A Four-Cut Method for Squaring a Crosscut Sled (no math, no feeler gauges)

[Paul Ruud]

Best wishes to everyone! I dropped one of my crosscut sleds and knocked it out of square. Now with a little break from teaching I had a chance to try an idea I had for squaring the sled. It worked nicely so I thought I would share it and see what you all think of it as an alternative to BIll Ng's 5-cut method.

This method for squaring the fence of a crosscut sled relies on an old method for checking whether a cut is square. Make a cut, flip one piece, and put the two pieces together to check for a gap. Because two of the cuts in this method are only required at the start, this method is almost a two-cut method and goes through less wood.

--Paul

Method

1. Cut a board so that two sides are parallel. Keep the distance between these side less than the width of your crosscut sled fence.
   To make the parallel sides on a crosscut sled, cut the board on one side and then slide it over to cut the other side. Do not turn the board. Just slide it sideways.
2. Put one of the parallel sides against the fence and cut an edge.
3. Flip the board end-over-end, placing the other parallel edge against the fence, and make a second cut. The offcut will be your gauge for adjusting your fence. Mark the parallel edges to keep track of whether the (bottom) edge against the fence after the second cut is narrower or wider than the other (top) edge. Use calipers to measure which end is wider.
   In theory, the angle between the last two cuts is twice the correction needed for the fence to be square. If you have a keystone shape, with the widest edge at the top away from the fence, then the fence needs to rotate counterclockwise.
4. From the pivot screw that holds the fence, measure along the fence twice the length of the offcut. You can use the offcut itself for this.
5. Place one end of the offcut at the measured point. If the fence needs to rotate clockwise then put the widest end of the offcut at the point you have just measured. If the fence needs to rotate counterclockwise, place the narrowest end there. Butt a pointed stop up against the end of the offcut and fix it there on the sled. I use two pieces of blue tape and CA glue for this.
6. Unscrew the nonpivot end of the fence and move it so that the opposite end of the offcut is between the stop and the fence. Clamp the fence in its new position and fix it there with a screw in a new location.
7. Return to step 2 and repeat until the offcut has the same width at both ends.

Comments

• This method does not amplify the angle as much as Bill Ng’s 5-cut method and can be less accurate, all other things the same.
• Both methods rely on approximations that cause the adjustment to be slightly less than required. This explains why at least two iterations are often needed. In my experience, a second round with an additional two cuts is sufficient.
  One can remove the math from Bill Ng’s 5-cut method but probably with a sacrifice in accuracy. For example, cutting the final offcut into quarters length-wise gives the adjustment to apply at the length of the offcut. This effectively divides the difference in end widths by 4 (not accounting for the saw kerf).
• In my experience, measuring wood cuts at single locations gives a misleading impression of accuracy. Along a piece of wood, there is some variation (as much as a couple of thou) in the width of the offcuts that is caused by imperfections in cuts and inconsistent use of the calipers. Measure at several locations to get a sense of the accuracy.

Here is a Sketchup image to help understand. (I do not know how to make it a reasonable size.)
Screen Shot 2020-11-23 at 7.49.40 PM.jpg
On the right, it shows the offcut repeated to find the point to place the stop block. On the left, is what happens after switching the end to get the adjustment.

[Kevin Jenness]

I use another method which works well enough for my purposes. Using a panel at least 2' x 4' with a straight edge I crosscut it at one end and in the center, then flip the first piece end for end keeping the reference edges against the crosscut fence. Butting the two newly cut ends together will show whether the cut is square; any gap is twice the error. The wider the crosscut, the more obvious any error will be. Obviously, the exercise depends on a truly straight fence and reference edge on the panel.

I tried the 5 cut method but found it time consuming. This has the benefit of simplicity and I am in no doubt of which way to adjust the fence. I have scribed marks on my sliding saw extension table at front and back of the outboard end of the fence so I can see at a glance if the fence has been displaced from square.

[Paul F Franklin]

Interesting variation. You might also want to check out Tommy Mac's Three cut method; he has a video at his web site.

[Johnny Barr]

I've found the best way to align a sled without one single cut. https://www.youtube.com/watch?v=tZM1OBcC6ok
Well, the only cut is a test cut at the end but each time its spot on. It relies on a dead straight fence and an engineers square with a smooth edge that has to be bang on 90 deg.
When I first tried this I ended up doing the 5 cut method as a check and it was near perfect.
Simply have a pivot on the fence and keep moving the fence against a dial indicator until there is little or no movement then clamp it down. Check again and then screw in the fence. The video gives a better explanation. Takes 5 minutes without any mess. This works really well with a miter gauge that has a small fence as the 5 cut method of adjustment isn't practical. Anyway I swear by this method and have used it on every sled I've made.

