Table Saw Blade Runout

[Jay Knobbe]

Looking for some feedback on a blade runout concern.
The saw is a Dewalt 7491. I have 3 blades that all have .006-.008 runout measured with my dial indicator following common measuring practice. The blades are a Freud, a new CMT ITK Extreme and a new Ridge 80tooth. Rotating the blades 90 degrees, then 180 degrees on the shaft and measuring gets similar results.
When I measured the machined flange on the motor I have no more than .0002. The needle barely wiggles while rotating, probably due to the torque applied to rotate the shaft.
Basic math says if I have .0002 TIR, 1" from the center I should have no more than .002 @ 10" out from the center.
I attached a pic of my dial indicator set up. The dial shaft isn't exactly perpendicular to the flange in the pic but it is when I take the actual measurements.
BTW, I sent the Ridge blade back and they tested it finding only .001 TIR. They sent a print out and test cuts they did. Very prompt and professional.
Thoughts on what might be causing this much runout?

[David Buchhauser]

I can't tell from your photo exactly what you are measuring. Your assumption is that the saw blade when mounted is perfectly parallel to the mounting flange face. This may not be true, and it doesn't take much to magnify any slight deviation from parallel at the flange out to the blade circumference 10" away. Possible causes might include contamination, burr on the flange or washer, etc.

David

[Andrew Hughes]

My guess is cheap Chinese bearings on the saws arbor.
They are probably about the same size as ones in a Dewalt router.
Good Luck

[David Buchhauser]

My guess is cheap Chinese bearings on the saws arbor.
They are probably about the same size as ones in a Dewalt router.
Good Luck

Now that I understand what he has, I would tend to agree. I wouldn't necessarily expect the build quality of the inexpensive job site saws to match that of the more expensive and possibly higher quality table saws. I think these job site saws are more intended for typical "job site" work and not necessarily "cabinet work". As I recall, I measured the blade runout on my Grizzly 1023 5hp cabinet saw and it was a few thousandths or less with a Freud blade installed.

David

503215

[Bruce Page]

Jay, I’m assuming that you are trying to measure blade wobble? If that is the case your inspection methodology is a bit wonky, or, I might not be seeing your picture correctly. The best way to check wobble is with the blade raised and locked and the travel indicator perpendicular to the blade. I’m using a test indicator in my picture, but the principle is the same.

[Bill Dufour]

I agree with Bruce. I think he may be measuring axial runout on the edge of the flange. A meaningless number on a probably stamped surface. He says using normal measuring methods? A 1 or 2 deviation needle move is not flutter. That is a measurable movement. Flutter is the needle moves just enough to see it is not perfectly still. less then 1/2 needle width total motion. Good eyes, good location, good lighting you should be able to estimate 1/10 of a deviation
Bill D

[David Buchhauser]

I agree - Bruce shows a correct method to measure blade tir (total indicated run out). That's the set up I use. The OP indicated (I think) that he did use this method to test his blades and then sent one them back to the seller. So the photo he posted above makes it unclear exactly what he is measuring.

David

[Jay Knobbe]

First, thanks for all the replies. Now I'll add more info to clarify.
1. I am measuring blade runout similar to Bruce's set up. Placing the dial indicator plunger just below the teeth in a location without any deformations, etc. Of course my table surface is non-magnetic.
2. The flange that the blade is clamped against in my picture is a machine ground flange and has no movement up/down or in/out. I am measuring the machined face of the flange where the blade contacts. After discovering the blade runout I cleaned everything, used a small deburring stone on the flange to ensure any possible burrs were dressed down, then measured again with the same results. I'm guessing the motor shaft/arbor is mounted in tapered roller bearings since there was no movement of the shaft whatsoever, but I don't know this for certain.
3. When I bought the Ridge blade I purchased 1 blade stiffener from them. It slides on between the blade and the washer and nut. I deburred that also. It provides no improvement in blade runout. FWIW I also can't detect any improvement in cut quality with the stuff I normally cut like birch ply, 1' hardwoods, etc.
Background: I bought the saw from my buddy who used it only to breakdown sheet goods. I knew the history and care of the saw. Since this is a non-money making hobby for me I spent my time "tuning" the saw and getting it as dialed in as possible, building a longer, taller fence and a table to support longer pieces safely.
Early on in my work career I spent 10yrs working in an automotive machine shop so using measuring devices, working consistently, and tracking down the slightest variance is a benefit and curse all at the same time.
If my runout was .003-.004 I'd just call it good considering the saw but I think .006-.008 is too much.
Could I be over tightening the flange nut? I'm snugging it up pretty snug.
Could the washer that contacts the blade be tweaked somehow? There are no burrs on that either.
Thanks for your help.

