Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

This is the most versatile work station you will ever use

table saw calibration

Posts

table saw calibration

#1

table saw calibration

Ronster

I've got this new Unisaw that I'm tinkering with.

Arbor runout is 0.002" at most. That's near the end of the arbor. I guess that's ok.

Using a Freud 10" reference blade (no teeth) and measuring first from the miter slot to the front of the blade, and then rotating the blade so that the measuring spot is at the rear, and then measuring from the miter slot to the blade at the rear, I get a difference of 0.01". Therefore, over a distance of almost 10" the blade is out of parallel with the miter slot by 1/100". On the one hand that seems like a lot (~1/32" over 30"). On the other, adjusting it to the better will take a while. Am I obsessing and in need of a slap?

Ron

Seems sacrilegious to be posting so soon after losing Col. Dart.

Re: table saw calibration

#2

Re: table saw calibration

Robin Corell in Atascadero,CA

If it were me I'd crank the blade over to 45 and check it. If you get something close to what you have at 90 of call it good. If out is way out then I'd work on getting something better than .010 after shimming the trunions.

Robin

Re: table saw calibration

#3

Re: table saw calibration

Lee Schierer McKean, PA

I think I would try to get the saw closer to zero if it were mine. .01 misalignment in 6-8" that the blade sticks above the table will give you burn marks when you cut wider boards and leave noticeable teeth marks. It will also make ripping more difficult.

Lee

Re: table saw calibration

#4

Re: table saw calibration

John in NM

I typically wait until I have to start fixing the results of this kind of thing in the finished product. While 0.01" may seem like a lot, when is the last time you cut a piece larger than about 8-10" using the miter slot? I doubt you'll even notice it when cutting a bunch of 2-3" rails.

Another way to look at it, that 0.01" is probably already closer than the 90 degree detent on the average chopsaw. Just the mechanics of those spring in notch detents means if you set it, cut, then change the setting and go back to the original setting, you are likely to be off by more that 0.01" over 10".

Just sayin', don't sweat it :D

Re: table saw calibration

#5

Re: table saw calibration

Dennis E Peacock

My uncle taught me many years ago how he setup his tablesaw and kept it aligned. Now grant it, his shop was in his barn, but he made some of the finest cabinets and furniture that I ever laid eyes on.

He would get it close with a "story stick" and then he'd check to see how the saw cut....rip with the fence and then when he was happy with how it cut with the fence, he would check how it cut with the miter. Typically that's all he ever did and his stuff was awesome.!

Sometimes, I think we have strong tendencies to "over engineer or over think" our setups and pieces we make. After all, wood moves a LOT more than our too tolerances from season to season and the truth comes out in how the machine cuts the wood for you.

In short, if it cuts good and performs like you expect? It's all good. :D

Re: table saw calibration

#6

Re: table saw calibration

Dick Coers

The old Unisaw was just a matter of loosening 4 bolts that hold the table to the base, and bump it over. You didn't have to mess with the trunion at all. Is the new saw different? If it's as simple as the old machines, I would make it closer than that.

Re: table saw calibration

#7

Re: table saw calibration

Mark Mandell - Gone Round In New Jersey

Thought cab saw alignment was a matter of loosening 3 of the 4 bolts holding the top to the cabinet and tapping with a dead-blow hammer to bring the slot parallel with the blade.

My miter slot (JET cab) is dead-nuts parallel with the blade. Not that it means much since I've never used a miter gauge anyway (use sleds instead). My fence, however, is deliberately set .02 wide at the rear of the blade to help prevent kickback and binding.

Re: table saw calibration

#8

Re: table saw calibration

Rick Christopherson

This perceived problem of requiring the blade to be perfectly parallel to the miter slots is drastically over-blown. Contrary to initial intuition, it does not impact the squareness of your crosscuts. Even if they are drastically non-parallel, the crosscuts will still be square, assuming your miter gauge is square. You could literally have your miter slots and blade be several degrees out of parallel, and although you will have a very rough cut, it will still be square to the fence of the miter gauge.

Even in terms of cut quality, you have to be drastically out of whack before they become noticeable. If the workpiece touches the blade plate, it is probably because the plate is wobbling, not because the workpiece is traveling skew to the blade.

Re: table saw calibration

#9

There are two planes to consider

Lee Schierer McKean, PA

Actually it does make a difference. While you might get a square cut in one plane, it won't be square in the plane perpendicular to the table surface. Try this test: If you hold a blade up in front of your eye and look straight onto it you will see a straight line. However as you rotate the blade left or right you will begin to see a curve, which is what the cut edge of your board will look like if the blade isn't parallel to the miter slot. This curvature can throw off your joints so they aren't square.

Alignment of most saws is relatively easy to do and well worth the time and effort. A cheap dial indicator from Harbor freight can get you as close as you need to be. Attach the dial indicator to a piece of wood with a wood screw and clamp the wood to the miter gauge and you are set to make adjustments. Once you get the blade aligned to the miter slot, you can align the fence to the same slot. Close alignment eliminates burn marks when you push a piece of wood through the cut past the back teeth. It will also eliminate teeth marks on your board edges when you make rip cuts and dramatically reduces the tendency for a kick back on rip cuts.

Lee

Re: table saw calibration

#10

Re: There are two planes to consider

Rick Christopherson

Lee,

I was well aware of the geometry change on the angle, but if you do the math, you will see just how trivial your argument is. To accurately go through this, I created a SolidWorks model representing a 10" circular saw blade crosscutting a 3/4" workpiece, at typical blade height, and where the front-to-rear blade skew was 0.010".

The difference in length of the cut at the top and bottom of the workpiece is 0.0011" (0.001"), and the angle is 0.09 degrees off perpendicular. The curve of the cut is so trivial that I can't get this $8000 cad system to even measure the radius.

Just blade wobble alone is going to far exceed these numbers. That's not to mention, that I guarantee that your blade bevel setting is not accurate to within 0.09 degrees from perpendicular. Simply turning on the saw will knock this calibration out of alignment due to gear lash in the worm gear.

The principle of this need for parallelism is blown out of proportion, and the people that keep harping on it are simply repeating what they've heard elsewhere without critically performing the mathematical analysis of it.

As you read through the majority of internet postings on this topic (not just this forum), you will see that most people assume that this is a critical adjustment because it impacts the square of the cut relative to the miter fence. This is why so many people mistakenly harp on the topic.

It doesn't impact squareness to the miter fence. It impacts squareness of the "bevel", but it is so trivial of an impact that it is not worth messing with until it becomes visually noticeable when looking down on the blade.


Re: table saw calibration

#11

New unisaw is the same in this way


Re: table saw calibration

#12

Interesting discussion!  Glad I asked.


Re: table saw calibration

#13

Re: There are two planes to consider

Lars Parrington

fyi, the reason solidworks wouldn't measure a radius is because the curve you describe is part of an ellipse, i.e. not a circular arc.

Re: table saw calibration

#14

Re: There are two planes to consider

Rick Christopherson

fyi, the reason solidworks wouldn't measure a radius is because the curve you describe is part of an ellipse, i.e. not a circular arc.


I wasn't trying to measure the curve directly because I already knew it was not circular. I was trying to emulate the curve using a 3-point arc, but the points are so close to being linear that SolidWorks assumed I was trying to do something different, and would error-out.

Nevertheless, the point is that the cut is so close to being true that it is well beyond worrying about.