[Paul Ruud]

Thank you for your replies!

Kevin, I think your method is what I was trying to describe as an old method for checking square. I guess I thought my insight was using an offcut to gauge the movement of the fence rather than measuring with calipers and doing some math.

Paul, I tried to find Tommy Mac's 3-cut method but could not find it. Do you have a link? Your comment and Kevin's made me realize that I made a mistake. I should just use one reference edge and not try to make two parallel edges. My method for getting two parallel edges won't work unless you are starting with a straight edge. Given a straight edge, I can use two cuts to get a physical measure of the gap (the offcut) that is uncovered by flipping one half of the board from a single cut. I just need to flip the piece so that the reference edge stays against the fence. No need for a parallel edge. Then I use the offcut from that to set up a stop for moving my fence.

Johnny, that method is so cool. It's a zero cut method! Thanks so much for mentioning it. I think that is the method that I will use also. And it can also be used for 30, 45, and 60 degree angles if you have triangles with those angles. A large combination square would work well for a 45.

[Paul F Franklin]

Thank you for your replies!

Kevin, I think your method is what I was trying to describe as an old method for checking square. I guess I thought my insight was using an offcut to gauge the movement of the fence rather than measuring with calipers and doing some math.

Paul, I tried to find Tommy Mac's 3-cut method but could not find it. Do you have a link? Your comment and Kevin's made me realize that I made a mistake. I should just use one reference edge and not try to make two parallel edges. My method for getting two parallel edges won't work unless you are starting with a straight edge. Given a straight edge, I can use two cuts to get a physical measure of the gap (the offcut) that is uncovered by flipping one half of the board from a single cut. I just need to flip the piece so that the reference edge stays against the fence. No need for a parallel edge. Then I use the offcut from that to set up a stop for moving my fence.

Johnny, that method is so cool. It's a zero cut method! Thanks so much for mentioning it. I think that is the method that I will use also. And it can also be used for 30, 45, and 60 degree angles if you have triangles with those angles. A large combination square would work well for a 45.

Paul,

If you search for "tom mclaughlin three cut method" you should get to the youtube video. Or you may be able to get to it from the Shop Talk Live podcast page over at Finewoodworking; they interviewed him an episode or two ago and talked about the method.

[ChrisA Edwards]

I just made myself a new crosscut sled this weekend. I did use the 5 cut method and fortunately got it dialed in on the second adjustment.

But I do like that dial indicator method, posted by Johnny Barr. I'm going to use that to verify the 5 cut method just for kicks and giggles.

[Paul Ruud]

Paul,

If you search for "tom mclaughlin three cut method" you should get to the youtube video. Or you may be able to get to it from the Shop Talk Live podcast page over at Finewoodworking; they interviewed him an episode or two ago and talked about the method.

Thank you, Paul! I just watched it. He does have the same geometry that I have in mind. The difference is that I don't use feeler gauges or tape to figure out the offset. As I said in my previous reply, I thought it was cool to just use an offcut. That seems more direct to me and, perhaps, less vulnerable to measurement error.

[Mark e Kessler]

This is on a slider but i use a starrett 24” square and a dial gauge. You so have to clamp the square down with some blocking but take at most 1/2hr to get it dead nuts with almost zero diff over 24” which at 8-9’ is negligible. Even better would be to use Brian lambs square, using the Starrett is a bit clumsy but works and is better than 5 cut any day of the week

[Steve Demuth]

I've found the best way to align a sled without one single cut. https://www.youtube.com/watch?v=tZM1OBcC6ok
Well, the only cut is a test cut at the end but each time its spot on. It relies on a dead straight fence and an engineers square with a smooth edge that has to be bang on 90 deg.

Yep, if you've got a large machinist's square and a dial indicator, this is a great method. Basically using tramming technique for table saw crosscut fence.

I would hesitate to assume that most amateur woodworkers who would benefit from a crosscut sled have those two tools, however.

[Wes Grass]

I use a 12" combo square on my Felder, just to verify it's 'on'. Eyeball a tiny gap on a tooth, run it to the other end of the square and eyeball the gap is the same. In good light my progressives are good enough to a couple thou'. Otherwise, Magnavisor ...

However, I re-engineered the attachment hardware for the crosscut fence so there's near zero play in it. I'm not relying on the threaded shank of the pivot hardware anymore.

[Paul Ruud]

So, do any of you use the combo-square-and-gauge approach for a 45-degree fence?