[mike calabrese]

You are going to get far more runout than you are measuring when the saw is actually running and adding all kinds of excitement energy to the rotating element and the rest of the saw base.
If you want to eliminate some of the blade wobble simply cut some paper shim strips install them as indicated between the blade saw hub and remeasure.
At the end of the day this may make you feel better about your static runout condition but ultimately not do much to improve the cut.
There are a million other factors that will come into play especially with a portable "job site saw".
Try this simply raise your blade place a screwdriver on the saw table about 6 inches from the blade. Be sure to have your saw fence between the blade and screw driver so the screwdriver could never get to the bade.
Start the saw allow it to come up to speed and then shut it down. IF the screwdriver dances around you bigger issue is not " If my runout was .003-.004 I'd just call it good considering the saw but I think .006-.008 is too much."
If the screwdriver makes you nervous try the same test with a nickel standing on end.................................
calabrese55

[Bruce Mack]

[QUOTE=Jay Knobbe;3262107]Looking for some feedback on a blade runout concern.
The saw is a Dewalt 7491. I have 3 blades that all have .006-.008 runout measured with my dial indicator following common measuring practice. The blades are a Freud, a new CMT ITK Extreme and a new Ridge 80tooth. Rotating the blades 90 degrees, then 180 degrees on the shaft and measuring gets similar results.

Jay- Perhaps you could do a test crosscut and rip. If you are satisfied with 90° on the crosscut and a fairly smooth surface vertical to the board and a similarly acceptable surface on the rip you can work with the saw. If later you are unhappy with the results and your ability to improve them, you could get a Sawstop contractor saw. I have a Ridgid that works well and I don't do enough to warrant a change at this time in my life.

[Robert Hazelwood]

Looks like you are measuring the flange runout correctly and it seems pretty good. That doesn't look like a tenths indicator so I would round to .0005 if the needle just barely moves.

I don't think measuring the blade is as meaningful because it won't necessarily be the same at rest as it is while spinning at 3000+ rpm. All you can really do is compare the theoretical kerf with the actual kerf, and subtract from that the runout you think is attributable to the flange (maybe 2.5 thou in your case). The remainder is the effective runout, which could be an issue with the blade itself or something dynamic in the setup (probably a lack of rigidity somewhere). You have several blades from quality manufacturers so we can assume that they are not all out of whack. If you see a consistent effective runout on all of these blades beyond what the arbor flange TIR will cause, then it must be something in the saw.

How are your actual cuts? I assume everything else is well aligned. It's a jobsite saw so I think you will reach diminishing returns pretty rapidly with fine tuning everything.

[Edward Weber]

How are your actual cuts? I assume everything else is well aligned. It's a jobsite saw so I think you will reach diminishing returns pretty rapidly with fine tuning everything.

This is what I was going to ask.
What prompted you to start measuring, is there a suspicion that something is wrong.
BTW, using a blade is not he best way to check. A calibration disc is far more accurate.

[John TenEyck]

I really doubt the arbor bearings are anything other than standard ball bearings. That's what's on my Unisaw where I measured about 0.002 - 0.003" on a 10" blade - after I cleaned the arbor flange and made sure the blade was dust free. Before I did that, I got values as high as 0.005 or 0.006" IIRC. FYI, the position of the blade on the arbor definitely influenced the values I recorded, too, even after everything was clean. It would go as high as 0.005" and as low as 0.002" depending upon where I put the blade on the arbor and was basically consistent from blade to blade. I suspect it was due to tolerance stack up of the arbor and blade, but don't know for sure.

FWIW, I don't recall seeing any difference in cut quality from the highest to lowest runout values. Blade design and manufacturing had a much greater impact.

John

[Jay Knobbe]

Thanks for all the thoughtful replies. Once I had the saw I wanted to make an outfeed table, a tall fence and crosscut sled. Before I started any attempt at accurate cuts I began by measuring everything and making my adjustments where everything worked out well except the blade runout. My cuts are ok and I figured any issues were more likely due to operator error. I liked the idea of reducing errors where I could but I suspect I'll need to live with what I have and just find ways to compensate as needed.
I am grateful for all your guidance.

