to indicated what part it was, which side was the foyer side, and which edge was “up”.
The blanks were fairly flat. The cores had been jointed and planed and the relatively thin skins were planed to uniform thickness before they were glued on. Nevertheless, I jointed one side of each blank. I also jointed a few pieces of 8/4 poplar about 5” wide and about a foot long. These were intended for making test cuts.
The widths of the middle and bottom rails were nominally 8” and 9” respectively – too wide for my 6” jointer. For these parts I used my planer. I’ve used 2 methods to do this. For one method, I made a torsion box sled 48” long and 12” wide. I fix the work to the top of the sled with a couple of dabs of hot melt glue and run them as a stack through the planer a couple of times taking light passes, jointing the top surface parallel to the flat bottom of the torsion box sled.
The second method which I think I like better, is to attach runners to the edges of the work. The runners are thin-ish strips of scrap that have been jointed on one edge. I put the work on my table saw top and attach each runner, jointed edge down so that the jointed edges of the runners are co-planar. The runners extend 3” beyond the ends of the work so they also eliminate any snipe when planing. I then run the assembly through the planer, taking a couple of light passes until the top is jointed to my satisfaction.
I then proceeded to thickness all the parts to a final thickness of 1-3/4”. As I worked down in thickness from the rough thickness to the final thickness, each part was flipped as needed so that the final thicknesses of the skins on the part were approximately the same thickness. The poplar blanks were milled to this final thickness at the same time.
After thicknessing the parts were edge-jointed on the jointer, and finally ripped to final width on the table saw.
Pretty standard stuff.
I used a carefully calibrated cross cut sled to true up one end of each blank, and then to trim the other end to final length. It was probably silly to do, but I planned these cuts so that even if the sled was not precisely calibrated to 90 degrees, the parts would still come together at tight joints. I marked each corner that was against the fence when the corner was cut, and I planned the cuts so that once the parts were assembled in the door, these corners would be at the lower left and upper right of each part. The door would at least be a parallelogram even if it wasn’t exactly square.
I had decided to use the Amana entry door making router bit set. Having never made a door before I decided the prudent thing to do was to make a small practice door. So I got some 8/4 poplar to use for this. Once I got it home, I remembered that I had parts leftover from the solid wood front door that was originally on my house. I replaced it because it leaked and eventually the bottom joints started to come apart. Nevertheless, I was able to mill up some of the remaining parts to provide all the blanks I needed. I took notes from this process and from that developed the procedure I would follow when milling the parts to the pocket door.
Here is a picture of the practice door:
I intended to follow the instructions provided with the entry door making router bit set, but I had to deal with the fact that the thickness of my panel blanks was going to come up shy of 1/2 the door thickness.
This set does not provide a cutter for the panel slots, and I bought the Timber Line Multi 3-Wing Slot Cutter Router Bit Set to handle whatever slots I was going to need. My panel blanks were going to plane out to 13/16”, a bit less than 1/2 the thickness of the door ( 7/8” ). I considered separating each pair of back-to-back panels with a 1/8” thick spacer panel to compensate for the under-thickness panel blanks. Instead I decided to cut separate slots for each panel. In this case the slots would be 3/16” thick, and there would be a 1/8” strip separating them. I still have the poplar parts I used for test cuts. Here is a picture of one of them that is milled up as a rail and shows the panel slots.
I set up a slot cutter for 3/16” slot width. The bit height was adjusted so that the outside edges of the slots were 1/2” apart. That is, the total thickness of the 2 slots and the spacer between them was 1/2”. With the poplar blanks I tested the setup and made sure that the mortising chisel fit exactly between the outside edges of the slots. On the test door I had gotten a bit of chip out when slotting (probably because I cut the slot in 1 pass) and I wanted to eliminate any chance of that happening. I used a sacrificial fence on my router table fence to provide zero clearance support for the cut. I also made the slots in two passes.
I laid out and then cut all the mortises. I used a mortising attachment on my drill press. I don’t think this would have been possible with a typical benchtop mortiser because I had to cut mortises into the mid and lower rails that were 8” wide (or more). I don’t think they would have fit. Another advantage of using a drill press is that you can use a very low rpm and minimize heating of the bit. I set the belts on my drill press for the lowest rpm.
The next step was to route the coping profile. The (flat) top of the bit is aligned with the outermost edge of the slots. For each cut I clamped a backing board to the work piece. I elevated my fence to be a bit higher than this so that the entire edge of the work was always fully in contact with the fence – there was no gap in the fence around the router bit. I cut this profile in 3 passes, adjusting the position of my fence for each pass. Consequently, I ran all the parts through the first pass, adjusted the fence, and then did the second pass.
The final step was to route the sticking profile. This was also done in 3 passes, but the bit height was changed between passes after aligning the fence with the bearing on the bit. I also elevated my fence on this cut so that it could be as close as possible to the bearing.
I had to do a lot of fine-tuning of each one of the mortises, widening them with a chisel. I deliberately set up the coping bit to leave the tenons just a smidgen thick just to be sure no joint came out loose. I went joint-by-joint, shaving the sides of the mortise until the tenon fit to my satisfaction – not quite so tight that I couldn’t seat the joint without using a mallet.
-- Greg D.