Replies

1. 
   SteveSchoene | Apr 16, 2007 03:23pm | #1

   If you get the two test pieces by just flipping one end of the board you are right.  But, if you get the pieces by cutting two separate test pieces and using the same end on each (ie. both ends from the left of the blade, or both from the right) then the test described in the book will work, I think.  And it is probably a little easier to just lay the boards on a flat test surface than to manipulate them so you can accurately measure the right angle. 

1. 
   FRA | Apr 16, 2007 05:57pm | #2

   Hi Steve,That's exactly what I thought when I used the apparently "perfect" setting of my blade on expensive timber.
   When you intersect two parallel lines (top/bottom of timber) at any angle, the resulting (wooden) angle at the long end of the cut must be identical to the (air) angle of the short side of the cut. It follows that you can insert another piece newly cut at the same angle and flipped over without any gap. If you don't believe me, sketch 3 examples: 15 degrees, 45 degrees and 90 degrees. Perfect fit each time.
   Either Euclid or David Freedman must be wrong. Or I make some incredibly stupid mistake in applying David's technique.Fritz

1. 
   BobMc | Apr 16, 2007 10:25pm | #3

   Go with Euclid.  And accurate angle blocks, squares, and 45s.  Then cheap wood and a mockup. 

   I got sucked into the "test cut" bs <not Band Saw> before, too.  Spent a bit of time with scissors and paper to reaffirm Euclid.  I was lucky because I used cheap wood to make a mockup before I ran my wallet through the blade.

   Be sure to post pictures of the final product.

1. 
   BobMc | Apr 17, 2007 06:14am | #10

   Amateur speaking, again.  The problem with test cuts is that they tell you that you need to adjust either the bevel or the miter--take your pick for the type of the cut.  Cutting a bevel on a 1 inch piece of stock requires that the angle be 45 degrees +/- about .1 to keep the gap at each corner less than 1/128"  For the gap to fall less than a sheet of paper, say .002 on 1 inch stock, you have to be within about 3/100 of a degree.  The 45 you used to set the bevel (or miter) is lying to you, otherwise your cut would have been correct--unless maybe, you're using a hollow ground blade, and don't have an accurate reference on the blade.

   So, you make the test cut, it says the bevel is wrong.  You have to set the blade to your 45, again, but now you also have to deal with mechanical hysteresis (slop in the gear train).  Your best bet to have a chance to get dead on is to use a dial indicator.  Say the gap was on the inside--the outside corners touched , the blade has to come toward the vertical.  If you put a dial indicator on the upper edge of the blade, (use a magnetic base on the table) and zero the indicator. you can see very minute movements.  You can always get back to your initial point by resetting the blade to return the indicator to zero.  I always try to approach the angle from the same direction (moving toward vertical).

   A miter gauge can also be set with a dial indicator.  Set the miter gauge to 45 degrees, put the long side of a 45 on the fence, so a short leg should be parallel to the blade.  Zero a dial indicator on the end of the short leg away from the fence, and push the miter fence  forward.  The indicator should read zero all the way.  If not, adjust until it does.

   Then learn to use fillers, shooting boards, or whatever.  I personally enjoy the futzing with the machinery, and will use days when I'm too tired to put flesh near cutters to tweak stuff.  I made an octagonal frame out of mahogany for a round table, 46" diameter.  Using an Incra miter gauge that I spent a while tweaking, the accumulated error was minute.  One gossamer shaving wasn't enough to close the gap, and two gossamers were too much (;>).   (Now if I could actually use a plane effectively, and could get to where I make gossamer shavings.)

   <oops emoticon, realizing the topic isn't setting the saw, but to not use ineffectual tests>

   <reminder emoticon that "Any Jokes???" in the Cafe is working toward 300 posts, but it needs help>

    

    

    

    

2. 
   FRA | Apr 17, 2007 01:37pm | #11

   Thank you very much for your moral support in my belief in Euclid and for your valuable advice.

   For those of you who are curious about my project, here are a few pictures (including David's non-Euclidean setup):

   http://www.tg29.at/winebox/

   The concept is a somewhat elaborate winebox with a 240V / 50W heating pad together with a 10 - 30 degrees centigrade thermostat. I like to drink my red wine at a slightly higher temperature than recommended and after taking it out of THIS box, no guest of mine dares to question my preference.

