Another runout question

[Walt Stevens]

Dear all,

I'm just getting started in woodworking and am in the process of setting up a new Grizzly G1023SL table saw with a Forrest Woodworker II blade. I'd like to get the setup as accurate as possible. I measured the runout of the arbor flange, and it was less than 0.001 inches. I mounted the blade, and checked the runout at the perimeter, just inside the teeth, by rotating the blade through 360 degrees. The blade runout is 0.007. Is this reasonable? Have I missed something?

Everything else looks good. I had no trouble getting a Masterplate parallel to the slot to within 0.002. Setup blade angles are dead on 90.0 and 45.0.

This is a great forum and I'm learning a lot. Thanks to all of you experts for taking the time to share your knowledge and experience.

[David DeCristoforo]

Actually, it's very difficult to gauge runout on a blade that is not running at speed. Most blades will have some degree of "warp" which will flatten out at speed. Don't get too hung up on a few thousandths of an inch...wood is not that critical.

YM

[Eddie Darby]

When you measure the run-out at the edge of the saw blade, you are running the risk of having the error in the blade, and the error in the arbor coincide with each other to magnify the mistake.
Instead locate the place on the arbor where it is out, and then try locking the blade in at different places until you find the least run-out. Then mark the blade to coincide with the arbor.

Lee Valley use to sell a giant washer that had 'set screws' located in it around the outer perimeter, which allowed the user to deflect the blade to be able to zero it out. Wonder what happened to it?

[Walt Stevens]

Actually, it's very difficult to gauge runout on a blade that is not running at speed. Most blades will have some degree of "warp" which will flatten out at speed. Don't get too hung up on a few thousandths of an inch...wood is not that critical.

YM


I've been reading Kelly Mehler's table saw book, and he suggests 0.006 as the maximum blade runout. Forrest claims their blades have a maximum of 0.001 "wobble". I've sent them an e-mail, but they haven't responded yet. I guess it's easy to get hung up on specs, especially when you don't have the experience to know when they really matter.

[James Suzda]

I know this is sort of basic, but check to make sure there is no "crud" on the saw blade or the arbor back-plate that could throw the blade off. I know you said you had no problem with a masterplate, but run your finger along the arbor to make sure there are no little burrs that could be the problem.
One time I put on a blade and I thought I had a wobble blade on! I checked and there was a small piece of sawdust squashed against the backing plate.
If no problems found I would check with Forrest to see what their "specs" are on the blade? Maybe it is out of tolerances and they will replace it.

[Walt Stevens]

I know this is sort of basic, but check to make sure there is no "crud" on the saw blade or the arbor back-plate that could throw the blade off.


Everything appears to be clean, but I will double check this morning. I sent an e-mail to Forrest and I will call them today

[Phil Thien]

Problem is, you need a better indicator to measure your arbor runout. That is because your blade will magnify it by 10x. So if you're seeing .007 at approx. 9" of the blade (measuring right below the gullets), and the blade is in fact flat to .001 (and unless you have a bad blade I trust Forrest's spec), then your arbor runout is as high as...

(.007 - .001) / 9 = .000667

To check whether your blade is flat, loosen your arbor flange nut just enough so the blade can spin on the arbor, but doesn't wobble around. Now spin the blade against the indicator and note the deviation on the scale. You'll soon know whether your blade is flat or not, and to what degree.

Finally, some will argue that .007 isn't enough to worry about. It translates to a little more than two sheets of paper of wobble. If I can see the blade wobble as it spins down, I feel there is too much and I would look for a solution. Mind you, I wouldn't stop all work until it is corrected. But I would work towards getting it lower, working in this order:

(1) Clean blade washers, nut with lacquer thinner and toothbrush.
(2) Experiment with changing orientation of washers, nut
(3) Replace washers, nut with new ones.
(4) If 1-3 fail, find a machinist with a good dial indicator that can read into the "tenths" (shorthand for ten-thousandths, .0001). A good Interapid or Starrett. Check the arbor with this.

[Walt Stevens]

Problem is, you need a better indicator to measure your arbor runout. That is because your blade will magnify it by 10x. So if you're seeing .007 at approx. 9" of the blade (measuring right below the gullets), and the blade is in fact flat to .001 (and unless you have a bad blade I trust Forrest's spec), then your arbor runout is as high as...

(.007 - .001) / 9 = .000667

To check whether your blade is flat, loosen your arbor flange nut just enough so the blade can spin on the arbor, but doesn't wobble around. Now spin the blade against the indicator and note the deviation on the scale. You'll soon know whether your blade is flat or not, and to what degree.

Finally, some will argue that .007 isn't enough to worry about. It translates to a little more than two sheets of paper of wobble. If I can see the blade wobble as it spins down, I feel there is too much and I would look for a solution. Mind you, I wouldn't stop all work until it is corrected. But I would work towards getting it lower, working in this order:

(1) Clean blade washers, nut with lacquer thinner and toothbrush.
(2) Experiment with changing orientation of washers, nut
(3) Replace washers, nut with new ones.
(4) If 1-3 fail, find a machinist with a good dial indicator that can read into the "tenths" (shorthand for ten-thousandths, .0001). A good Interapid or Starrett. Check the arbor with this.

