Hello,
I noticed today that as the blade on my unisaw spun down there seemed to be some wobble. So after a bunch of measureing I have concluded the arbor face has .009″ runout at the blade just under the teeth. I used feeler gauges to get these measurements and was not able to easily check it at the arbor. The blade is a new Forrest and I am sure that it is not the problem because I rotated the blade 180 degrees and rechecked the measurements. The the high spot was on the opposite side of the blade from the first time. Looking in the Taunon Tool Guide it seems that there should be .001 or less. I cleaned the flange, blade, and washer which didn’t help. The belts have a lot of “set” in them and I wonder if this could be causing it. I’ll try to check that tomorrow by loosening the belts completely, but I don’t think this is the problem. I am not generally a .001 kind of guy, but if I can SEE the wobble that must be affecting the cut. If fact I can hear it at the end of a cut on the backside of the blade. WHat do Yall think?
Thanks for your time,
Mike
Replies
How is the run-out affecting your cut?
How old is the Unisaw? Is it possible the bearings are starting to wear?
My Forrest WW II had a terrible "argument" with a piece of 8/4 hard maple, which bent the blade slightly. These things are thin rimmed, and fairly fragile; have you tried doing your measurements with another blade?
I am suspect of your method of measuring runout. All blades have some and the only way to neasure the flange with certainty is with a dial indicator which can be picked cheaply at http://www.use-enco.com. Loosening the belts won't change anything. The first thing to do is try another blade.
Hello all,
Thanks for the replies. I know that a dial indicator is easier and a little more accurate. However I am certian that my measurements are within .001". This is hard to explain, but here I go. It is not the blade because first I marked the points on the blade where the runout is highest (most to the left) and lowest (to the right). I put reference marks on the blade and arbor flange as to their orientation to each other. Then I loosened the arbor nut and rotated the blade 180 degrees relative to the flange and tightened the nut again. If the max/min readings were in the same location on the blade, it would be the blade, if the max/min readings were opposite from where they were the first time it would be the arbor. The latter senerio proved to be true.
The blade is new. The saw is maybe 5 years old, but when I got it it had so little use that the throat plate wasn't scratched, there was no impacted dust, no scratches on the table, side table, cabinet, anywhere. The saw had not even been "broken in" I bet it had less than 200 feet for cutting on it.
I cleaned the blade, arbor, washer and nut and there is no diference. there are no burs. the arbor feels solid but it is hard to feel when the belts are tensioned.
Yes, I think it does affect the cut quality because I don't get as smooth of a cut as others claim from a blade of this quality. Plus I can hear the runout at the end of the cut when the back of the blade catches the piece. It is not bad but I can hear a very fast (the speed of the blade turning) ting-ting-ting-ting.
What do ya think now?
Thanks again,
Mike
Edited 6/4/2004 1:50 pm ET by mike
"I have concluded the arbor face has .009" runout at the blade just under the teeth."
I'm not sure what that description means.
Where exactly did you refernece the feeler gauge. I'm still suspect of the measuring methods. A dial indicator is a lot easier and a lot more accurate than a feeler gauge. Did you try another blade?? I've never seen a blade that didn't have some kind of wobble. You might try clamping some piece of metal near the face of the flange and getting feeler gauge readings that way. I repair machinery and it's been my experience a feeler gauge is usually a ballpark within a few thousandths.
Rick,
I clamped the ruler from my starret combo square to my miter gauge and locked the miter gauge in place to the right of the blade to where it just touched the blade at the "highest" (farthest right) point. I angled the ruler so that only one corner touched the blade about 1/4" below the teeth (toward the center of the blade). Basicly I did the exact same thing you do with a dial indicator (as shown in the pictures posted in this thread, and shown in mag articles, cataloges, and advertisements for the dial indicator kits marketed for saws). My reference point is the same used with these systems. The only difference is that I used feelers to measure the gap, which I agree is not as accurate. However feelers are used to measure many things on machinery to a high degree of accuracy. The blade deflects easily with little side pressure which affected the measurements. With one finger I could "bend" the blade to remove the .009 gap. So I was carefull to not push sideways on the blade while measureing.
