Replies

1. 
   UncleDunc | Aug 30, 2002 02:56am | #1

   I think I would try to adjust the depth first. Run two pieces, turn one of them over, fit them together, adjust until the faces are flush. Then adjust the fence. Theoretically, the same fence setting should work for both the horizontal and vertical pieces, but it wouldn't shock me to find out that a little different setting for each gave the best results.

   In theory, theory and practice are the same. In practice, they're not.

2. 
   John_Casteen | Aug 30, 2002 05:34am | #2

   Gene,

   The only trick I could suggest is to make life easy by leaving about 1/64" of thickness unshaped when you set up for either the face or edge cut.  That way you can leave your fence faces in a single plane for the edge pass, which eliminates one possible source of discrepancy between the two cuts.  The small flat that's left (a little inside corner, actually) when the joint is assembled can be sanded or planed out.

   Aside from that, you're doing it right.  Height first, then depth... I think someone said that already.

   John Casteen

   http://www.fernhillfurniture.com

1. 
   Boxduh | Aug 30, 2002 07:04am | #3

   The geometry of the joint is a 45 miter, with a tongue/groove zigzag along the line.  The reason it is so tricky is that the zigzag has a centerpoint that needs to fall dead center along the miter, and the miter length is dependent on the thickness of the stock.  Thus the need to be working with scrap during the setup that is PRECISELY the thickness of the finish stock. 

   The question is, how do you know when your bit height fine tuning has done the job, and it is time to work on the precision placement of the fence?

1. 
   UncleDunc | Aug 30, 2002 07:07am | #4

   Is the method described in message #2 inadequate somehow?

1. 
   Boxduh | Aug 30, 2002 07:58am | #5

   I am taking a break from the shop to study this problem using CAD.  After having done some modeling, I cannot see any other way to do this than to attempt first to get a bit exposure so that the diag is exactly equal to the hypotenuse of the stock thickness, i.e., either cut, the one with the stock flat to the table with edge to fence (cut 1), or flat to the fence with edge to table (cut 2), results in a perfect no-loss 45 knife edge.  Put 'em together.  If the edge of #2 overruns the joint, the tongue/groove detail is too far "up."  Lower the bit a few thou, and move the fence enough away so the bit exposure will do those knife edges again, and cut two more test pieces.  See what I mean?  I do not think one can independently move the bit up or down without some corresponding movement of the fence.  I think when I get this booger dialed in I will want to make the cuts and archive the whole jig, table, router, and bit.  A friend of mine does this, buys routers at auctions, estate sales, etc., and leaves them in the jigs he has made for cuts he will do repetitively, when setups are tough like this one.  I just checked my test stock and it is .810 thick as compared to the cherry stock at .825.  No goodski.  I will mill some test stock as close to .825 as I can get and start over.

1. 
   UncleDunc | Aug 30, 2002 10:15am | #6

   To set the bit depth, I need to be able to move the stock past the bit in a straight line, but I don't necessarily need the fence to do it. Assume that I take the fence completely off the table use some other method, such as a miter gauge, to move my test pieces in a straight line. I run two test pieces through the router, both lying flat on the table. I feed both of them far enough into the bit that the whole hypotenuse is machined, all the way out to the thin edge. Then I turn one of them over, both of them still flat on the table, and try to fit them together. If the bit is the right height, they will fit together with the surfaces flush on both sides. If the bit is too low, the thin edge on both pieces will stick up past the other piece. If the bit is too high, the thin edge will be below the other piece on both sides.

   I guess what I'm trying to say is the joint cut with this router bit doesn't have to be a miter joint. The shape of the machined surface is always the same, but if you cut one piece on the edge and one piece on the face, it's a miter joint, and if you cut both pieces on the edge, it's like a real short, keyed, scarf joint. It's easier to cut the test pieces and fit them together as a scarf joint. You get exactly the same information about the depth of the bit as you would testing it as a miter joint. And you only have to deal with one variable at a time.

   If the axis of the router bit is perpendicular to the table in both dimensions, and stock of the same thickness will be used on both sides of the joint, and the fence is perpendicular to the router table, this method will set the bit at the right height, and I should then be able to adjust the fence without moving the bit again. If any of those conditions doesn't hold, then you're right, I may have to adjust the bit height and the fence distance simultaneously.

   As far as archiving the whole setup, yes, that would save a lot of hassle on repeat jobs, if you can afford all the routers.

2. 
   Boxduh | Aug 30, 2002 02:57pm | #7

   I see!  You are right, and I will change my setup method as soon as I mill my new batch of test stock.  Height first, then fence.

3. 
   nigelUsa | Aug 30, 2002 03:15pm | #8

   All the above explaines why the bit I have is still in the drawer!

3. 
   SawdustSteve | Aug 31, 2002 06:49am | #9

   If you bought the blade from Roeckler or several other companies, they sell a set-up gauge for the bit.  It's about $10 to $15, but it should give you a 99.9% accurate setup the first time.  As soon as I need to get one of these bits, I will invest in the set-up block too.  There MUST be a reason they list it along with the bits.

   SawdustSteve