Replies

1. 
   JohnWW | May 10, 2006 01:00am | #1

   Jasmin,

   As you've discovered, your process is generating a cumulative error, and a tiny one at that, of about .003" per step, which is pretty good for the technique you're using.  In fact that is about as good as you can expect indexing each cut off of the previous one.

   The way to do this is to set up a long fence on your miter gauge. The fence should be a bit longer than the stock you will be notching.  Next put a sturdy stop block on the fence at the appropriate distance from the blade for the first notch and cut all of the notches in all of the pieces at this setting.  Now move the block to cut the second notch and cut all of those, and so on for all of the cuts.  By proceeding in this manner, the notches will be in line with each other because they are always being measured from the end of the stock rather than from a previous notch.  Be sure that after making the notches you don't get some of the strips turned end for end before you start to assemble them into the steps.

   For production work, like your flight of stairs, you can group all of the eight or so strips for one step into a rectangular bundle and notch them as a unit.  A good way to make up a group is to make the first notch in each piece, using the technique described above, and then tie the group together with a snug fitting cross strip in the first notch, the cross strip can be held in place with a screw into the first and last pieces of the group if necessary.

   John White 

1. 
   jasmin | May 10, 2006 06:51am | #2

   Thanks, JWW

   I appreciate your input. Limits my error to one increment of my starting point discrepancy between pin width and dadoe stack sizing. I can then adjust the size of my perimeter field for that one increment of error in both tread & landing length & width.

   To pass along to anyone else interested in a similar project, my solid/space/solid pattern in the lattice is derived from the root of the 3/4" stacked dadoe. I've found efficiency in dadoing the widest stock available to create tread stock, then ripping & thickness sanding that stock as close as possible to dadoed increment for the perpendicular interlock. Less tear-out issues, less machining.

2. 
   jasmin | May 18, 2006 04:00am | #3

   Thank you John,

   You've provided me with a good alternative to my procedure, probably better than my proto-type first attempt.

   My lattice field is made of interlocking 3/4 x 1  1/2" members; a frame 2 1/4" deep x 'yet to be determined' width is to be side-notched to house the lattice. My safety valve here for error accumulation was to vary the end width of the supporting frame to halve the discrepancy over each end. Again, each end of the interlocking lattice would have to be oriented correctly.

   With your procedure, the need for 'perimiter' correction could be eliminated.

    

1. 
   JohnWW | May 18, 2006 05:26pm | #4

   When making a lattice all of the notches have to line up otherwise the lattice can't be assembled.  Your approach of making the correction on the end of each strip won't do anything to keep the notches in each strip properly aligned, that is why the process I outlined works better because the position of each notch is always located from a fixed starting point.

   Good luck with the job,

   John