I don't know if you understand what I meant when I said that I repositioned the blade on the arbor. But, if the blade is not true the wobble would be the same no matter how the blade is mounted, even if mounted the wrong way. When I remount the blade the wobble is not in the same spot on the blade, but it is in same spot relative to the arbor.
I want to know if any of you more experienced machine heads know of any other cause for the run out. I have ruled out the belts because the tight-loose cycle due to set is independant of the blade rotation because the motor pulley and the arbor pulley are not 1 to 1. What is the best way to check the bearings? Remove the belts and spin by hand feeling for rough spots?
I will pick up a dial gauge set this weekend. If I end up having to call Delta I want to be able to tell them the run out on the arbor itself with acurate measurements. I really don't want to tear apart the saw to fix this but I do think it is affecting preformance. I wonder about shimming between the blade and arbor. Though that sounds like a PITA, and probably wouldnt work any way.
Thanks again for helping me sort this out,
Mike
Mike...that picture in this thread with the indicator is to make sure the blade is parallel to the miter slots. That kit and the others don't measure run out. You need an indicator and a magnetic base with an adjustable arm so you can indicate the face of the flange. I still think it's your blade. You still haven't mentioned trying another one. It's a cheap easy test. Take off the belts and spin the arbor by hand. You can have bad gravelly bearings that run true.
Rick,
If the jig is secured in place and you rotate the blade by hand you can then measure the run out at the blade. This, of course, would include the blade and the arbor. If you slide the jig back and forth in the slot you can check for parallel as well. Whether the base is magnetic, tack welded, or clamped to the table it works the same. By repositioning the blade as I described I was able to check whether it is the blade or the arbor. Hopefully tomorrow I can get to Rocklers and pick up a dial indicator set to check it at the flange. I tried using a bent piece of steel clamped to the table but couldn't get repeatability on the measurements. I also know that the bearings can be bad yet still run true. Which is why I think it is the arbor. I am hoping that I can beat it into alignment with a hammer. Don't worry I know what I am doing; I'll use a brass ball peen hammer. ;-)
Mike
"Don't worry I know what I am doing; I'll use a brass ball peen hammer. ;-)"
I'm not worried. It's your saw.
< You need an indicator and a magnetic base with an adjustable arm so <you can indicate the face of the flange.
I second this advise...Also, buy a dial indicator from someone other than rockler, unless you are happy paying triple.. (sorry about type but couldnt change it)
>>What do ya think now?
Call Delta.Howie.........
Take a long metal straightedge,preferably from a 20-24" protractor and check for alignment of the flange and the pulley. They must be parallel. Adjust accordingly. Then the real test is cutting a kerf and checking tooth width to kerf width.
Jackplane...Been repairing machines for over twenty five years and thought I heard them all. That's a new one. Bogus!
I goofed. Upper pulley must be parallel with the lower pulley not the flange.
Not enough coffee today.
Edited 6/4/2004 9:24 am ET by JACKPLANE
mike ,
You say you notice some wobble when the blade slows down. On my very old unisaw I also notice some wobble at slow speed, however the saw leaves a fine kerf or cut edge with little or no distortion . When the blade is spinning at full speed is obviously when the cutting is going on, and there is no visible problem. The blank or bell of the blade may have a bit of wow in it but I'm thinking that when all spun up it smoothes right out.
good luck dusty
dusty,
I recently put an old Unisaw back to work and in the process, installed new belts. I did a bunch of heavy ripping and shaping for about 50 large picture frames. I noticed a vibration on shut down and traced it quickly to the fact that the new belts had stretched. I retensioned them and have had no problems since. This might be your problem.
Regards,
Mack"WISH IN ONE HAND, #### IN THE OTHER AND SEE WHICH FILLS UP FIRST"
Friends,
Would still say that it is very difficult to distort such an arbor. Only if it has visible scars, or other damage would it be due for a replacement or machining.
If run-out is meticulously measuerd at the arbor tip, that still does not mean that the arbor is not true. For instance, the points at which the arbor bearing(s) is supported could just need tightening.
-mbl-
You should have less than .003" runout, anything more and it should be fixed. Also, try some Powerflex Plus drive belts, you will be amazed at the improvement in the performance of you saw. Get you runout problem fixed first.
