Miter sled tuning

[Michael Merrill]

I don't know how many folks here use a miter sled but I'm curious on the method to accurately tune the miter sled for absolute accurate miters? My sled is built based on a plan from Wood magazine some time ago, 3/4 ply base with maple runners and maple miter rails.

Right now I would guess it cuts about 44.8 degrees or so but not right on the money 45.0. I just built some picture frames and there pretty good but just off a touch.

I'd like to get this on the money for future projects, any ideas? What is the most reliable way to adjust this?

[Doug Shepard]

When I did mine I just put a finish nail thru the end closest to the blade to allow it to pivot. Then I used a speed square to get in the ballpark and clamped the other end to a scrap block that I tacked to the sled. In between trial cuts and checking the results I would insert busness card or paper shims between the scrap block and the fence. Once I had it where I wanted it, I pivoted the fence out of the way and added glue. Then I clamped it back against the scrap block with the shims in between and screwed the fence down to the sled.

[tod evans]

michael, 95% of the mitersaws on the market are accurate and the non-sliders are really inexpensive nowadays......is there a reason you must cut miters on the tablesaw?..tod

[Michael Merrill]

What was your method of test cuts and measurement? I assume just using some straight jointed stock and two mitered cuts, then fit checked with a framing square?

I also, wonder if only one of my sides is off, I guess I could fid out by making two cuts on each side and rotating one to check that miter joint along the framing square?

[Michael Merrill]

Tod, my experiance with my miter saw has not been so good. It's a Craftsmen 12" with laser guide. I bought it to cut all the baseboard and crown in finishing my basement and that was fine. But I've never been able to get super tight miters for picture frames using it. Maybe it's not the best? Maybe it needs tuned?

[tod evans]

michael, try using some mdf scrap ripped parallel to the width of your mitersaws capacity at 45deg. turn off the darn laser:rolleyes: and using stops try cutting four pieces into a frame, be sure you lock the saw head so you don`t unintentionally nudge it when you`re cutting and see what you get? i suggest mdf for two reasons; cheap-n-flat......02 tod

[Doug Shepard]

What was your method of test cuts and measurement? I assume just using some straight jointed stock and two mitered cuts, then fit checked with a framing square?

I also, wonder if only one of my sides is off, I guess I could fid out by making two cuts on each side and rotating one to check that miter joint along the framing square?

I was using some Cherry sapwood scrap that I ran through the jointer and planer. I was checking with the best squares I had at the time - a 14" Freud cabinetmakers square and a 12" Starrett combo square. At first I was just checking a 90 deg. corner. Once I thought I had it right I cut the 45's on the end of 4 pcs, then clamped a scrap stop block on the fence and cut the other ends all to the same length and checked the resulting square frame to see if the joints looked like they still needed to open or close up. I think I still need to nudge it a paper shim or two after checking with a 4-sided frame vs. just checking 2 pcs at a corner. I've noticed though that my sled is a about a business card thickness off from 45 since switching to a PM66. I made it when I still had a contractor saw and gluing the fence down means I can't adjust it for the newer saw. It's about time I make a better one anyway.

[Michael Merrill]

Doug, good ideas on checking both the single corner then a full frame. Also just thinking that I want to use something at least 12" in lenth to make sure it's good, anything shorter will be tough to tune.

[Jim DeLaney]

Michael,
If you're cutting 44.8° on one side of the jig, that means the other side is 45.2°, which, when joined, will still yield a perfect 90°.

So, the solution is to cut one side of your miter on the left side of the blade, and the other side of it on the right side of the blade.

If you do this consistently on all four sides of the frame, it'll come out perfectly square.

[Michael Merrill]

Jim, as the two side rails are independant it's not for certain the the sum is 90°. If the sled is off like I think just a bit I think it would lead to assemblies being a parrellelagram and not a rectangle or square.

I think I need to invest some time in tuning, so I can trust the cuts more.

[Jim DeLaney]

Jim, as the two side rails are independant it's not for certain the the sum is 90°. If the sled is off like I think just a bit I think it would lead to assemblies being a parrellelagram and not a rectangle or square.

I think I need to invest some time in tuning, so I can trust the cuts more.


I guess your sled is made quite differently from mine. Mine has a base the full width of the saw's table, and uses both miter slots. In the center, it has a 90° triangle centered on the blade kerf. I can cut one miter half using the left side of the triangle as a rest, and the other piece using the right side of the triangle as a rest.

