Dovetailing large casework with an Incra jig

[John Schroeder]

So I have an Incra LS positioner, which is a very nice router fence. What I didn't factor in prior to purchase was the difficulty of dovetailing large upright pieces. Since I badly wanted to make a bookcase for our daughter with a dovetailed top, and I am too much of a newbie neander to risk cutting the dovetails by hand in birdseye maple, I set out to see if I could do it on the Incra router table.

Just to be clear: given the choice, I can't say I'd recommend the Incra as your preferred tool to make dovetailed casework. And although I think there is an Akida jig (and definitely an LN dovetail saw) in my future, the Incra is what I had to work with right now. Necessity is certainly the mother of invention, and it worked well enough that I though I'd share the relatively minor design changes I made to make it work for me. There are also a couple of other threads that touch on routing large upright pieces on an Incra (and the desire to see such a thing) so pics are attached!

The biggest issue was the right angle jig. I found the design to be subpar, and with an idea or two from some creative australians over at woodworkforums, I made a few improvements:

• Replaced the plywood runners with longer UHMW ones. They provide more stability and greatly improve the sliding action
• Added a jig handle (rockler, about $10 bucks) for better control
• added an adjustable UHMW piece with two bolts that slips onto the fence above the runners to prevent tipping forward and only allow forward/backward movement
• added a larger phenolic face plate to provide a bigger area to clamp to and more rigidity. It can be removed for smaller work.

So, after a bit of a delay getting all the stock ready to go, I finally put my new and improved right angle jig to use. I must admit I was quite nervous as the two side panels were 48" long, and the 36" top is a beautiful piece of birdseye maple. Funny enough, it worked! And it worked surprisingly well at that. The sliding action was smooth and precise, and the pieces were quite stable and machined well. The two end pieces were done together, and when clamped upright with a backing board produced a wide enough surface that was very stable.

The thing to watch is the fence itself will flex if you aren't careful. It's aluminum and lacks the rigidity and precision I'd like to have. The wider the piece, the more it can act as a lever and flex the fence if you aren't careful to advance it evenly and smoothly. That said, if everything is set up properly I can attest to the fact that you can dovetail a 48" high board with precision.

One thing I didn't take into account was the height of the backing board. As you can see from the pictures, unless you use a 12" high backing board, the height of the larger phenolic plate is largely wasted. I think I may buy some cheap flat pine and make a couple of 24" high by 12" wide boards specifically for backing large case work. I'll just cross cut the dovetailed end off after each use, so that should give me 12 to 15 or so cuts. I do think the plate provided more stability and a more rigid clamping surface even with the smaller backing board.

FWIW, I also needed to make a sliding dovetail for the bottom shelf, and tried a couple of different approaches to make the male ends. I ended up using two rockler right angle plastic clamp guides to stabilize a 36" shelf that I dovetailed the ends on. After extensive testing on scraps, I got it to the precise width and a tight but smooth fit. Boy did that work well, and I'll be using that joint more in the future. Being able to tune the fence .001" at a time really made the fit precise, and no issues with the glue altering the fit. I'd read a few horror stories, including one by David Charlesworth, so I was prepared for the worst. A couple of taps with a deadblow hammer at the end was all it took to assemble perfectly.

Pics attached. Again, it may not be anyone's first choice, but it worked out well enough that I wouldn't hesitate to do it again in the future.

Cheers!

[Paul Johnstone]

nice.. thanks for sharing that solution. Case came out well.

[ian maybury]

Very nicely done John, and something i've been wondering about. I have a similar system and have been thinking about the possibilities of doing that sort of work. So thank you.

I guess to be fair to Incra it's quite some way beyond what they pitch the system as being capable of. As you say it's the precision positioning of the LS that makes it possible, after that it's down to your fixtures...

ian

[Joe Leigh]

Nice job. I was wondering though, would it be better to put the tails on the side boards to make the dovetail more visible?

[Bill Arnold]

Nice job, John. I'll keep your solution in mind if I have a project that large to do on my LS system.

I've done thousands of half-blind dovetails on my Incra but haven't tried through DTs yet. Did you find the additional trimming an issue?

The largest pieces I've worked with so far have been drawer parts up to 10" wide and 24" long. I had to keep a good grip on the parts but it worked fine. For anything larger, I'll be looking at your solution.

[John Schroeder]

Nice job. I was wondering though, would it be better to put the tails on the side boards to make the dovetail more visible?

I wanted to build a bookshelf that derived its strength from joints rather than nails or glue. Thus the tails keep the sides from pulling out at the top, and the bottom shelf holds them in at the bottom with a sliding dovetail. Since I wasn't sure how finicky the sliding dt would be, I decided to use stopped dados for the others, as I thought it provided more a little more lattitude should I be off a hair here or there.

I'm also toying with the idea of putting molding around the top and bottom, in which case I definitely wanted the angled tails on the top rather than the sides. I like the dovetails as they are, so I may skip the molding, but that was my thought process. I've actually never seen a piece with the tails on the sides, but that may be just my lack of exposure to different options.

[John Schroeder]

Nice job, John. I'll keep your solution in mind if I have a project that large to do on my LS system.

I've done thousands of half-blind dovetails on my Incra but haven't tried through DTs yet. Did you find the additional trimming an issue?

The largest pieces I've worked with so far have been drawer parts up to 10" wide and 24" long. I had to keep a good grip on the parts but it worked fine. For anything larger, I'll be looking at your solution.

Thanks Bill. FYI the additional trimming with a chisel is a snap. As Incra states, it's just a little tiny triangle on each side of a pin. The top of the pin is flat and correct, so I just slap a sharp paring chisel onto the top of the pin and slide it down. The little triangle pops right off.

You do need to spend the time getting the setup absolutely correct though for TDT, as the second pass with the pin board flat on the table has to take off the same adequate amount on each side of the pin. I find that to be the most challenging part - final paring with a chisel sounded like it might be tough but in reality it's one of the easiest parts. The quality of the joint is already determined by that point.

[Bill Arnold]

Thanks Bill. FYI the additional trimming with a chisel is a snap. ...

Thanks for the feedback, John. I'll have to give through DTs a try soon.