Having trouble with miter....

[Charles Palmer]

I'm trying to create a repeatable 60 degree miter on a RAS but am having a heck of a time getting my miters exact enough to have no (or at least an even) gap between the pieces when put together.

I'm using a professional quality RAS. I've set the saw to 60 degrees, cut the first angle, then used it to place a stop block on the saw's miter table. At 60, the triangle has gaps in the center. To compensate, I set the miter to 59 degrees. Now the gap is on the tip of the triangle.

I've gone through several 1*4*6' trying different tweaks, but no luck. Any suggestions? I've uploaded a couple of pictures showing what I'm trying to do and the angles I've come up with.

[roger wiegand]

Only way I've ever had success with miters is to keep tweaking on test pieces until it comes out tight. The required movement can be small fractions of a degree. If it never does get tight you may be running into the inherent limits of your saw-- if it moves the slightest bit when you apply pressure to it you'll probably never get perfect results. I've always had trouble doing this with a table saw (don't have a RAS) and end up tweaking the final fit on my stationary belt sander for small pieces. I clamp a temporary fence in place, tweak it on test pieces, then adjust my sawn pieces by sanding to the final angle. Crude but effective. Using a good quality protractor to strike a knife line to work to helps to set the initial angle close, the scales on most machines are too coarse and seldom accurate enough.

[David Eisenhauer]

If you use hand tools, perhaps you could clean up or final tweak the angle with a hand plane and shooting board. I have a very good, older RAS with the bearings still in good shape, but I can influence a square cut one way or the other by the pressure of my hand on the yoke.

[Dave Richards]

Can you set the saw up to make accurate 90° cuts?

You might find it easier to set up an angled fence and adjust it until you get the required angle. The problem with any radial arm saw is that unless you can pull the saw straight along the arm with no side loading whatsoever, the blade won't follow the desired course exactly.

[Andrew Pitonyak]

If you use hand tools, perhaps you could clean up or final tweak the angle with a hand plane and shooting board. I have a very good, older RAS with the bearings still in good shape, but I can influence a square cut one way or the other by the pressure of my hand on the yoke.

This is exactly what I was thinking, but either way, a perfect joint is difficult and seems to require perfect technique on a table saw based on really small issues.

I can't help but wonder if you were able to create one perfect example and then use a router with a template and a guide bit to make copies.

[Ken Fitzgerald]

Can you set the saw up to make accurate 90° cuts?

You might find it easier to set up an angled fence and adjust it until you get the required angle. The problem with any radial arm saw is that unless you can pull the saw straight along the arm with no side loading whatsoever, the blade won't follow the desired course exactly.

I'm with Dave on this one. When I had to make a lot of repeated angled cuts for some 45º wainscot, I found it easier to set an angled block attached to the fence/table to make 90º cuts referenced to the original fence and 45º referenced to the angled fence.

[Ian Moone]

Wixey digital angle gauge and sliding bevel!.

You can measure to 0.1 of a degree.

So, when setting your RAS and test cutting your test piece off the fence, to get your angle, if you fit your sliding bevel to the cut angle you wish to measure for accuracy & lock the wing nut at that angle - then you can use a flat steel saw bench as your level reference angle on the wixey gauge for 0 degrees & laying the back of the sliding bevel back against that reference 0 degree surface measure the angle of the blade with the gauge to see exactly what you cut so you know how much you need to alter the arm of the RAS!

You can even tap your sliding bevel to exact 45.0 off the flat reference surface & keep altering your RAS arm degrees until you do get a perfect fit with the sliding bevel.

When doing multiple compounded angles (i.e. 22.5degrees) - the requirement for accuracy goes up by the number of pieces!

Adjusting the RAS angle sometimes its a matter of leaving the tightening lever just partly done up so that you have to gently tap the arm sideways a fraction of a time with light hits of your palm...and keep test cutting scrap until you get it right... but that 0.1 degree of the wixey makes all the difference!.

http://www.wixey.com/anglegauge/

If it's just 45degrees you want you can reference (*zero) your wixey off the saw bench and then stick it by its magnetic face to the saw blade on a tilting arbor saw and wind the tilting blade over until you have 45.0 degrees exact - lock the tilt blade and do a test cut and keep repeating them with the sliding table and fence past the tilted blade.

It's hard to be that accurate with old hand tools - not impossible just harder and requires patience skill and trial and error (and wasting a lot of scrap on test cuts).

OMMV

[Doug Hepler]

Charles,

I cut lots of segments for making turning blanks. The difference between a close fit and a gap is much less than a degree. From your description, you overshot the correct setting when you went from 60 deg to 59 deg. You can waste a lot of time (and patience) by trial and error.

