Share your craft projects
Make new craft buddies
Ask craft questions
Blog your craft journey
Good thread George. As a machinist-turned-engineer, I have some thoughts on using wood for fixtures/jigs that I generally try to keep to minimum. But since you opened the door...😆
How precise should you dial it in when you make it? Well, as you said, over time it's only going to get worse. So I say as precise as you possibly can. Regarding sleds specifically, the 5-cut method is a sound one IMO and a method that most woodworkers can use with measuring tools they already have. I'll go with the supposition that (a) the blade is set dead parallel to the miter slots and remains that way and (b) the sled fence is dead flat and can't move and (c) the miter runner is locked in solid and can't move and is consistently tight through the entire cut in the slot.
Let me back up a bit... In mechanical design, we use a principle called exact constraints. What this basically means is that you control a design by making sure that no components are over or under constrained. Applying this to a table saw sled, you have 6 degrees of freedom to consider:
How precise should you dial it in when you make it? Well, as you said, over time it's only going to get worse. So I say as precise as you possibly can. Regarding sleds specifically, the 5-cut method is a sound one IMO and a method that most woodworkers can use with measuring tools they already have. I'll go with the supposition that (a) the blade is set dead parallel to the miter slots and remains that way and (b) the sled fence is dead flat and can't move and (c) the miter runner is locked in solid and can't move and is consistently tight through the entire cut in the slot.
Let me back up a bit... In mechanical design, we use a principle called exact constraints. What this basically means is that you control a design by making sure that no components are over or under constrained. Applying this to a table saw sled, you have 6 degrees of freedom to consider:
- moving vertically (up-down)
- moving laterally (side-to-side)
- moving longitudinally (front-to-back)
- rotation around lateral axis (back or front lifts)
- rotation about longitudinal axis (left or right lifts)
- rotation about vertical axis (spins l/r on the table)
Exact constraints tell us that we must constrain (prevent) motion in any of these that we don't want the sled to be able to move in. So for each DOF (degree of freedom) we need a design control to either (a) restrict it to a desired range (b) allow unrestricted motion or (c) eliminate the possibility of motion.
- moving vertically (up-down) Generally gravity will take care of this unless the sled is so light that the blade is able to lift it in which case we have to find another method.
- moving laterally (side-to-side) For a sled this is generally done by using a miter bar engaged in the slot in the table but could also be done by straddling the fence or by something more elaborate like a sliding table found on some saws.
- moving longitudinally (front-to-back) We obviously need to allow motion in this direction unless our blade can move through the work instead of the work moving through the blade. For a table saw, we'll assume we want full free motion in this direction.
- rotation around lateral axis (back or front lifts) Provided our sled is fully supported by the table and #1 is fixed, the table and gravity will take care of this.
- rotation about longitudinal axis (left or right lifts) Same as #4.
- rotation about vertical axis (spins l/r on the table) The miter slider engaged with the miter slot will constrain this DOF in theory but there is more to consider.
So, for our sled, we want the only motion to be unrestricted freedom to move front-back through the cut. We want all other degrees of freedom eliminated.
Let's assume our sled is heavy enough that it's unable to lift off the table. So the weight of our sled constrains #1.
#2 is a little trickier. Our miter slider has to stay engaged with the slot through the full cut for us to get the results we want. No problem keeping the bar IN the slot right? As long as the bar is at least as long as the sled, we're good. BUT, from experience I can say that it may not be that simple. My miter slot on the right side of my table is pretty worn and measures several thousandths of an inch wider in the center than it does at the front and back of the table. So in order for my slider to constrain the sled against motion, it not only has to be IN the slot but it has to engage the slot fully through the full range of motion. So on my saw, that can't be done with a simple bar (I'm not even going to get into sliding fit tolerances in this particular diatribe) because if I use a bar that can slide freely through the infeed and outfeed sides at the front and back of the table, it will be "sloppy" in the center of the slot and the sled could shift while the work is in the cut. So what's a girl to do? There are some options:
Let's assume our sled is heavy enough that it's unable to lift off the table. So the weight of our sled constrains #1.
