Question about 5 cut method for RAS

[Louis Brandt]

Hello,

I've read about the so-called 5 cut method for truing up a table saw. As far as you know, would this same method work for a RAS, and if so, are there any differences that would be required to use it for a RAS?

Thanks,
Louis

[Rod Sheridan]

Yes it would work for a radial arm saw, you make a cut, rotate the wood so that the cut edge is now against the fence, and repeat until you've made 4 cuts that take off about 1/4" of material.

Now make the fifth cut placing the freshly cut edge against the fence and cut a about a 1" strip off.

Measure both ends of the strip for width with a pair of calipers or a micrometer, if it'

they're equal you're square.

Regards, Rod.

[Tim Sgrazzutti]

There is a much simpler way for adjusting a RAS square. Take a wide board (8 or 10 inches) with a jointed edge. Place the jointed edge against the fence to the left of the blade, and make a cut off the end of it. Next, flip the board over like you're turning the page of a book, so that the jointed edge is against the fence to the right of the blade. Make your cut, and check the thickness of your cutoff piece at the front and back. The cut is out of square by one half of the thickness difference over the length of the cut.

[Louis Brandt]

Thanks to all. I think that my RAS is pretty close to true, but I'd like to check it with the above methods. I only use the RAS for 90 degree crosscuts or dados, so I want to keep it accurate.

Louis

[Steve Rowe]

Tim is exactly correct and is the method I use for my chop saw. I didn't even know that a RAS could cut square (at least based on the saw I had).

[Louis Brandt]

Yes, my early 1960s 10 inch Craftsman RAS cuts pretty darned square. I just wanted to do a test to get it as perfect as possible.

[Peter Quinn]

I like Tim's method for short stroke radial saw. The 5 cut method multiplies the error by 4, thus making it easier to read small errors on a panel saw. Perhaps it will lead to greater accuracy? I would think the amount of error that would cause a problem over the typical stroke of a RAS would cause a comparatively greater problem over the 8'-10' stroke of a large panel saw. So for a panel saw the 5 cut method may be preferred, for a miter or RAS with a stroke typically 24" or less, the cut and flip method should suffice. I often just use the "make one cut. put a square on it, bump the saw until it squares up" method for the miter and RAS, but when squaring a panel saw that simply isn't good enough.

[Rick Christopherson]

I like Tim's method for short stroke radial saw. The 5 cut method multiplies the error by 4, thus making it easier to read small errors on a panel saw. Perhaps it will lead to greater accuracy? I would think the amount of error that would cause a problem over the typical stroke of a RAS would cause a comparatively greater problem over the 8'-10' stroke of a large panel saw. So for a panel saw the 5 cut method may be preferred, for a miter or RAS with a stroke typically 24" or less, the cut and flip method should suffice. I often just use the "make one cut. put a square on it, bump the saw until it squares up" method for the miter and RAS, but when squaring a panel saw that simply isn't good enough.Actually you have this backward. The shorter the cut, the greater the need to compound the error to accurately measure it. On an 8-foot cut, there is little need to compound the error because the real error is quite easy to measure.

Back to the original poster, I have instructions on how to perform this adjustment, including a built-in calculator for the compounded error. It was written for the Festool Kapex manual, but I extracted the necessary pages because they apply to any similar tool, including your RAS.

Oh, by the way, the need for the 5th cut is a math error by some woodworking magazine author, but the name sticks. It really takes only 4 cuts.
http://www.waterfront-woods.com/festool/4-cut_pages.pdf

[Myk Rian]

Oh, by the way, the need for the 5th cut is a math error by some woodworking magazine author, but the name sticks. It really takes only 4 cuts.
No way dude! Rag editors don't make mistakes.

[Louis Brandt]

Let me follow up my original question by asking this, and I don't mean to imply that the answers that I've already been given aren't good ones, but since I'm still pretty much of a novice woodworker, I need to learn.

Why wouldn't the following method work just as well as the (four cut or five cut) method for truing up a RAS? Just put a known perfect edge against the fence, make a 90 degree crosscut, then check the cut edge with a square, and if necessary, make adjustments to the arm until the square revealed a perfect result.

Louis

[Chris Padilla]

Louis,

That works...but how good is your square? The reason for the above 'tricks' is to MULTIPLY or INCREASE the error so that it is more discernable.

[David Winer]

Just put a known perfect edge against the fence, make a 90 degree crosscut, then check the cut edge with a square, and if necessary, make adjustments to the arm until the square revealed a perfect result.
Louis
A variant of this method I have always used does not involve cutting: Place a rafter square against the fence, then with a tooth in contact with the square, pull the blade across the table. The tooth should make contact all the way; if not, adjust the saw. Obviously, this is only as accurate for squareness as the rafter square is.

[Rick Christopherson]

The difference between the methods is simply a matter of the "resolution" of the measurement. Using a square against the blade provides a 1x resolution. Cutting a board and flipping one of the pieces (Double-Error Squaring (http://www.waterfront-woods.com/Articles/Double-Error-Squaring.html)) provides a 2x resolution. The 4-cut method provides a 4x resolution.

The higher the resolution, the closer you can confirm your setting to the desired position. It should be pointed out that none of these methods actually make any adjustments. They simply show how far out of square the actual cut will be.

Choosing which method you use is a matter of how accurate you need your cuts to be. This will vary depending on the type of work you are doing.

I do a lot of precision woodworking, so I never check square by using just a square. You can have a cutting error that is small enough to not be detectable with a square, but when you magnify that tiny error for the length of a cabinet faceframe, it becomes more significant.

On the far end of the extreme, when I created an 8-segment circle, I compounded the error 16-times to adjust my tablesaw sliding table. I needed that level of resolution to join the components without a gap. This level of precision would be pointless on a miter saw because there is too much variability from cut to cut. My sliding table is the only tool that can be locked down to this level of precision with repeatability.

[Louis Brandt]

I understand what you're saying, that the more "complex" the test (4 cut versus 1 cut, for example), the more likely the error will show up. But that also means that you really need to have a quality measuring device to measure the variance in something like the 4 cut method.

[Alan Schwabacher]

No, it means that if you use the 5 (4?) cut method, your measuring device needs to be only 1/2 as accurate, or alternatively the test piece 1/2 as big, to be able to measure the same error.

It's absolutely true that you need only 4 cuts to do the 5-cut test. Of course you need a straight cut edge to start, and the question is whether you consider that the first cut or not. The reason not to is that this test is commonly repeated several times, and thinking of it as a 5-cut test tends to make people unnecessarily repeat that preparatory cut between tries where it's clearly not needed.