I found the info below at The Wood Shop.
Setup and Align the Table Saw
This is the process I follow to setup and align the table saw and miter gauge for square cuts:
View ImageSince the Set & Sand and arbor together have about .003" runout, I placed a mark on the Jesada Set & Sand and use that as a reference point.
Life is what happens to you when you're making other plans .
Measuring runout, and in fact anything, gets harder the more precise one is trying to be. The accuracy of the measuring instrument starts coming into play when your measuring less than 10 thousandths as does the geometry of the relationship to the indicator tip and that which is being measured. In addition, it becomes harder to know where the runnout actually is.
Personally, I can't imagine one can accurately measure with most cheap import dial indicator closer than 2-3 thousands. In addition, measuring at the saw blade edge introduces 4 sets of tolerances.
Arbor runout, blade runout, blade surface (rough spots), accuracy of the indicator, error in locating indicator tip. The last one will tend to reduce the readout but it is still affecting the accuracy of measurement.
In addition, the surface of the blade may not be flat even though the teeth themselves may be in perfect alignment with the axis of rotation. Also, you are taking your measurement of the blade staticly, the blade may be set so that it is running true when the stresses of rotating "flatten" it out.
Your ears are pretty sensative devices and the sound your hearing could, and I would check with Forrest on this, indicate that the blade is relaxing as it spins down and starts to wobble. Either that or your saw is shot...
Michael,
The sound is while the saw is running not winding down. I really don't want to do anything to the saw at all. I would like to keep working like before. However, I didn't spend that kind of money on a saw that isn't at least in spec. I am most concerned about unevenly wearing the $100 blade.
I changed the blade, and the run out is in the exact same place relative to the arbor. Remember I marked the arbor and the blade. I know that even the best blades have runout, hell NASA's gyros have runout. Do you understand what I mean by repostioning the blade and rechecking so as to rule out the blade? If the runout is consistint to the arbor, it is the arbor; if it is consistant to the blade, it is the blade.
Do me a favor for a moment and assume that my measurements are right +/- .001. If the runout is .009 at a point 4 inches from the flange I am thinking that the runout on the arbor flange itself would be something like .00225. I have access to a Starret dial indicator set and I will use it. My question is how much runout is acceptable (or out of spec) at the flange?
Thanks again,
Mike
Dear Mike,
Seeing wobble while the blade winds down is not OK.
The causes for wobble are unlikely to be the blade nor the arbor.
Once, the arbor nut was the problem. As nuts are used, they loose 'face' since they end in a pitched spiral, or if the nut was slightly out of true, it can increment with time. When the nut looses its normality, as it is tightened, it tends to tilt the blade minutely, and that minuteness is amplified at the rim of the blade. The same thing can happened if there were some debris at the saw support places.
Try turning the nut the other way, or check with another nut. The other problem can be the disk. If your saw uses two disks, interchange them. The bore can become slightly eliptical with time, or with undue stresses.
Also, because the way they are manufactured, some disks have their bore slightly out of line with the shaft, and is not noticeable initially, but that error amplifies with use.
The disks used here are specially first bored and then machined-turned using that bore as basis.
-mbl-
Mike, I hate to be the bearer of bad news, but your saw is stuffed. Just pack it up and ship it, prepaid, to me. I'll recycle the scrap for you.
Mike, I' ve read most of the threads responding to your 'Wobbling' problem.
A simple tack to check the arbor's flange, is to apply Dykem blue to the perriferry of the flange. then, withthe power off, rotate the arbor while advancing a soft pointed brass prod toward the inked area
The prod must be ridgid in it's holder (A magnetic surface guage will do
When the point marks and shows, they'll either be a small arc shown or a full circle. The marks indicate high spots on the flange (arcs) a full circle is 'Right on'
Also, try the same thing while the arbor is under power (Careful there?)There should be no lateral lost motion IE 'Slop'
A good machinist can true up the flange on a lathe.
By the way ,blades can develop a twist from overheating.
