Need Help with Table Saw Cut Issues

[Brad Ross]

I have a strange issue with the cuts on my table saw. I have an older 3 hp Delta Unisaw. The beginning of any cut has a slight bevel leaning to the right and the back of the cut piece has a slight bevel angle to the left. It's like there is a twist in the end of a crosscut. This happens with ripping too. I have tested multiple Freud blades and get the same result with each one. I have tried full kerf and thin kerf blades. I can't figure out the problem. I have the blade set at 90 degrees and am sure it is set correctly. I use machinist squares and I have a Wixey digital angle finder to set it. My saw blade is parallel to the miter track within .0005" and has been rechecked many times with my dial indicator. The blade at 45 degrees is parallel to the miter track within the same tolerance. It was slightly off but I shimmed the front of my table and got it parallel. Runout on the arbor is .003" and around .008" at the end of the blade. There is no play in my arbor when I try to move it. I can't figure this out. The angle is not very severe but it is very easy to see when making shoulder cuts on tenons as this causes there to be gaps. Any help would be appreciated.378805378806378807378808

[Mel Fulks]

That looks like construction spruce ,and it is not as dry as cabinet wood, so it's moving some. I would not make any fence adjustments without first test cutting some good cabinet wood.

[Peter Christensen]

Check the table for twist/flatness.

[Ray Newman]

You said you have an "older" Uni-saw. How old?

Check the bearings as I had a similar issue with my Uni-saw before replacing the bearings. The below link to the Saw Center shows how to check the bearings using a shop-made stethoscope --piece of wood -- in picture 13. Crude but it works. http://www.sawcenter.com/unisaw.htm

Also check the belts condition.

[Ted Derryberry]

I agree, check the table and make tests with better material.

[Brad Ross]

I’ve tested with a few types of hardwoods. I used the 2x4 for clarity in the picture. I first noticed this when my tenons on an oak wardrobe build had curved shoulders. It may be arbor related.

[Brad Ross]

I thought this may be the case but the same thing happens when cross cutting from either side of the blade and happens with my crosscut sled that is very flat.

[Brad Ross]

Thank you. I will check the bearings. The arbor was rebuilt in 2009 but there’s always a chance it could need work again. The belts were replaced with link belts a few months ago to remove the vibration I had.

[Don Jarvie]

How does the blade look on the indicator as you move the blade around? Could be that bit of arbor runout is causing you problems and there is a slight wobble to the blade. You can mark the high points on the arbor and file them down.

[Matt Day]

You shouldn’t need v-belts on a Uni, which might tell you something. As said above, check arbor bearings, arbor flange, and table flatness.

How are you making the crosscuts? Sled, miter gauge, sliding table?

[Lee Schierer]

Thank you. I will check the bearings. The arbor was rebuilt in 2009 but there’s always a chance it could need work again. The belts were replaced with link belts a few months ago to remove the vibration I had.

Doesn't your Unisaw have 2 or 3 belts? If so I don't think link belts are generally recommended for multiple belt drives. If your saw has multiple belts, get good quality Gates belts, they are all precision made so you no longer need to buy matched belts.

When you you say your arbor has run out is that in the vertical direction or is the flange running in and out?

[Brad Ross]

Thank you. I will try some new belts. I had some vibration in the saw so I switched to link belts and they removed the vibration. The runout is in the flange running in and out. From things that I have read .003" is within the tolerance for good cut quality.

[Brad Ross]

I am making most crosscuts on a crosscut sled. I will sometimes use a miter gauge but mostly the crosscut sled. I may try buying some new belts. I had some vibration so I switched to the link belts and it removed it. My table is has a .003" slight gap in the center and on the left and right side of the table when measured with feeler gauges and a machinist straight edge. The highest spots are around the miter gauges. That seems pretty good. Is there a way to test the bearings without removing the arbor? I cant hear any issues when spinning it by hand. Arbor flange has a .003" runout. I don't feel any movement in the arbor when trying to move it around.

[Robert Engel]

Have you checked this with other blades?

Are you using thin kerf blades?

The reason I ask is I'm thinking there is something causing it when the blade first enters and exits, which in my mind makes me think that.

