Replies

1. 
   Jamie_Buxton | May 08, 2006 07:39am | #1

   Whew, that is one wide piece of lumber.

   Using that standard incremental finger joint jig seems like it wouldn't work.  Error buildup would be a problem. 

   I'd locate the fingers from some other source which has been precision-milled.   For instance, I can borrow a shelf-pin template which is 3' long, and has holes bored exactly every 32 mm along it.   Maybe you have access to something similar.  Another (more expensive) approach would be to use some of the Incra-Jig parts to build your jig.  

2. 
   citrouille | May 08, 2006 01:27pm | #2

   Cut the two sides of a corner together and trim the ends afterwards.
   That way you are multiplying the error the same way on both sides and the joint will look tight.

   C.

1. 
   jasmin | May 08, 2006 04:54pm | #3

   I don't think my explanation of the situation was complete enough. The project components consist of stair treads and landings that have the form of a lattice, made from perpendicularly interlocking components that measur 1 1/2" deep by 3/4" wide with 3/4" dadoes spaced 3/4" apart. I've contructed a jig along the lines of Lon Schleining's article in No.148 and end up with accumulation error of 1/16"+/- over 32".

   Are there any jigs like Incra or others out there that resolve this?

   Thanks

1. 
   BobMc | May 14, 2006 07:12am | #7

   I don't understand how you're using the joints for your stairs and landings, but am attempting to answer the root question.

   I'm not familiar with the article you mentioned, but you might look at the book "The Ultimate Jigs & Fixtures Handbook"  It has a box joint jig that uses 10-32 rod and inserts to adjust the key setting for width and distance from the the blade.  I use a dial caliper to set it up, and have had great results--but the longest joint I made was .25 width teeth over 22 inches.  I would probably make a few mods to handle a 7 foot board.

   Reality test:  32-inches of 3/4 inch cuts = 21 cuts

   1/16-inch error on 21 cuts = .003 error per cut. 

   1/16-inch error on 84-inches of 3/4 inch cuts = 56 cuts at .001 error.

   My solution would be to leave the stock a couple inches longer, and cut all the teeth on each mating piece.  Join the mating pieces in a flat configuration, and then trim each end square to yield 84".

    _____________________  |                     ||  _   _   _   _   _  | |_| |_| |_| |_| |_| |_|    _   _   _   _   _   _   | |_| |_| |_| |_| |_| |  |                     |  |_____________________|

    

    

   Edited 5/14/2006 12:18 am ET by BobMc

2. 
   QCInspector | May 14, 2006 09:36am | #8

   "stair treads and landings that have the form of a lattice" If I understand correctly you're trying to make a grate so that you look through the surfaces your standing on.I would look for a high quality piano hinge with 3/4' spacing and pull the pin out. The halves will serve as the pattern.Make a sled for the tablesaw with a stop that will fit between the teeth of the hinge that you can move (by clamp or screws) to adjust for fit.Take the 1 1/2" deep by 3/4" wide 7' pieces of wood that are taped or clamped together to gang cut ( can also use wide boards and rip after), and attach the piano hinge to the top of the bunch with a couple/three small short screws.With the dado blade set to cut a slot to make an easy sliding fit, push the sled through to make the cuts. Always applying pressure in one direction of the finger/hinge with all the cuts to prevent any tolerance errors.If you've got lots to do you may want to see if you can find someone with a CNC Router. They could make them with the precision you want especially if you give them wide boards to cut. And if you ever have to go back for more (hopefully not because of a mistake) they already have the program made and can slip it in quick.

2. 
   jasmin | May 13, 2006 06:24am | #4

   Thanks, Citrouille.

   That's the conclusion I've drawn, too. For further precision, all I can think of doing is giving my stacked dado set to a machinest to mill a more precise spacer block.

1. 
   John_D | May 13, 2006 07:33am | #5

   Well I drafted an idea earlier, and it got eaten by the software.

   I think I would try using my Leight D24R jig.  Use a straight bit on the through tails side.  Use setup blocks to get the fingers set at the exact right interval, then you could cut 24" at a time with essentially no error.  Move the stock once, cut another 24".  With your stock length, you'd only have to reposition, what, three times?

   It's odd but I think it would work.My goal is for my work to outlast me.  Expect my joinery to get simpler as time goes by.

3. 
   cowtown | May 13, 2006 10:49am | #6

   stair treads have to have lateral strength, so yer talking about 7' of box joints set up to split off. Ergo, no rational reason to do it.

   Box joints normally run the other grain direction, ie castellated into the end of the board. A grain orientation which is, if you pardon the pun, at 90 degrees to rational stair tread construction.

   Seems to me like you got an architect or designer who doesn't know much about even basic material science at all. And wood has been around from day one. Give a moment's pause to let that sink in.

   OTOH, you might be subbing this off an even more naive flooring contractor. I dunno, and you hasn't elucidated on details much

   Hold on .....

   There are on stair treads only two joints which could run close to 7' (and they certainly aint the rise and run...)

   one is where the tread rests on the riser, and the other one is where the riser rests on the tread.

   in the first case stair treads have to have nosing, ergo a box joint is not only unconventional, not likely conforming to codes, and also likely unsafe, not to mention impractical to cut,

   in the second case , a box joint would be completely hidden unless you were underneath the stair case, and even then it would be a kludgy methodology, not pleasing to the eye, and certianly an expensive option with no redeeming factors, like strength, visual simplicity, etc.

   To me neither the first or second case just don't make sense STRUCTURALLY. Neither do they make sense from the economics of production. This lack of "common sense" is so flagrant that it suggests that there may be an alternate explanation.

   Take a second, and third look at the plans,

   What you might be looking for is an indication that the "box joint" look required is merely a " finished appearance" and it may well serve you to query the architect or designer on this. If you got reason to query this forum, you got reason to query them.

   I've encountered errors on blueprints before, and sometimes I've even thought that they may even be intentionally put there just to assess the competency of the folks bidding. OTOH, if it was a genuine mistake, and you pick up on it and query the detail, it's likely the designer architec ain't likely enthused about the error, and if yer tactful about it, you win. Find a discrepancy either in design or dimension, and it can translate to a win/win situation when you query it.

   Where the 7' tread rests on the riser, it might not be improbable that if it is merely the appearance of a box joint that is required, that you countersink screw holes well into the stair tread, maybe to a depth of 1/2", and then make a cut in a square hole and fit a square plug into it and sand it off. The appearance of a box joint would of course mean cutting about 45 such holes across the each tread. (charge for this feature!!!)

   Nuff said, not that there ain't more, but this post is long enuf.

   Hey, tell us what transpires.

   Good luck.

   Eric