Thanks for the tips. I've played around with this a little bit after cleaning everything again. I mounted the blade, measured the runout, and marked the two extremes. They are about 120 degrees apart. Mehler's book suggests that arbor runout would most likely show up as extremes that are 180 degrees apart. Anyhow, I experimented by rotating the blade relative to the arbor (about 60 degrees at a time), retightening the nut, and remeasuring the runout. The largest runout I measured was 0.009 and the smallest was 0.004. The extremes were always near the places that I originally marked. I suspect that the different measurements arise from the addition or cancellation of the intrinsic arbor runout and the actual blade runout. I understand what you are saying about amplification of the arbor runout at the perimeter of the blade, although I'm not sure if the factor is ten or five (radius). In any case, I left the blade mounted in the 0.004 runout position and made a few test cuts in softwood I had laying around. The rips and crosscuts are beautiful, even without a zero-clearance insert.

The current setup seems very good, but I still suspect the blade is not within the 0.001 spec. I'm just being picky - blade cost $100!

[David DeCristoforo]

"I'm just being picky - blade cost $100!"

"Real world" test: fire it up and cut some wood. (No dial indicators needed.) If you are dissatisfied with the blade's performance, ask for a replacement.

YM

[glenn bradley]

Yoshi's got it. That being said I would check at a point on the blade about an inch from the outer edge. Right between the teeth would be my most suspect area for consistency. As the tooth is wider than that area . . . you get the point.

[Paul Simmel]

I need to jump in here... and also ask for help on the same subject. I also have Forrest blades (WW-I) and (WW-II), and the same Grizzly saw (brand new).

I posted a thread some time back on blade-wobble. At the time I thought someone had smacked my blade while it was up, or over-heated it while ripping 2 x construction lumber with the 60 tooth blade. I had noticed a wobble during startup and spin down... and some slight vibration where at least I did not RECALL it being there before.

I've since put on a never before used WW-I blade and have the same wobble.

Simple question: If the arbor IS running out, is there a way to correct it?

I had a run out issue with my shaper when I first got it some years ago, and the only solution was to have Griz' ship me a new shaft which solved the issue.

I'm unfortunately recovering from surgery, but do plan on some light tuneup cleaning, etc., this week, and will follow the above steps regarding a possible bit of junk around the arbor and/or washer/nut/blade.

ITMT... I just can not comprehend how a running out arbor could be corrected short of replacement. Am I missing something?

[Eric Haycraft]

The best options is to replace the arbor. Someone on here mentioned an adjustable washer, but I certainly never found it and think that it is probably a discontinued product.
That being said, I decided to fix mine using the rambo approach after someone on here found success doing it. Basically, I mounted a 4.5" angle grinder to the fence, fired up the saw and the grinder, moved them together and let the sparks fly. I think that I let it run for 15 or so minutes of total grind time checking and adjusting as it went. Basically, I just waited for the sparks to almost completely end and called that done. I can barely measure any runout now.

[Art Mann]

The best options is to replace the arbor.

If the OP measured a runout of 0.001" at the arbor, then replacing the arbor is just as likely to provide worse results as improvements. Woodworking machines are not machined to be more accurate than that. I think the OP's biggest problem is that his instruments are too accurate. If he had never gone down this path, he would have spent more time cuting wood and less time fiddling with the machinery and the difference in actual results would have been indistinguishable.

[Eric Haycraft]

True. .001 isn't worth fixing. Mine was worse than that and caused poor cut quality, so it needed to be repaired. It was also eating up my ZCIs. If the cut quality is fine, there is little need for drastic repairs.

[David DeCristoforo]

I am in agreement with Eric. "Adjusting" an arbor that is not specifically designed to be adjusted is no simple process. A machine intended to perform certain tasks should be able to perform those tasks. If it is not, for whatever reason, it should be on the manufacturer to correct the problem. So if your saw arbor has more runout than it "should" and the manufacturer has not provided an adjustment, you should get either another saw or at the very least, a replacement arbor assembly.

YM

[Bill Jepson]

First, Getting back to the original post if you got the TIR (total indicated runout) to .004 on a 10" blade that is fine. If the arbor has .001 then it will multiply. With respect to the arbor runout most people would say you would need to replace the arbor to get it any better. Grinding it in place is difficult. Most setups would just make it worse so leave it alone and be sure everthing is super clean when you assemble. SOmething a lot of people fail to take notice of is if your blade has paint near the arbor, that is lettering, you need to be sure it doesn't go beneath the arbor flange and washer. If it is on one side only it will throw the runout off. I doubt that this is the problem on the WWII but check anyway.
Bill J

[Walt Stevens]

First, Getting back to the original post if you got the TIR (total indicated runout) to .004 on a 10" blade that is fine. If the arbor has .001 then it will multiply. Bill J

I agree that 0.004 is fine. I've run some test cuts, and everything looks really good. I guess I'm too much of a geek. I worked for the National Bureau of Standards for 25 years, and when someone claimed 0.001 accuracy we held them to it.

I talked with someone at Forrest. As you might expect, they are willing to replace the blade, no questions asked. But, after reading the responses to my original question here, I think I'll stick with it until I actually see some performance problems. While I have only made test cuts so far, I'm really happy with the G1023SL and the WWII blade. Now it's time to do something constructive!! I wonder what I should buy next?:D

Many thanks to all who responded.

Walt