Hello all,
A couple of months ago I posted that I was showing an unacceptable amount of runout at the blade on my late 90’s right tilt Unisaw. To clarify this saw was in showroom condition when I bought it, not more than 2 hours of use.
Previously my methods of measuring the runout were …. circumspect. So this weekend, armed with a Brown and Sharpe dial gauge and a Master Plate, I proceeded to investigate. I came up with the same measurements as before (the feeler gauges were pretty accurate after all) .011″ run out at the blade (Forrest WW II). It is my hope that my findings will help other picky Unisaw owners.
The arbor flange itself showed about .0015″ runout on the dial gauge. I don’t have a test indicator that reads down to .0001 so this measurement is not all that reliable as it is at the edge of what the dial indicator can read. I mounted the Master Plate with the arbor at four positions (12, 3, 6, and 9 o’clock) and compared readings. The Master Plate cannot be rotated because it is a big square. I came up with runout of .007″. After some helpful ideas from yall and a lot of head scratching I decided I had nothing to loose and got out the sand paper and 3 in 1 oil. 🙂 OK, for the record what I did next may seem unsafe to some, if so please feel free to not try it. I first held oiled 60 grit paper on the flange with my thumb while the motor was running. This removes metal REALLY fast!! I then polished up with 220 grit paper and took another round of measurements. It worked! A little. I forget now but I think that the readings were like .002 better. I was knew that doing this free hand was not going to be very effective as the paper was removing metal all the way around not just the high spot. So I used an engineers precision square to grind the arbor as shown in the pic below. The technique (redneck as it may be) is similar to flattening a wood plate on a lathe. I used adhesive backed 120 and then 220 grit paper lubed with oil stuck to the surface of the square that engaged the flange. The danger in this operation is if the paper grabs the flange the square will go airborne or trap your fingers in the throat opening. However, I felt more secure doing this than using a panel raising bit in the router. Regardless, I was successful!! The runout after this measured .002″ using the Master Plate. Keep in mind that the Master plate offers considerably more measurable area than a saw blade, so at the blade the runout would be less. The last thing I did was to stick a piece of 120 followed by 220 grit paper to the 1″ hole in the Master plate and slip it over the arbor to flatten the face. Using the square and my fingers had slightly convexed the machined surface of the flange. This operation would be potentially disastrous if the plate bound on the arbor. Imagine installing a dado set while the saw is running! However the master plate has 1 inch holes that gives ample clearance.
I knew that this operation could have possibly ruined the arbor, but I was prepared to replace it any way so I had nothing to loose. Especially considering the replacement might have been just as bad! And I was not looking forward to the time involved in that repair.
I took all of the measurements many many times using great care. Through all this measuring I found a few interesting things about the Unisaw in general.
1. Belt tension affects the arbor. My belts have a lot of set in them. (Though I have no experience to compare it to) As I rotated the arbor the belts got tighter and looser, this is not due to pulley run out. The arbor deflected as much as .002″ because of this. (The bearings are solid and clean). I changed the belts so that they were no longer in phase with each other. In other words the tight spots don’t line up now.I intend to get link belts soon. Interestingly pushing on the table would change readings I wonder if a Powermatic is more solid.
2. When the blade is raised all the way the trunion deflects. Using the Master Plate and dial gauge the reading jumped .0015″ or so when the trunion topped out. So 90 and 45 degree stops should be adjusted at less than full height.
3. My $120 Forrest blade has more runout than my Freud Ultimate Cutoff blade! I’ll ask Forrest about this.
4. I replaced the arbor nut and flange with Deltas better set (the nut and flange are integrated and machined better). It made no difference other than it is a lot easier to install blades now.
5. Setting up a cabinet saw is a lot easier than dialing in a contractor saw!
I am sure that many will attack my methods of taking measurements, all I can say is I repeated everything many times and archived perfect repeatability. The runout with my Forrest blade is .005″ and the Freud blade .003″. That is better than twice as good as before. While I spent about 8 hours messing with it I think that it was better than tearing down the saw to replace the arbor.
