On the way back to the shop to resume Mr. D's wall unit project, I stopped by the plywood place for some sheets.

For the last 15 years, it's been my practice to stand up plywood, as is similarly done for the lumber - and for the same reasons. When plywood is created, it has the same moisture content as permitted by the ovens in the plywood mill.

I like to think they know what the moisture content should be for my shop, but experience has shown it isn't so. And since it hasn't done any equalizing laying face-to-face in a huge stack in a warehouse, letting it air out for a few days is good insurance I won't have warpage problems later.

Sure hope I don't need the jointer any time soon...:)

The main task for the day is routing the raise in the panels. This is done on the router table using a horizontal cutter and my new router (more later).

The depth of the profile can really be anything. Some woodworkers like to leave the backs of their panels flat, so they mill their panels skinny or leave the drop-off around the field pretty steep so they can use thick panels. I prefer to keep my panels full thickness and in plane with the frames, especially since some of them are going to be drawer faces, and I don't like the looks of having my panels stick way out. To each his own.

To keep the 3/4" panels in the same plane with the frames, just align the front of the tongue with the front of the groove (photo right). After the panels are raised, a straight bit is exchanged in the router table and the backs of the panels are rabbeted to leave a 1/4" tongue.

Because we have a new router - the biggest, baddest, hand-held router there is (more later), we can tackle cutting the raise in one pass. I like it!

Of course, in a perfect world, with perfectly flat panels and a perfectly router table top, all our profiles will be routed to exactly the same depth. In my case, some of the corners exhibited a little offset in the profile. Even if you can't see them easily, they're sure easy to feel with your thumb.

Since we want our profiles perfect, even if the world isn't perfect, any high corners are re-routed with a bit more downward pressure applied in the proper places

It's always been my practice to invest a little money into the shop at the beginning of each large project. The last time was a new exhaust fan. I do this 'cause at the end of a project, I may not need the tool for a while, or I may not have the money...:( 

As I mentioned earlier, I got a new router - the big Porter Cable. It's called a Speedmatic and it has five speeds so you can dial it down for turning those large panel-raising bits.

The router it replaced in my router table is also called a Speedmatic, it's the little brother of the big PC and it's about 15 years old. Back then, variable speed was relatively unknown and it always gave me the creeps running large bits, in the little PC, at 21,000 RPMs.

Since the big PC is a whole lot bigger than the little PC, the dust shroud for the router table wouldn't fit - I needed to make a new one. I really didn't want to stop and build a new shroud and besides, I only had a few panels - how much dust could they make?

Judging from the picture at right, you'll see a new shroud next time!

Now that the raises are profiled, it's time to back-cut the edges to form the tongue. this is done with a straight bit mounted in the same router table with the speed adjusted up to 21,000 RPMs instead of the 13,000 RPMs used for the raises.

For all the panels, I back-cut the rabbets to leave about 3/32" of wiggle room all the way around. This lets the panels in the doors and drawer faces expand 3/16" before doing any damage to the frames - more than enough for an 8" wide walnut panel in mid-season (between dry winter and humid summer). Denser woods like oak or hickory might need a little more.

If this were the middle of February, I would probably double this space to about 3/16" all the way around. If this were the depths of summer, where panels can only shrink with humidity changes, I could get by with hardly any clearance.

The large panel (photo right) over the fireplace is a different story. At 30" wide, it needs proportionately more space to expand. I moved the fence back, on the rotuer table, more than an 1/8" for the two long edges, giving them a wider rabbet and more room to grow.

Without doubt, the trickiest part of this design is handling expansion and contraction around the fireplace. If we were using plywood for these areas, it would be no problem, but part A is directly above the fireplace and, at 8" wide, it will definitely heat up and want to shrink.

To allow this shrinkage, the fireplace header (part A) will mostly float in the long stile (part B) except for the bottom couple inches which will be glued tightly. I'm also going to throw in a couple biscuits between parts A and C to further lock in this intersection. The gap at the top of the header will be covered by the back edge of the thick mantle (not shown...er...not even made yet...:)

Further, the groove in part C is splined to the groove in part B. It's what's holding it in place now.

From here, all the parts are labeled for re-assembly with glue, and the raises are sanded for pre-finishing (more later).

While working on the pieces, I was a bit disappointed to find a check had opened up in the large panel. I think this was there all along only I didn't see it. Normally when this happens, I ram some glued wedges in the crack and call it good. This crack was too narrow to get any wood to insert so I just worked in some glue. To open the crack as wide as possible, I cantelivered part of the panel over the edge of the bench and pushed down. I know I have enough glue in the crack when I can let up on the overhanging edge and have the glue squeeze back out of the crack.

I'll have to watch this spot to see if it stays tight. If not, I'll have to cut in another piece or inlay a dovetail key in the back side...groan.

The final task for this workday is pre-finishing the panel raises. They are hard to get to, after the frames are assembled, and should they shrink badly in winter, you won't see bare wood around the perimeter.

I'm only putting on one coat and that should be enough for now. Mr. D's walnut floor is finished with water-based polyurethane so that's the white stuff in the old peanut butter jar.

Let's call it a day, OK? I'm pooped!

  The show will go back online Monday. If you have any questions or comments about Mr. D's Walnut Wall Unit, please post them at the Info Exchange.

Jim Mattson