If you've ever watched David Marks' TV show, and seen his miter jig, mine is pretty much the same, excepting his is machined from aluminum, and mine is Baltic Birch.

[Nissim Avrahami]

This is one of the pictures

http://i50.photobucket.com/albums/f321/avrahami/Frame%20cutting%20sled/sld6.jpg

The full threads are here:
http://www.sawmillcreek.org/showthread.php?t=38482
http://www.sawmillcreek.org/showthread.php?t=36590

niki

[Michael Merrill]

OK, I took a shot to try to explain the design a bit better. This is much like you have mentioned except two rails. In building it you do have a triangular guide to set the rails but then it is removed.
http://home.comcast.net/~sm.merrill/Posted_photos/SawMillCreek/miter_sled.JPG

This show now what I'm saying about the two rails, once again not my design, but from Wood Magazine. I'm not looking to scrap it and start over as it is very close, just needs tweeked.

[Jim DeLaney]

From your pictures, it seems like what I originally said will also work with your jig: Cut one half the miter on the left side of the jig, and the other on the right side. As long as your triangular guide is a perfect 90°, then the resulting miter halves will go together at 90° as well.

[Michael Merrill]

Jim, at the end of the day I'm not sure I can count on the sum being 90° with the two sides separate at this point. I think I'll have to play with it this week.

[jamie shard]

Reviving this thread...

I've spent way too much time trying to figure something out... it might be easy for someone smarter than me on the creek! Here goes:

It sounds like one common way to test a 45 degree miter is to make mitered frame. Cuts are made on one end of four pieces of wood. For the other end, a stop block is used on the miter fence, so that the next set of cuts result in all the pieces of equal length. In theory, all the angles are the same, all the lengths are the same. If the frame is assembled together, the last corner will have a cumulative error that can be used to determine if there is an error.

So here's the question: How to >calculate< the error based on the gap?

For example, if the miters are all 44 degrees, then the gap should be 8 degrees right? 360-(8*44)= 360-352 = 8 degrees. So taking the angle of the gap and dividing it by 8 is the error. 1 degree error on the miter fence.

Now obviously the error is hopefully going to be much smaller than that, so what I'm trying to figure out is how to calculate the error not based on angle, but based on the thickness of the gap. For example, using feeler gauges to measure the error.

It seems like gap would be proportional to the number of sides and to the length of the cut. So let's say the gap is .08". And the length of each miter was 2". What would the slope of the error be? And how would adjust the fence using feeler gauges?

???

[glenn bradley]

OK, I took a shot to try to explain the design a bit better. This is much like you have mentioned except two rails. In building it you do have a triangular guide to set the rails but then it is removed.
http://home.comcast.net/~sm.merrill/Posted_photos/SawMillCreek/miter_sled.JPG

This show now what I'm saying about the two rails, once again not my design, but from Wood Magazine. I'm not looking to scrap it and start over as it is very close, just needs tweeked.

One or both rails will need to be adjusted. there's no magic pill ;-)

Are the rails glued down or are they adjustable? As stated, the sum needs to be 90* but, it is always nice to have either side be 45*. If your miters are open at the outer corner, adjust the rails so that they are more open. If the miter is open at the inner corner, do the opposite. If the rails are not adjustable, remove one and make a replacement that is.

[bill tindall]

I just built one like it. I put a bolt through each end of the legs with an oversized hole in the lower part of the leg and a tight hole at the top end. The bolt hole through the 3/4" plywood is tight in all cases. I made a test cut on two sticks on the left side, put the miter together and checked square. Loosened the lower bolt in the oversize hole and moved leg to adjust out the error. Repeated the two cuts and test fit. When it was perfect I tightened the lower bolt in the oversized hole. Repeat for right side.

[jamie shard]

I think I figured it out. This page helped me visualize it better: http://www.startwoodworking.com/post/use-test-cuts-accurate-machine-setups

If each miter creates a .01" gap off of an ideal 45o miter, then eight of them will create .08" if you put them together and let one corner contain all of the gap.

If each miter is 2 inches then that gap was created over 8". So the error is .01" per inch.

To correct the miter sled angle, take the distance inches between the pivot point and where the correction will be made and multiply that number of inches by .01".

So let's say there is 10 inches between where the fence pivots and where the correction is made. If you put a stop block with a .01 x 10 = 0.10 feeler gauge against the fence, then take the feeler gauge out and move the fence to the stop block and screw it down... that will be very very close to a perfect correction of the error.

Basically it's similar to the 5 cut method, except with miters.

Sound right?