[Bill Dufour]

I am just finishing moving my tablesaw from my old house. I had to remove the top. I could not get the mitre slot less then 10/1000 parallel to the blade. after removing the top and enlarging bolt holes.. grinding the bolt necks down etc I got it down under 1/1000". Turns out one corner of the top was touching the lathe and would not shift that last 10/1000. All I had to do was pull the saw out away 1/4" and it lined up fine.
The top alone weighs in around 160 pounds so hard to tell what is moving or not. I ended up using my lightest dead blow hammer for the final adjustments.
Now I get to add the 60 pound side tables and line them up with a DTI.
BilL D.

[Patty Hann]

I am just finishing moving my tablesaw from my old house. I had to remove the top. I could not get the mitre slot less then 10/1000 parallel to the blade. after removing the top and enlarging bolt holes.. grinding the bolt necks down etc I got it down under 1/1000". Turns out one corner of the top was touching the lathe and would not shift that last 10/1000. All I had to do was pull the saw out away 1/4" and it lined up fine.
The top alone weighs in around 160 pounds so hard to tell what is moving or not. I ended up using my lightest dead blow hammer for the final adjustments.
Now I get to add the 60 pound side tables and line them up with a DTI.
BilL D.

What does DTI stand for? Digital-something-indicator?
(It is not in the list of acronyms that someone very thoughtfully posted elsewhere). Thanks

[Bruce Page]

I'm assuming Digital Test Indicator?
Acronym fail? :rolleyes::)

[glenn bradley]

I would hesitate to get too far down the rabbit hole on a job site saw. These are built for a purpose and their function focus is specific to their use model. I believe this is one of the better job site saws. If you are doing job site work and find the tolerances to be too loose you could move to a CMS for those operations. A CMS is also pretty loose when compared to some sort of cabinet saw but on a job site can miter and cross cut within the required tolerances. Trying to force a job site saw or a contractor saw to be a cabinet saw will only bring you frustration; I have been down this road. I would say your saw is operating decently for lopping off 2x4's and ripping siding as it is intended to do. Are you doing something else?

[Patty Hann]

I'm assuming Digital Test Indicator?
Acronym fail? :rolleyes::)
If it is, it sounds rather hokey to me :rolleyes:

[Bruce Page]

Bill will probably elaborate.

I was reamed years ago buy a member for using a not too common acronym without explaining the definition. The member was quite irate with me, lol. (laugh out loud) :eek:

[Bill Dufour]

DTI: Dial test indicator. Simple tool, no mount needed. Not to be confused with a dial indicator which requires a mount of some kind. Used to check alignments. Flip the little probe down and it measures how far up and down it moves as you slide the tool across a joint. It just lays flat on its back.
Claim is a flat surface mismatched by 1/1,000 of an inch is visible to the human eye. I doubt you could feel that.
bilL D

[Bruce Page]

Bill, you are right, of course. I've always called them just test indicator and the one Jay is using, a travel indicator. I have a several of each.

[Patty Hann]

Bill, you are right, of course. I've always called them just test indicator and the one Jay is using, a travel indicator. I have a several of each.

You obviously stumbled upon a "Dial Test Indicator" garage sale. :D

@Bill...thanks for the explanation.

I picked up this one a while back.
Having worked 25 years in a metrology lab (aka calibration), I'm very partial to Mitutoyo, then [US made] Starrett for linear measurement tools.


503341

[andy bessette]

...When I bought the Ridge blade I purchased 1 blade stiffener...

Besides the cheap saw, this is likely your problem--a blade stiffener on one side and a different size on the other.

[Bruce Page]

You obviously stumbled upon a "Dial Test Indicator" garage sale. :D

Lol, I used to buy a lot of my machinist tools from retiring machinists. Now I’m looking at downsizing my own tools. :(

[Andrew Hughes]

I’ve posted this before but I’m so proud of it let’s see it again. :)
My Federal indicator I paid 40$ Craigslist. Seemed to be new or hardly ever used
My favorite indicator for setting jointer knives. I also have a shars.
Good Luck

[Bill Dufour]

I’ve posted this before but I’m so proud of it let’s see it again. :)
My Federal indicator I paid 40$ Craigslist. Seemed to be new or hardly ever used
My favorite indicator for setting jointer knives. I also have a shars.
Good Luck
I have something like that but I have no idea where it is in the shop.
BillD.