Making a precise bevel rip

[MarkSHancock]

quote:Originally posted by RainShadow

This post is way beyond my little pea brain...............

Good job guys, Randy from Oceanside, CA

Yep, getting Engineers discusing a problem is a dangerous thing.

[RodKirby]

Well - a Software engineer's opinion - FWIW

Timber/lumber (compared to metal) is very forgiving - and "cheap".

However, I DEMAND precision in my WW - one of the reasons I bought my BT3000 - 10+ years ago.

I treat the angle marks (SMT), and the rail scale, as, at best - a guide only. As Loring mentioned, many of us do.

If I have multiples to do, I always test cut/fit before I hit the real stuff. There is no substitute for taking this step.

One thing I do know - once I have set up stops etc., the BT will ALWAYS give me identical work pieces.

If the BT cuts wrong - I screwed up!

[MarkSHancock]

Rod,
FWIW, I'm also a Software Engineer (though a BS/MS EE by training). I have flip flopped a few times. Right now I'm doing control software for precision machine control (speeds up to 8 m/sec with positional control to 1-5mm). So, I get to play with a lot of cool mechanical stuff.

But, on the home front I have two little boys so for now the table saw is just a functional thing for home repair and finishing my basement. The only time I usually have is from 10-midnight and sometimes on weekends. This project is to build box columns around my basement support Lally-Columns.

Here is some concept modeling I posted to a sitehttp://www.hometheaterforum.com/gall...g2_itemId=1475. If you note from the floorplan there is a plan for a woodshop.

Mark

[RodKirby]

Looks great, Mark

[JimD]

I am a ME and I agree with Loring. I do not think you can create a table saw where the rip fence scale works regardless of the angle of cut. If the blade had no height and you located the pivot at the center line of the arbor you would get the result you desire except a blade with no height doesn't cut anything. Anybody who says their saw will do this needs to check again.

I use the scale and consider it a drag that it doesn't work well when cutting bevels but that is the way it is. I also wish the BT3100 was left tilting but it isn't (and neither is any saw much under $1000). You can measure the offset if you want but you need to remember that it is for the angle and if you are measuring on the top of the wood it will also be only for that thickness. I just use scraps to set the rip fence. If I had to make the column a little smaller to avoid trying to use every single rip, I would make the columns a little smaller. You may be able to set the fence and then flip the sheet goods around but I find it hard to get real accurate bevel rips on big pieces of sheet goods. I like the suggestion of cutting rectangular pieces and then ripping on the bevel (or you could use the router table even at that point).

Jim

[gmack5]

You guys need to remember your High School Geometry!
The end view of your work piece is a Trapizoid, with 45 degree angles at the ends. On a 45 degree triangle (the end view of your "bevel" cut) the thickness of the material X 2 added to the inside dimension (the top edge of your trapizoid) should get pretty close to your desired outside dimension. I assume that you're going to use 3/4" thick MDF, therefore the top-side dimension for the 8" wide piece would be (8-[2 X 3/4]) or 6 1/2" wide at the inside surface.

[LCHIEN]

Here's a drawing of how the BT3x is really set up... the pivot is below the table:



You'll note how the pivot (blue dot) is located below the table distance "B" and the bevel angle offsets the blade by Distance "A".
If the bevel is exactly 45° then you have made a 45-45-90 isosceles right triangle and A=B.

Someone uploded this pic of the underside of the BT3 the other day, I indicated where the bevel points are with the green lines:



The bar across was the picture's original poster's dust curtain hanger rod. I feel like I"m looking up a rubber skirt...

[RodKirby]

I would add to Loring's post:

QED

[MarkSHancock]

Great photo Loring.

Today I tried a few more things to verify I can use the 5/16 offset I measured to set the fence for a 45 deg bevel cut. I ripped a short 2x4 scrap and it was true to scale within 1/16. As discussed before, I will be repeating the cuts for opposing pieces without changing the saw/fence settings so 1/16 should be way good enough. Since, as you noted from your diagram A=B for a 45 deg bevel (more specifically A=B*tan(bevel angle) ), and since the part that sets B seems from the photo to be a casting, I bet most people would find the bevel offset for a BT3100 to be 0.32*tan(bevel angle).

Another couple of questions:
1) Several people mentioned doing straight-cut rips first then adding the bevels. Why would this help? Since both edges are beveled, the final cut that sets the width of the piece must be done with the already beveled other edge against the fence anyway.
2) I have found with the experimental cuts I have been doing that my staight cuts are not perfectly straight and that the bevel cuts have not been perfectly 45 deg (at least compared to my speed square); they seem to be a fraction of a degree off. What is the best technique for getting the angle as close to perfect as I can? What is a reasonable expectation for how close I can get to true (in this case I mean the cut piece not the scale).

P.S. Here is my present cut plan for a 49" wide piece of MDF

First, I will rip the piece down the middle to create two 24-7/16 wide pieces. Then I will cut the as follows, ripping every cut with the saw set to a 45 deg bevel and making the same rip on each piece before changing the saw settings.

|--------- 24-7/16"------>| Rough-cut rip with circular saw
|----------24"---------->/. Bevel the rough edge, ref to fact edge
.................<-15-1/4"---. 1st piece, ref to top of beveled edge
./<-7-1/4"---................... 2nd piece, ref to top of beveled edge

Note:
1) Turn 2nd piece over before cutting to widest portion on top.
2) The beveled edge will rest against the approg 1/4" recessed area on the fence so I will have to account for that too (above number don't)
3) Test Cut, Test Cut, Test Cut, Test Cut .... before any real rip

What the cross section would look like:
________________________
|/______\_______________/|

Accounting for a 3/16 kerf width, alog the top you get:
3/4 + 3/16 + 6-1/2 + 3/16 + 16 + 3/16 = 23-13/16
This gives me only 5/8 of scrap - very tight!
But... (if all goes well), I will end up with a standard 8x16 column out of a single piece of MDF.

TTFN,
Mark

[gonesailing]

Hey, I agree that you could not put the pivot point at the table surface using a hinge or pin to swing it. However notice in the picture that there are circular cutouts on both sides of the saw, one allows the sawdust exit to move, the other has the grear train to move the saw mechanism. Get rid of the pins, and replace the slots with a guide rail that the whole mechanism rides on, raise it a half an inch or so, and the VIRTUL pivot point will be at the surface of the table. Heck, I have an automatic cat liter box that follows a much more complex pattern than that.

Also, many telescopes use much the same type of mechanism, i.e. rather then hanging on a pivot they ride on a carriage that in turn rides on a track, and thise things are far more precise than a table saw ever need be.

Given the complexity of some of the jigs that Rod has made, I would just about bet he could design and build such a carriage sustem for the BT3k that would work fine.

Mike