[Brad Ross]

Yes, I have tried with multiple thin kerf and full kerf blades and get the same results. I thought the same at first. My first thought was blade deflection when I saw what was happening but when I tried my full kerf blades I got the same result.

[michael langman]

.003 runout on the face of the flange on the spindle is too much in my opinion. It could be runout because the flange is not flat rather then not running true to the bearings and motor.
The pictures show that in what is happening to final cut.

[Alan Schwabacher]

Did you clamp the stock to the miter gauge to make sure it isn't shifting during the cut? Is the miter bar loose enough to slide but not wobble in the miter slot?

If the blade is not parallel to the miter slot, it would cut a wider kerf at the bottom than the top, but it would also push sideways on the stock, making the situation more complicated if it shifted, perhaps leading to what you see. Runout should make the kerf wider at the top. You might compare what happens when you clamp to the miter gauge fence, and make the cut guided by either the right or left miter slot.

And check how parallel the blade is to the slot by measuring to the same tooth, while rotated to either the furthest forward or back position. You don't care about the exact distance, only the difference between the two.

[Brad Ross]

I used a dial indicator to set the blade parallel to the miter slot. It is within .005". I thought slop may be the cause in the beginning but the same thing happens with my crosscut sled that has no play at all.

[Earl McLain]

I use link belts on a few machines that just weren't going to get any better through set-up--and they have done a nice job of MASKING the vibrations (think cheap band saw)--but I don't believe link belts would ever "remove" vibrations. In some tools--masking is the only choice at the end. In some tools--Unisaw--a vibration is a sign of a problem and masking it won't remove it.

Just my opinion. I apologize that it sounds harsh, but I learned that lesson.
earl

[Brad Ross]

I appreciate the feedback. The vibration I had seemed to be caused by V-Belts that were on the saw for the last 10-12 years. Once replaced with link belts the saw passes the nickel test and even a dime will sit on edge. I read a lot of mixed reviews about putting link belts on multi pulley machines but it seems to be the best thing I could have done for the saw.

[Brad Ross]

Last night I decided to remove the tilt lock handle and clean inside the crank wheel. I was able to lock it a little tighter after that. This helped correct the angle of the cuts but didn't eliminate the issue entirely. I'm going to see if I can remove the .003 flange runout next.

[Mike Hollingsworth]

You could have a Cuisinart Blade on it and your cuts would be square if your table is straight and true. Maybe nasty cuts, but still straight and true. Your table is the problem.

[andy bessette]

.003 runout on the face of the flange on the spindle is too much...


I used a dial indicator to set the blade parallel to the miter slot. It is within .005"...

These are both too much.

[Matt Day]

Regarding the belts. Try a new set of gates belts and play with the belt tension. Mine has the least vibration with the belts looser than I would have thought they should be. No issues with power or slippage either.

[Lee Schierer]

Thank you. I will try some new belts. I had some vibration in the saw so I switched to link belts and they removed the vibration. The runout is in the flange running in and out. From things that I have read .003" is within the tolerance for good cut quality.

If your arbor flange is 2" in diameter and is running out .003", your teeth tips can be running out as much as .015"

[Jack Frederick]

Assuming you have a flat table I think I would look at the arbor assembly. You say it was replaced in '09, but by whom, and how? Getting rid of the vibration with the link belts just masks the actual issue. I know, because I did the same;) I was close to the Saw Center in Springfield, MA when I was having problems with my '48 Unisaw. I discussed it with them and disassembled the saw and took the arbor assembly to them for repair/replacement. I got new V-belts, put it back together and it was just so much better and so smooth. Like a new saw, well, with the new motor and electrical.

[Brad Ross]

Typo on the miter slot. It is within .0005”. Arbor flange runout I agree is too much. Going to try to grind that out.

[Brad Ross]

I think saw center did the repair but not sure. I bought the saw last year and the previous owner gave me some receipts from the new motor he put in it and he gave me a receipt from the new starter/switch he got from them. I don’t have the receipt from the arbor rebuilt but I can see that it has the newer style bearings in it instead of the older Unisaw bearings with the extended race. I will probably go ahead and replace the bearings to be safe and I am going to get some new V-Belts as you recommend and see what happens there. Thanks for the info.