Thank all of you for the input and suggestions on my original post.
Mike
Replies
Mike,
While I find your report interesting,it occurs to me you might rather be a machinist than a wood worker. The tolerances you mention are acceptable for any woodworker I know, and the tablesaw is only one many tools used to get wood in its final condition.
Best of luck!
Ha Ha,
I know what you mean and I agree fundamently with you. I am not a .00000 kind of guy in general. I will never own any Incra micrometer fences! If it fits and looks good, job done. However I was suprised when my new Unisaw and Forrest blade were not cutting any better than my old set up. After posting my findings several people who I respect (you are one of those) posted that .011" of run out was not good. That's almost 1/64" and I try to build to the 64th.
Besides metal is just plain ugly compaired to curly maple and walnut burl.
Mike
please excuse my spelling.
Yeah, wood is good.
With parts made for heavy machinery produced all over the world and not as localized, it's always good practice to check for trueness, flatness, etc. I came across a Delta chopbox from Taiwan- had a 1/4 inch warp over 14" (fence). About as bad as it gets...
It's just that when you mentioned feeler guages I started thinking about adjusting valves on my truck <G>
LOL, I don't even change the oil in my cars any more. Although I have a 26 foot '85 class C motorhome on a Chevy chassie (sp?) that has 130 k miles. That is a lot for 350 in a motorhome. I do all the maintance (and road side repairs). I will be rebuilding that beast next month! Although I will buy a remanufactured long block so no feeler gauges for me!
Mikeplease excuse my spelling.
Mike,
I am sorry you felt we were attacking you, I don't recall anyone saying you were stupid we were just questioning your methodology.
Here goes another one, using the dial gauge in one position to measure the arbor flange is correct. Using the master plate and sliding the dial guage along the slot you are measuring two things at once. The parallelism of the mitre slot to the plate combined with any runnout of the arbour flange.
Also, as to runnout on blades I could be WAY off base here but the dynamically test the blades and I am not sure testing them statically yields a useful measurement. I think quality of cut, which you said isn't as good as you expected with the Forrest is a better indicator.
I would agree with you that .011 off is WAY to much for and decent saw, let alone a Unisaw, heck my Craftsman contractor does better!
Good luck and I like you method of truing the arbour flange, I am going to try that on mine, it also has the effect of squaring it to the table.
Michael,
I assure you that I did not think I was being attacked. I was getting anoyed at the insistance that the blades were the cause. But, that was because of my inability to explain how I was repositioning the blade on the arbor to cancel blade runout. I was pretty confident in my measurements, even though I knew my methods were crude. The doubts and criticisim (positive in my view) helped me work it out in my own head. For me that is what this site is all about; many heads are better than one. If I didn't respect yall as craftsmen (and FG) I wouldn't bother. In fact I quit posting on some other forums for that very reason.
Using the master plate I did both; keep the gague in the same position and changing the arbor position, and sliding it in the miter gauge slot. Sliding in the slot worked because I could compaire the various readings at the different arbor positions. For example I zeroed it out at the front and read .006 at the back, then repeated the process at 12, 3, 6, and 9'oclock. By subtracting the diference between 12 and 6 or 3 and 9'oclock. I found the runout (actually the runout doubled, I think).
I know that Forrest blades are "hand trued and ballanced" or something to that effect. I will call eventually to get the scoop.
As far as using the square as I described to grind. I dont think it was accurate enough to keep the face of the flange perpindicular to the table. However the last step I did where I used the master plate to flaten the face will flatten it. Then I adjusted the stops. Remember I was fully prepaired to replace arbor if the results were unsatisfactory. I probably wouldn't recomend this operation unless you were simularly inclined.
thank you,
Mikeplease excuse my spelling.
This forum post is now archived. Commenting has been disabled