Old time Sawyers could 'adjust' their bent blades with a few judicious whacks of a hammer (In the right area of course and on an anvil) Steinmetz
Edited 6/5/2004 4:28 pm ET by steinmetz
Steinmetz,
I did what you described with out the powder, Instead I listened for the sound. I can hear a high spot. I really don't want to tear down the machine. It looks like an all day ordeal. How much do you think a machinist would charge for reface it? I wonder if the part is cheaper from Delta. Of course, the replacement could be worse. I would rather have a very good machinist true it up.
Well I'm putting this on the back burner untill I get a dial indicator. I use my tools to pay the bills so I really don't like down time. I will only fix it if necessary. Again, my main concern is uneven wear on my expensive blade. Sandpaper will remove the cutting marks. ;-)
Thanks again for all of yalls advice,
Mike
Mike, to achieve the accuracy needed to true up the arbor, the machinist would have to center the arbor on a fixture and grind off the high spot on a surface grinder .
Second choice, would be centered on a good lathe and reground using grinding attachment
Since you depend on your machine and can't afford 'Downtime',you MIGHT try to scrape the flange while the arbor is in place and running??
Dykem blue is not a powder,but a fast drying lacquer used to lay out metal parts for accurate centering/removal etc
Years ago, I knew a guy from Texas who invented and used a neat device to repair and true up out of round crankshaft bearings and journals in the car with motor cranking but not under power.
I even did my own Dodge's crank using the same device . (1960)
It consisted of curved abrasive inserts to replace one half of the bearing insert (Shell)
With a remote starting switch, and plenty of kerosene, you crank the engine a few revs then tighten up the rod nuts a bit and continue 'till the egg shape disappears
If you could clamp a used blade to your table so it wouldn't rotate and coat one side of the blade with valve grinding compound to fit snug (,but not tight) to the flange
Then 'bump' the start switch a few times and check your progress
Clean away ALL abrasive and re blue and check for less high spots. Steinmetz
If you order a new arbor, save the old one for emergency replacement ED from Connecticut
Edited 6/6/2004 2:59 am ET by steinmetz
Steinmetz,
I am not sure I understand. Basicly I fix the blade to the table somehow, I can think of a couple good ways. then apply the valve grinding compound between the arbor flange and the blade, then run the saw for a second to grind the high spot. Is that right? I have a right tilt saw, and the nut is reverse threaded so that even if I lightly tighten it I think that it would imediately tighten and wreak havoc. I like the idea, and it seams that it should work It just seems a little redneck to me. Of course being a redneck the idea sounds good to me!
I could also see backing two nuts against each other and using a spring and washer to put the pressure. That should prevent the blade from tightening.
Again I am not going to do anything untill I get a dial indicator and check for certain.
Thanks guys,
Mike
Mike, using jam nuts occured to me too.
Also, instead of using a blade, use a steel bushing to fit the arbor shaft snugly.
No need for any nuts though, just grasp the bushing with a large Vise Grip and slide the bushing along the shaft to 'lap' the flange'(With power on)
Try not to get compound on to arbor shaft! Steinmetz
I checked my Unisaw with a dial indicator and the runout was .003" to .004" with 2 different blades.
Life is what happens to you when you're making other plans .
Jerry,
Thanks, I am wondering how much is average. As I said before I am not really a .001 kind of guy. There is not, nor will there ever be, any Incra stuff in my shop. I get great results from the close one eye and squint method. :-) However I think that .009 is a little much. I mainly think this because I can see and hear it. After buying a Forrest blade I was suprised that there wasn't signifigantly fewer saw marks than my $50 blade left. Though the blade definately cut a lot faster. I think that this is the reason.
Mike
Mike, I highly recommend you try a set of Powerflex Plus belts when you get your runout fixed. You will notice an improvement in the cut along with less vibration and your Unisaw will sound good. Read the reviews at Amazon.com. I have an extra set if your are interested.
I also recommend the flange and nut shown below. The flange grips and flattens the blade better than the original equipment flange.
View Image
Quick-Change blade flange & nut
Easily change blades with this blade flange & nut set. Large knurled knob permits removal of flange without blade wrenches. Fits all Unisaws,10" Contractors saws, & 9" Builders saws.
P/N 36-659 (right tilt) Price: $21.50
Life is what happens to you when you're making other plans .
