4 cuts accuracy test

[levib1]

Working with my crosscut table and coming up with the following numbers. Please let me know if I'm doing this incorrectly...

First cut only to get reference edge.

Rotate/cut piece 4 times so the final cut is the reference edge.

Measure cutoff piece from final cut and get 12.01 and 10.84 over 610mm distance.

Is the equation 12.01-10.84 = 1.17

1.17 / 610 = .001918

And finally, .001918 / 4 = .00048.

Not sure how if this is degrees or what the error measurement units are?

Had watched some videos, but want to verify if I am doing correctly.

Thanks

 

[levib1]

Adding to my post, diagonals with tape measure measure 801 and I'd guess 801.5.

I am prepping shop for a kitchen cabinet build and just want to make sure everything dialed in...

 

[woodferret]

Your .00048 means that's the mm/mm deviation per cut.  So a 1m length cut would be 0.00048*1000mm = 0.48mm deviation +-.  If you stick to using the front as your reference edge when breaking down goods, then that's the deviation you get.  If you round-robin, you'll be X times each new edge you put up against the fence.

 

[usernumber1]

i thought it was 5 cuts
https://windridgewoodcrafts.com/utilities/five_cut/

 

[bertv]

i thought it was 5 cuts

https://windridgewoodcrafts.com/utilities/five_cut/

The difference may be that the fifth cut (parallel to the fourth) is to obtain a piece that is about an inch wide, to be measured with a caliber. Please correct me if wrong!

 

[tsmi243]

Yeah, the 4 vs 5 thing is mostly semantics.  The 1st cut is to make a clean edge to start from- then 4 more cuts.  If you start with a rough edge, then what Bert said is correct.  Either way gets you the same numbers, so it doesn't matter which way you think of it.

It's really the "4 angle" test.  Since the angles are what's between the cuts, you have to make 5 cuts to get 4 angles.  Unless you trust your starting edge, in which case you can really just do 4 cuts and call it good. 

1.17 / 610 = .001918 

 

[Michael Kellough]

That’s not how I understand this.

The point is not to find the angle in degrees but to find out how much to move the fence to set it exactly 90* to the saw path.

In the last step you cut off a bit of the first side. If your fence is set to true 90* then both ends of the cutoff will measure the same. In the real world, less than .5mm difference is very good. You can get even better.

The purpose for making four cuts (after straightening the first side) is to amplify the possible error.
You could just make one cutoff if you started with a know perfectly square panel, and measure that difference,
but it would probably be too small a difference to easily remedy, unless your fence was way off.

The difference in widths of the ends of the four cut cutoff divided by four is how much you need to move the fence,
at the point equidistant to the length of the cutoff from the kerf. Just put the cutoff against the fence and even with the kerf and mark the fence at the end of the cutoff.

Simplified example. If the cutoff is 4mm wider at one end than the other and is 500mm long then you need to move the fence 1mm at 500mm from the kerf. No degrees or advanced math required.

Clamp a stop against whichever side of the fence needs to move away and then move the fence. Put the 1mm shim between the fence and the stop and move the fence back to the stop and tighten. Do the 4 cut test again and you should have less than .5mm difference between the ends of the new cutoff.

If the near end of the offcut is larger than the far end the fence should be moved away by 1/4 of the difference.

Your piece was 12.01 on one end and 10.84 on the other. The 1.17mm difference / 4 = .2925. That is the thickness of shim needed to correct your fence position. In cad this looks like .03 degrees. Not too bad.

 

[tsmi243]

That’s not how I understand this.

The point is not to find the angle in degrees but to find out how much to move the fence to set it exactly 90* to the saw path.

Good point.  I was focusing more on determining IF an adjustment was needed.  At some point, you gotta stop dialing and move on.

 

[raylaray]

Michael has the right approach. I recently did the same on my guide rail ( moving one end of the rail rather than the back fence, but it's the same idea ) - 5 cuts, measure thickness at front and back of the offcut, clamp a block next to the rail front bracket, loosen the rail front bracket and shim that over as needed with a cheap $5 set of feeler gauges from harbor freight, tighten the bracket back down and do the set of cuts again to test my changes. It might take a couple goes around of the cuts to get things to where you want them. I nudged the front rail bracket the wrong direction on the first go and had double the error to correct. I got to .007" difference from front to back of the offcut piece, with about 30" per side of cut length, so, pretty square (about the thickness of a couple of sheets of paper over the length of the 4 cuts). I then promptly made up a giant square, about 2 1/2 feet to a side, out of plywood and trimmed one side square with that track positioning which now becomes my precision square for future calibration of the rail guide.

 

[kevinculle]

I recently did the test on my Kapex and tweaked the position of the angle setting plate.  The method I came up with was to measure the starting gap between the inside left end of the angle plate and the saw casting, where the gap was small enough to use feeler gages stacked.  I did the initial test, got the result which was off enough that adjustment was needed.  I got a ballpark measurement of the radius from the pivot to the spot I was using the feeler gages to calculate the amount of movement I needed, which was 0.005".  I eased off the 3 screws on the plate, added a 0.005" gage to my stack and reset the angle.  The follow-up test error was very close to zero.

 

[Laminator]

I used the same method as Raylaray except I clamped a dial indicator to the table and moved the front rail support and quickly got 5th cut deviation of .001/.002 on a 22" block of MDF.

 

[smorgasbord]

Note that the straightness of your fence can also come into play. That is, something that checks out fine at, say, 24", may not be as fine at 10" of the miter fence (miter saw back guide, etc.) isn't itself flat/straight. Just something to be aware of.