   As a final note, I still love David's book on box-making without which this project would haver never materialized.

   Fritz

2. 
   9619 | Apr 16, 2007 10:54pm | #4

   FRA,

   Sorry about your learning experience. Stuff happens. You have already gotten good feedback. My comment concerns the use of the term "expert" in the title of your thread "even experts get it wrong sometimes".

   I have often wondered what an expert is. I would guess that almost everyone on Knots is a woodworking expert. I have seen the book that you used, but I don't know the author. There is a tendency to believe that people who write books are experts. When I buy a book on woodworking, I really don't care that the author is an "expert". I do care that the author is an "expert writer on the subject at hand". What do I mean by that? In the field of cookbooks, there are billions and billions of cookbooks. The difference between the great ones and the trash is not whether the author is an expert cook, but whether the author is an "Expert cookbook author". To be good, a cookbook must have been tested and tested and tested and tested. Betty Crocker has good cookbooks.

   I would like the same thing from authors of woodworking books. Have they tested each chapter on hundreds of people like the ones that might read the book? Have they fed what they learned back into the book to insure that it is readable and understandable by the intended audience, and that the level of detail is sufficient for the reader to be able to do what is described.

   I would guess that this almost never happens in woodworking, or in most fields. IT DOES HAPPEN IN THE BEST COOKBOOKS. I bought Jim Tolpin's book on Table Saws. (the first version). I ran across numerous errors. I hear that the new version corrects those. Unfortunately I bought the first version, and there is no "money back guarantee", and there is no discount for buying the second edition.

   I am not picking on Jim. I don't know much about him, except he seems to have written a lot. After my experience with my first purchase of one of his books, it is highly unlikely I will ever buy another. Maybe he is an EXPERT woodworker. I don't know. It really doesn't make any difference. I wanted carefully written advice, and the quality of what I got was not up to my standards.

   Editors often check for grammar and typos. I'd like to see editors check to see if the writing is effective. I would like to see guarantees in the front of the book that the book was tested, and that the test group and the results are described.

   I know that is a lot to ask for. Until that day, I rarely buy a book that I have not leafed through, and read a bit of, to see if the author is a fairly understandable writer (to me). That ain't a guarantee, but it is the best that exists.

   Maybe Taunton will lead the charge to make woodworking books the same way that good cookbooks are made (with thorough testing).

   May all of your future projects go better. Rather than throw away the nice pieces of wood that didn't work in the box project, why don't you turn them into a nice sculpture. Pile them nicely, with glue between the pieces, and place a nice "dedication" to the author in a prominent place on the sculpture.

   Enjoy,
   Mel

   Measure your output in smiles per board foot. 

1. 
   joinerswork | Apr 17, 2007 12:55am | #8

   Mel,

   Expert, def:    "ex"= a has-been  +  "spurt"= a drip, under pressure

   Cheers,

   Ray

3. 
   DustyGeorge | Apr 17, 2007 12:30am | #5

   Fritz,

   I don't have David M. Freedman's book, but what you are describing is the test for a 90 degree cut.  To check for a 45 degree cut you would take two pieces cut from the left side of the blade or two cut from the right side, but them together and see if they form a right (90 degree) angle.  I generally use two pieces of 1/4 plywood about 6 to 8 inches wide, stack them and cut both pieces together.  When I check them with a carpenter square any error is readily apparent because the error is effectively doubled.

   Good luck, George

   You don't stop laughing because you grow old.  You grow old because you stop laughing. - Michael Pritchard
4. 
   forestgirl | Apr 17, 2007 12:37am | #6

   For those who crave pictures (click).

   forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-) 
   

1. 
   DustyGeorge | Apr 17, 2007 12:47am | #7

   Forestgirl,

   Very good, clear explantion.  I bet Euclid would agree.

   GeorgeYou don't stop laughing because you grow old.  You grow old because you stop laughing. - Michael Pritchard<!----><!----><!---->

   <!----><!---->

1. 
   forestgirl | Apr 17, 2007 02:45am | #9

   Thanks, Dusty.  I wrote that up and took those pics because the tests are so simple to do, but a royal pain to describe in words (and to understand, just reading the description). forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)