#2 is a little trickier. Our miter slider has to stay engaged with the slot through the full cut for us to get the results we want. No problem keeping the bar IN the slot right? As long as the bar is at least as long as the sled, we're good. BUT, from experience I can say that it may not be that simple. My miter slot on the right side of my table is pretty worn and measures several thousandths of an inch wider in the center than it does at the front and back of the table. So in order for my slider to constrain the sled against motion, it not only has to be IN the slot but it has to engage the slot fully through the full range of motion. So on my saw, that can't be done with a simple bar (I'm not even going to get into sliding fit tolerances in this particular diatribe) because if I use a bar that can slide freely through the infeed and outfeed sides at the front and back of the table, it will be "sloppy" in the center of the slot and the sled could shift while the work is in the cut. So what's a girl to do? There are some options:
- Use a miter slider like George mentioned previously with UHMW washers that expand is my personal choice. These elastic washers allow me to tighten screws to expand the washers to engage my slot fully in the wider center section but are elastic enough that I can push them through the front and rear of the slot with some resistance. "Some" resistance means that I haven't technically constrained the sled against #2 but I can still get the range of motion that I need and I can accomplish it with a readily available product.
- If we have a slot that is wider in some parts than others, eliminating #2 could be as simple as using the other slot (assuming you have one)
- But what if you do have two slots and they both have inconsistent widths? In this case you have an option of seeing if one side of your slot(s) are flat along the full length of the table. Let's say either the inside or outside edges flat on both slots. The you can use two miter sliders like pottz used above. If the inside slots are flat, you use bars that are narrower than the "tightest" sections of your slots and engage only one side, the flat one. So if your inside edges of both slots are flat, your two sliders will engage ONLY those edges. And it must be both inside or both outside edges otherwise you can still have motion side-to-side. And note that the edges must not only be flat but must also be perfectly parallel to each other otherwise you will either bind up or get sloppy at some point.
- "But Kenny, I have two un-worn miter slots that are perfectly parallel to each other so I should use two sliders that fully engage the slots, right?" WRONG! This is something I often see on sleds that makes me cringe - two adjustable miter sliders. We are looking for EXACT constraints. If you have one perfect miter slot and one slider that engages it fully (assuming you attached it rigidly to your sled) you have completely eliminated the possibility of side-to-side motion of the sled on the table. So when folks add a second slider, they see added security. I see OVER constraint. Each degree of freedom must be EXACTLY constrained otherwise our "added security" is a liability. We now unnecessarily designed in the possibility of other problems like binding to occur. If you have two slider that fully engage both slots, what happens when cold weather causes your table to contract (yes, cast iron most definitely expands and contracts with temperature)? Yes, the possibility that it will be cold enough in your shop for the table to contract and cause binding of the sliders is remote and yes, you can probably get by with two sliders with no problems. But WHY when only one will eliminate the freedom of movement you're trying to eliminate? EXACT CONSTRAINTS. Anything less is under constrained and anything more is over constrained even if it seems like a good idea.
- And there are other options. But I think I've said enough on this particular issue. If you do have the problem with sloppy miter slots though, feel free to PM me some specifics. I'd be happy to explore the design with you.
#3, as we said, is a DOF we want to retain. We want our sled to slide front to back freely and unconstrained.
#4 is in most cases constrained by the table and gravity like #1. Special cases may exist for very large sleds that aren't fully supported by the table during the full cut and can "tip" off the table either at the infeed or outfeed of the cut. Most of us would solve that by using either temporary or permanent infeed/outfeed additions to the saw. As far as the sled itself goes, most of us have "T" miter slots with wider sections at the bottom. Adding washers to the bottom of the slider that will engage the bottom portion of the slot can prevent rotation in this direction. But they should only engage the top/bottom of the wider part of the miter slot. If they engage the sides as well, we have committed a faux pas and over constrained #2 again!
#5 is again not a big issue in most cases unless you have a really big sled and a really small table. If your sled is fully supported by the table, gravity takes care of this.
#6 ties in with #2. Same basic resolution of the latter will prevent the former assuming your slider(s) are rigidly attached to the sled.
So there is a birdseye view of the Principle of Exact Constraints and how it can be applied to something like a table saw sled. We could dive much deeper of course and go into material selections, thermal considerations, sliding fit tolerances and if we really wanted to dig in, we would need to consider the design and tolerance of the flatness of the saw table, how the arbor is constrained to run true and stuff like surface finishes and hardness of the table and miter slots. So feel free to expound if you wish 😁
