Table saw blade cutting larger kerf

[Andrew Dedman]

I built a box joint jig and after my first test cut the joint was too loose using a CMT FTG ripping blade. Using calipers the teeth measure 0.123". Using a digital indicator I verified that the blade shows 0.001" runout. I also checked the arbor and it shows 0.001" runout there. I verified the miter slots are parallel to the blade and the jig is parallel to the miter slots/blade. The blade is also square to the table top. Trying to rule out the jiI've used both the table saw fence and a crosscut sled to cut a groove into a few different pieces of scrap and verified the blade IS still cutting a larger kerf. Roughly 0.135" to 0.140" so quite a bit more. So that rules out the jig being the sole culprit. I've tried the other blades I have resulting in a larger kerf as well. I have a SawStop PCS. How close should the kerf match the tooth width of the blade? Can anyone provide any suggestions on what may be going on here and anything else I should check on my saw? Thank you!

-Andrew

[roger wiegand]

Never thought about it, but it doesn't surprise me. When it matters, eg in cutting dados, I always just measure the actual width as that's what is important, then shim as required to get where I want to go, again measuring the actual width, not worrying about theory. I've never measured so have no idea how the kerf width compares to the offset of the two teeth that are sticking out the farthest on each side of the blade. That actually seems like a fairly difficult thing to measure without some fairly serious equipment.

Runout, slight differences in tooth offset, and just having some adjacent wood "pulled along" all occur to me as possible causes. It seems desirable that the kerf should be slightly wider than the teeth to avoid burning in cuts.

[Kevin Jenness]

It's probably coming from the blade or some schmutz between the arbor collar and blade. Clean the mating surfaces and try rotating the blade on the arbor to minimize out the wobble. Try another blade. That does seem excessive. As Roger said, work with the groove size you get.

[Rod Sheridan]

Also dynamic and static runout of blades is different.

Regards, Rod

[Roger Feeley]

Also dynamic and static runout of blades is different.

Regards, Rod


Rod,
What is the cause of dynamic runout? Are we talking about aerodynamics or resonance of some sort?

[Charles Coolidge]

A .001 inch run out measured at the arbor will produce a much larger run out at the tip of the blade so something is off with the measurements. I think the run out on my arbor is .0001 to .0002 inch.

[Lee Schierer]

A .001 inch run out measured at the arbor will produce a much larger run out at the tip of the blade so something is off with the measurements. I think the run out on my arbor is .0001 to .0002 inch.

Charles is exactly right. a 0.001 arbor run out means at least .008 on the teeth. Add the 0.001 each side to that and you could have 0.010. Your .123 blade will cut at least a 0.125 slot if everything was perfect. Add the 10 thousnadths of run out to that and you would have somewhere in the neighborhood of .135 for your slot, not counting any play in your fixture or miter slot.

To make perfect joints you need to adjust the spacing between cuts to match the width of the cuts, allowing a little clearance for ease of fit. Even though many of us have calipers and micrometers in our shops, they really aren't for wood working except for use in tuning up you machines. Woodworking is much different from metal working in that exact dimensions aren't essential. A good fit is the goal.

[Charles Coolidge]

Charles is exactly right. a 0.001 arbor run out means at least .008 on the teeth. Add the 0.001 each side to that and you could have 0.010. Your .123 blade will cut at least a 0.125 slot if everything was perfect. Add the 10 thousnadths of run out to that and you would have somewhere in the neighborhood of .135 for your slot, not counting any play in your fixture or miter slot.

To make perfect joints you need to adjust the spacing between cuts to match the width of the cuts, allowing a little clearance for ease of fit.

I will add do not assume higher end blades cough (Forest) cough are flat and true the last one I purchased had .007 inch runout. I use nothing but Ridge Carbide blades now for anything important. They don't have the fancy paint and graphics but are monster thick blades with huge carbide and extremely sharp. My Ridge Carbide Ultra 48 tooth combo just made a fool of a brand new Freud 80 tooth cross cutting 12/4 Mahogany.

[Greg Funk]

Rod,
What is the cause of dynamic runout? Are we talking about aerodynamics or resonance of some sort?I don't know if it applies to 10" saw blades but large circular saws stretch at the outer edge and will wobble around unless they are dished when at rest.

Provided the kerf is consistent it shouldn't matter for a box joint jig.

If you really wanted to check if the blade was wobbling when running you could used a stroboscope, ideally synced to the motor with the ability to offset the strobe frequency slightly. If set up correctly the blade will appear to rotate very slowly and you would be able to see any variations from planer movement. We used to used this technique for testing products on a vibration table. It allows you to slow down vibrations at a fixed frequency.

[Andrew Dedman]

Thank you everyone for your input. I checked the arbor again and didn't see any runout like I did before. My setup may not have been the best. Also, the indicator I'm using admittedly isn't anything special. I believe the arbor is good. I tried another blade I have and that one is cutting just a touch over the thickness of the teeth so it seems good. It must be blade related on the other two blades. I think it may have been as simple as that. I'll report back after I spend some more time with it.

Problem with the box joint jig is it indexes off of a 1" 8TPI threaded rod so I'm sort of stuck with that advancing the carriage by 1/8" each movement.

[Lee Schierer]

Problem with the box joint jig is it indexes off of a 1" 8TPI threaded rod so I'm sort of stuck with that advancing the carriage by 1/8" each movement.

You could go 1/2 or 1/4 turns to get 15ths or 32nds.

[Mark e Kessler]

I have improved runout conditions in the past by rotating the flange some retighten, measure, repeat as needed similar to what one would do on a drill press spindle. Could indicate a few things if it improves. Also if its your blade a good sharpening outfit can hammer your blade to get it more true.

[Rich Engelhardt]

My Freud box joint blade was cutting too wide a kerf - both 1/4" and 3/8".
Ended up being caused by me cranking down on the arbor nut too much.
That must have bowed the blade enough to push the outside apart and make the kerf wider.

When I backed off and just snugged the arbor nut, I could cut perfect 1/4" and 3/8" wide kerfs.

[Mark e Kessler]

My Freud box joint blade was cutting too wide a kerf - both 1/4" and 3/8".
Ended up being caused by me cranking down on the arbor nut too much.
That must have bowed the blade enough to push the outside apart and make the kerf wider.

When I backed off and just snugged the arbor nut, I could cut perfect 1/4" and 3/8" wide kerfs.


Cranking down on it should not warp the blade, i would check to see if your flanges are flat and if something is in the way.

[Rich Engelhardt]

Cranking down on it should not warp the blade, i would check to see if your flanges are flat and if something is in the way.Thanks - but - just snugging the nut works so I'll just stick with that.

[Rod Sheridan]

Rod,
What is the cause of dynamic runout? Are we talking about aerodynamics or resonance of some sort?

Normally a harmonic in the cutter at certain speeds.

[Charles Coolidge]

I have improved runout conditions in the past by rotating the flange some retighten, measure, repeat as needed similar to what one would do on a drill press spindle. Could indicate a few things if it improves. Also if its your blade a good sharpening outfit can hammer your blade to get it more true.

^^^ this. Plus clean the arbor and blade carefully. I bought a precision chuck for a $2k Powermatic drill press. First install the run out was pretty disappointing. Removed, cleaned, rotated 180 and the run out was almost 0.0000

[Jeff Bartley]

You can measure runout at the arbor, mark that spot, then measure runout with a blade installed. Sometimes the blade will have some runout and by rotating the blade about the arbor you can cancel out the runout on both.