Occasionally an arbor flange can get a small ding or burr on the edge that will make a blade run out of true, but the burr will be almost impossible to see, or even feel. It can happen from just dropping a wrench. If this condition appeared suddenly, that may be the proglem. I had this happen a time or two in production shops, and here's what I did:
I got a big file -- a 12" mill file does just fine, and laid it gently against the inside arbor flange while the motor was running. It wants to vibrate a little at first but as soon as the ding or burr is gone, the file rides smooth. I keep the file as parallel to the arbor surface as possible, use a gentle touch, and don't let my hands get anywhere they can get hurt -- You can keep your hands completely above the table, and there's nothing to grab the file.
This has been literally a five minute fix for me a couple of times. Obviously, you don't want to alter the shape of the arbor flange at all, just knock the burr off.
If you're not REAL comfortable around spinning parts, don't try this. If you are not completely comfortable, you are probably doing something wrong.
Michael R.
For all the time spent talking about it you can just change the arbor. I used to work for a Delta dealer as a tech and could do it in a couple of hours. Arbor is $88 for a right tilt.
Here's some instructions...
http://www.sawcenter.com/unisaw.htm
Pricing...
http://www.sawcenter.com/unisawparts.htm
I'd just order the arbor direct from Delta
I'd buy the bearings from a local bearing supply house or http://www.mcmaster.com Should cost around $5 a piece versus the $18 from Delta. I recall the bearing number is 6203 but it's been awhile.
Popular Woodworking has a special July issue on the table saw that many might find useful...deals with runout and such topics. If you still go for the indicator an import is more than enough in quality.
Edited 6/6/2004 8:59 am ET by rick3ddd
Edited 6/6/2004 9:02 am ET by rick3ddd
Edited 6/6/2004 10:17 am ET by rick3ddd
Rick,
Unfortunately I am pretty sure it would take me a whole day. I would also have to remove the side table, then reallign every thing. I can replace a large window, or a double door entry set in two hours, but not what I saw in those pictures. So I think that my time talking about it is well spent! Not to mention that my wife eyes glaze over as soon as I start talking about anyting having to do with my machinery, so I have no one to B.S. with except yall. I have an older customer who is a retired machinist and I think he would do it just to have something to do. I don't have any cabinetry work for 3 weeks so I wait untill I get him to check it out.
I just picked up that mag last night. I don't normally read it because of the articles about building a jig to resaw thick lumber on the router table, or this issue How to make dove tails on the table saw. But the picture of the dial indicator checking runout caught my eye.
Mike
Depending on your set up I usually unbolt the top and lift it high enough to slide the saw out. Put a saw horse underneath to support the top. Plus if you change the bearings you don't need a puller to get the one closest to the flange which would be the hardest part. Now you have that article on tuning up your saw which will help a lot to get the table lined back up. I thought you were going to pick up a dial indicator at Rockler? I usually buy a lot of my measuring stuff from http://www.use-enco.com or http://www.wttool.com I just bought a 12" dial caliper from the latter for $27. The catalogs are better than the websites. Well worth having around even for a woodworker.
Rick,
What do you think about Steinmenz (sp?) idea for grinding the flange can it be done. If not, why?
Mike
Mike,
I read the first posts and skimmed the last ones so I might have missed a detail but I think you need to back up and start simple. Your description of the performance of the blade points to saw alignment problems. The sound of the teeth touching at the back of the cut is indicative of a misaligned fence and this would also lead to a poor quality cut with saw marks, fuzzing, etc. I recommend that you check this first and then look for more serious issues rather than jumping into a saw overhaul.Charles M
Freud, Inc.
Charles,
Thank for the input. Your right If I open the fence away from the blade a bit the sound goes away. Unfortunately if my measurements are right the arbor does seem out of spec. An interesting note I picked this saw up from an extremely wealthy man (I don't use that lightly either) and he had the saw set up and dialed in by a Delta tech. Every thing is set up as perfect as I can measure. I am suprised to find this runout thing. Hopefully I am wrong and there is something else in play here, though I can't think of what.
Mike
Mike,
"If I open the fence away from the blade a bit the sound goes away."
Does the cut quality change if you do this?Charles M
Freud, Inc.
