I got a new table saw almost a year ago and when I sold my old saw, I included all the sleds I had with it. I’ve been meaning to make a new crosscut sled since and I’m finally getting around to it. I’m going to be making essentially the same sled which worked great for the 4 years I had and used it. Super simple, basic and no frills.
My last crosscut sled served me very well. I used what I had in the shop to construct it. The runners were made from purple heart, the base was 3/4″ mdf, the front fence was two strips of 3/4″ mdf laminated together, and the rear fence was a piece of a pallet.
That sled worked so well for me and since I personally don’t care for bells and whistles, I’m going to recreate it with what I have on hand. Many years ago, I took a scrap of purple heart and milled it into runner stock. The stock was a bit too wide for the miter slots in my tablesaw so I gave them a light pass with a handplane to get them fitting well. These should slide fairly easily in the slot but not have any side to side play.
The base on this sled is a quarter sheet of 1/2″ mdf which I’ve had in the shop for a while. The base gets screwed to the runners with some small #6 screws.
To make the whole thing slide much easier, I coat the underside of the base and the runners with paste wax.
Next I can add the rear fence. This fence’s only purpose is to hold the two halves of the sled together once the kerf is cut through the base. I’m using an offcut from the bed swing.
Now on to the most important component: the front fence. This is what the workpieces will reference against so it needs to be straight and stay straight. I like to use mdf for this as well since it has no predisposition for movement. I laminated 3 pieces of 1/2″ mdf and clamped the glueup to my bench. Since my bench is flat, it will ensure that the resulting lamination will also be flat.
Once the glue has set, the fence can be trimmed to size. I like to offset one of the laminations so it can be referenced off the table saw fence making trimming it down easier.
Last little detail on the fence is to add a chamfer to the front edge. This will give sawdust a place to go so it doesn’t throw off the work piece’s reference. I’ll then add a screw at one end to start attaching it to the sled.
That one screw allows the fence to be pivoted into square. I use a scrap of wood in the kerf of the sled and a square to move the fence until it looks like it’s square. Getting it precisely square at this point isn’t critical since we’ll check and adjust it in the next step.
To check and adjust the sled, I’m using the 5 cut method that was popularized by William Ng. You can find his video about this topic here. William goes into a lot of detail on this method in his video. To start, I make four cuts on a scrap rotating the cut face into the fence after each cut. On the 5th cut, I’ll remove a strip from the scrap. If the fence is perfectly square, that strip will have a consistent width. We can measure each end of the strip to find it’s deviation and use that to adjust the strip with a bit of math.
First take the difference in width and divide it by 4 (the number of cuts made). That number then get divided by the length of the strip. This gives you a deviation from square per inch of length. Lastly to determine how much to pivot your fence, take the distance between the two screws in the front fence and multiply it by the per inch deviation.
So here’s the results from my first test. I need to shift the right side of my fence forward by almost a 1/10th of an inch.
To know exactly how far to move the fence, I stacked a few feeler gauges together to get a thickness of .097″ and placed them between the fence and a block clamped to the sled. I can then remove the feelers, unscrew that side of the fence and push the fence forward until it contacts the block. The fence can then be screwed back down using a new hole for the screw.
After that first adjustment, I repeat the 5 cut method and here are my results. If I wanted to adjust this one further, I would need to move the fence .0015″. This margin of error is beyond acceptable.
What really matters to me is being able to hold a square up to my work piece and being able to see that it is perfectly square to the eye. You can detect a crazy small amount of deviation by eye by holding the square and work piece up to the light and looking for any light visible under the rule. This sample was spot on.