On a table saw, I make a sled. The secondary fence is set by trigonometry. The math is really easy. tangent(60 deg) =1.732= rise/run. So start with a piece of plywood 12" square. (Its a good idea to make sure that one corner is square) The longer the sides, the more accurate the cut angle will be.

Since you have a "rise" of 12" then the run should be 12/1.732 = 6 15/16". So measure 6 15/16" along the side opposite the right angle. Now you should have drawn a right triangle on the plywood with sides 12" and 6 15/16". Cut it out.

OK, so now you have a wedge that has angles of 60, 30 and 90deg. (actually, 59.95 deg) Recheck the lengths of the sides. Use this to lay out your 60 deg cut (it's more accurate than the angle marks on the saw) or use it as a secondary fence. That should get you the fit you want. If not, the tweak to 60 will be minuscule. I can't say much about the precision of your RAS. Earlier comments may be some or all of the problem. You can draw a layout line on the workpiece and plane to the line instead of re-adjusting your saw.

Another useful trick is to note that 9 segments should add up to 180 deg. So, one you are as close as you can get, go ahead and glue nine segments together into a half circle, as if they were perfect (no gaps). Then trim the resulting edge to a straight line. Repeat for the second three segments.

Hope this helps

Doug

[Charles Palmer]

Thank you ALL for your replies. Makes me think that it's not "all me."

I'm going to give it another shot tomorrow.

I've seen an adjustable table saw miter sled on that video site. I may end up trying to make the adjustable sled.

[Lee Schierer]

The problem with a RAS is that the saw carriage rides on rails and the bearings that travel on the rails are not precision fits with the rails. The raise lower column is a close fit but not perfect. The fences are usually made of wood and are not precise. Any variation in the density of the wood can deflect the overhanging arm, take the slop out of the fit in all joints, etc to give you the kind of variation you are seeing. They just aren't precision instruments.

I'm no sure sliding compound miter saws are any better.

[Roy Harding]

The problem with a RAS is that the saw carriage rides on rails and the bearings that travel on the rails are not precision fits with the rails. The raise lower column is a close fit but not perfect. The fences are usually made of wood and are not precise. Any variation in the density of the wood can deflect the overhanging arm, take the slop out of the fit in all joints, etc to give you the kind of variation you are seeing. They just aren't precision instruments.

I'm no sure sliding compound miter saws are any better.

In my experience, SCMS are no different in this respect than RAS. CMS (non-sliding) seem tighter, but are still subject to "slop". Depending upon a lot of factors (materials involved, what the finished product is, etcetera), I'll go one of two routes:

1. Power tools - I cut the mitre on the best saw for the job (table saw/CMS/SCMS etc), then use my edge sander, with an accurate jig (permanent, and built a long time ago to provide a dead on 45, or 22.5), to finish the mitre.

2. Hand tools - I cut the mitre on the Nobex mitre box (http://www.leevalley.com/en/wood/page.aspx?p=32926&cat=1,42884,43836), and then fine tune on a shooting board.

The second method is consistently more accurate, but also more time consuming.

[Mike Schuch]

My radial arm saw keeps very good angles but it would not be up to what you are attempting. Here is my picture frame jig on my RAS. The perfect 45's this jig produces is only half the battle. The other half of the battle is getting the parallel pieces the exact same length. I don't know how to make a jig for your project exactly but I think something like this would be a good start.

324337

I had a nice Bosch 10" cms... it was never close for picture frames. There were always gaps.

A good CNC router would probably be your best bet for "perfect the first time".

[Pat Barry]

I'm trying to create a repeatable 60 degree miter on a RAS but am having a heck of a time getting my miters exact enough to have no (or at least an even) gap between the pieces when put together.

I'm using a professional quality RAS. I've set the saw to 60 degrees, cut the first angle, then used it to place a stop block on the saw's miter table. At 60, the triangle has gaps in the center. To compensate, I set the miter to 59 degrees. Now the gap is on the tip of the triangle.

I've gone through several 1*4*6' trying different tweaks, but no luck. Any suggestions? I've uploaded a couple of pictures showing what I'm trying to do and the angles I've come up with.
Frankly, I'm amazed at how close you have it right now. To me that pattern seems to be an incredibly complicated thing to recreate perfectly. I think that the advice to make an adjustable fence and tweak that til you get best fit. On the other hand you might be able to stay with the existing fence and add a bit of a shim to compensate for a less than 1/2 degree tweak. That shim might only be a sheet of hardboard / paper. I don't think that you have a tracking problem with the saw however as your cuts seem to be very consistent. Good luck. It should be a very sweet finished design.

[Davis Young]

Second what Dave and Ken says. The bonus is if for whatever reason you need to cut from the other side of the blade, you can just flip the angled fence over (assuming it's registering against the original fence) and maintain a consistent angle.