UHH, I'll check.
Thanks. I have never really messed with the fence much. I just set it dead on to the miter slot an go. I'll try that.
MIKE
Mike,
Go to the library and check out John White's book, "Care and Repair of Shop Machines". He gives a very simple and cheap set up for very accurately measuring arbor run out. You will need a dial indicator which can be had for $20 or less from a number of sources including Grizzly. I have a magnetic base which I used for years--untill I got John's book. His little wood block set up is easier to use and probably more accurate than a magnetic base. Ingenious to the point of brilliance!
Your instincts are correct that there is a problem. With the blade removed, the run out at the arbor flange should be less than .003; my 1950 Unisaw has less than .001. If the run out at the arbor is excessive, pull up and down on the arbor and rotate by hand to check for any hint of click or rough sound. That would indicate bad bearings. They are not that hard to change and there's a couple of sites, owwm.com (or something to that effect) for one has a poster that offers instructions. If the bearings check good, plug your saw back in and wrap a nice square wood block with some 220 or 180 grit SP. Turn the saw on (still no blade of course) and apply the wood block carefully to the spinning arbor flange surface. Polish it up real good and recheck your run out. If you notice improvement, which you should, keep after it and eventually increase the grit to 320 and do a final cleanup. Try to keep pressure on the flange only, not on the shaft itself. I would shoot for run out of less than .001 at the arbor flange.
Belts are a separate issue and will not effect run out with the saw unplugged. Vibration yes, run out no.
Once the flange is tuned to "awesome" it's an easy matter to check any of your blades. For fine work I would expect run out of no more than .004 - .005 as far out as you can measure on a good quality blade. On your "crude" cut blade/s a bit more is acceptable.
While you've got John's book give the rest of your saw a real tight tune up.
Best of luck and please post results.
Regards,
Mack
"WISH IN ONE HAND, #### IN THE OTHER AND SEE WHICH FILLS UP FIRST"
To all,
Thanks again. I appreciate yall lending your experience to me. I am really waiting untill I get a dial indicator before I start jacking around with my machine. I'll definately post what comes up. I am confident in my admitedly redneck measurements so I expect I'll be fixing it somehow.
Mike
Mike,
I think your method has some merit but there are other factors that can also have a part to play. You are actually measuring the relationship between the blade and your miter way, not just arbor run out. Under flourescent lighting a spinning blade can give some visuals when the pulse of the light and the rpms harmonize. This often happens when the blade is slowing down so let's forget about the visuals for now. When a blade tings as you finish a cut, it is often a sign of misalignment. I would take the following steps. Unplug the saw, remove the drive belts. Use your miter way as a register for a combination square. Start at the front of the blade. Move the square blade up until it just ticks a blade tip and lock the square. Move the square to the back of the blade and you should have the same tick on the same tooth. If not you are out of alignment and have to adjust the trunion. Check the other side the same way, it is not unusual for there to be a .001 difference between miter ways. Remove the blade. Feel the arbor, spin it gently by hand, try to wiggle it up and down. You may be able to feel some grinding or movement which indicates bearing replacement. Take an engineers square and square down from the top of the table to the arbor flange up against the arbor. You may want to put a light inside the saw so you can see. Move the bevel control wheel until the flange is square with the top. Move the square out to the edge of the flange face, hold it tight and move the arbor. If the flange face pushes or falls away from the square instead of just kissing it all the way around, then you have some trouble. Next take a combination square and use it for a depth gauge from the table top to the arbor center. Again move the arbor. Just like the flange a push or pull on the square blade indicates a problem. Replace the drive belts, good time to check pulley alignment and your keyway set screws. The belts should not make much difference unless bearings are bad. Re-check everything again. It is always possible that a previous owner bent the arbor or flange. This will be evident with the above method. If you have the large Unisaw with the 1" interchangeable arbor make sure you check that first. Remove the arbor, clean and make sure that it fits the taper and is straight. You can buy a dial indicator and magnetic base but unless you get a good one they are known for being inaccurate. The mounting arms and fittings can also add their little fudge factor to the measurements. I agree that .001 is great if you're working finite math in a vacuum but .009 is too much.
This forum post is now archived. Commenting has been disabled