I just got an old 70’s unisaw moved into my shop. I bought it used and it looks to be in good shape except for one problem…
I have noticed significant runout which doesn’t appear to be in the blade.
As the saw slows to a stop, the blade appears to wobble quite a bit. I have tried a few blades, all which run very true on other saws that I have used.
Tomorrow I will bring home a dial indicator from my cabinet shop and check the runout at the arbor flange as well as on the blade and post the numbers here.
I read some other posts on this, which suggest the problem could be:
1- bad arbor nut
2- bad arbor flange (out of true, or perhaps it just has a burr or high spot
3- bad arbor bearings
4- arbor shaft out of true
5- bad belts
My question is, how should I go about determining which of these factors is the actual cause of my problem? Are there specific tests that can determine where the problem lies?
I am happy to get new bearings and belts, but if the problem is in the arbor I want to discuss this with the guy who sold me the saw because a new arbor is around $120…
thanks for you insight,
Vincent
Replies
What you are seeing is a harmonic flutter. At a certain point as the blade slows down, after the power is switched off, a speed is hit where the natural tendency of the blade and arbor assembly to vibrate is momentarily magnified. Most of the movement you see is actually in the blade not in the bearings.
It isn't uncommon and it will even happen on new machines, but any wear in the bearings will add to the movement. It usually doesn't indicate that anything is seriously wrong with the machine and won't affect the cutting. A simple test of the arbor run out won't necessarily detect wear in the bearings.
John White
Thanks john,Would you then say that a good first course of action would be to install new bearings, given the age of the machine and the strong chance that they are worn? Do you know of a good bearing supplier? Also, the belts do look pretty worn... I may go ahead and replace them as well. Do you have any favorite brands of replacement belts? I still plan to check the runout this eveing with a dial indicator. I may also measure the kerf of the cut on this saw vs. one at work. At least I know that what I'm seeing isn't actually just runout...
thanks for the tipvincent
I have worked on a lot of machines where the original bearings were well over fifty years old and they are as tight as the day they were new, so you shouldn't assume that you need new bearings. The chance that the bearings are truly worn out is probably no more than one in a hundred. Also, once a ball bearing starts to fail, it fails completely very quickly, they don't get a little loose, they rapidly fall apart with lots of grinding noises.It is worth considering that bearing replacement is a fairly difficult job and one that can easily leave the machine in worse shape than when you started. Unless you have some experience or training with pulling and replacing bearings you probably shouldn't make your first attempt on a Unisaw.I would go to a bearing supplier for the bearings and belts. A brand name bearing made in the U.S., Europe, or Japan will probably be of better quality than a Chinese made bearing but it will cost you considerably more. The same advice applies to the belts.John White
What about the belts, John?? Would you stick with the OE-type belt, or can link belts be substituted on a Unisaw?forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
You will pay a lot more for OEM belts. When I was working as a Service tech for a machine dealer I used belts from Grainger. No need to specify matched belts. Buying the belts at at the same time will insure a good enought match. The cog type belts are better than a Fenner. The belt is held in tension so there is no memory issue. This is the reason theyare recommended on contractor saws. The original Delta belts are an odd fractional size. Regular or cogged belts come in even sizes but there's a close match that will work. Cogged belts are better on smaller diameter pulleys.
Hi, Rick. I didn't meean OEM in a literal sense, just that the were V-belts or whatever they're called. What's a "cogged" belt?forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
http://meds.mmtc.org/background.asp?X=Cog%20BeltsLots of folks just go to the manufacturer and pay triple for standard off the shelf items of your local power transmission store. http://www.mcmaster.com among others sells cogged belts.I get bearings from http://www.bearingbrokersinc.com I pay about $1.60 for Nachi 6203 double sealed bearings for the Unisaw. Good Japanese bearings.
Edited 5/12/2008 6:02 pm ET by RickL
Thanks for the sources, I'll check them out
FG, either type of belt will transmit the power. If you install the three belts so the "set" is offset, you shouldn't notice much if any vibration from the stock style belts.
Theoretically the link belts will have less vibration. It is up to you to determine whether that decrease in vibration is worth the increased money you will have to spend.
Hi FG,RickL's response said about what I would say, the cogged belts seem to work slightly better than linked belts on smaller pulleys but the linked belts would be fine.John White
Ok john, here is an important update:I brought a starett dial indicator home from work today and tested out the runout readings by manually turning the motor pulley and looking at the numbers. When set on zero (i got an initial pressure of about 10 thousanths of and inch and then zeroed the scale) the dial read as high as .009 in one direction and .005 in the other direction from zero. I was using a nice, three month old ridge carbide TS2000 blade to get the readings, and I marked the high reading and rotated the blade 180 degrees and took the same readings and the high reading happened 180 from the mark... so I think blade runout is not likely to be the main factor here. I checked the FWW blade test and saw that the TS2000 had about .003 in measured runout in that test... so I'm going to assume that my numbers are only accurate to about three or four thousandths.I tried the same test with the indicator down close to the arbor, and I got lesser readings of about six or seven thousandths at that location. ...the bearings don't look or sound bad, and I tried to jiggle the arbor with a good amount of strength and could not get any play. The arbor itself looks to have some scratches and marks of age... but I don't know how to tell if it might be the culprit. The flange also fits a little loosely on the arbor, which worries me. What do you think my next course of action should be here? Vincent
Here is a link to a good set of instructions on how to replace/check the arbor bearings.
http://www.sawcenter.com/unisaw.htm
The also have instruction on how to listen to a bearing to hear if it is bad.
Hello Vince,
I agree 100% with John, it probably is harmonic flutter on coast down. The way I would determine if anything was wrong would be to make a few careful test cuts with a good blade. If the cut surface is accurate and up to par smoothness wise, I would put it to work.
Bob, Tupper Lake, NY
Vincent, now that the experts have settled on harmonic flutter, I'll chip in that when I first started using my old (now gone) blue Jet saw, it did the same thing. Problem disappeared when I replaced the belt with a link belt.
"Problem disappeared when I replaced the belt with a link belt."
Ditto that. Link belt and a set of machined (as opposed to cast) pulleys did wonders for my cheapo saw.
Mike HennessyPittsburgh, PA
I was lucky, didn't need to resort to a pulley change. For a free saw, that Old Blue wasn't bad.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
"For a free saw, that Old Blue wasn't bad."
I'd go so far as to propose that, for free, NO saw is bad. ;-)
Mike HennessyPittsburgh, PA
"...for free, NO saw is bad. ;-) " Well, not sure I'd go that far, LOL. I did have to sink $175 into a new Jet motor (retail usually around $300?), but other than that, pretty smooth situation.
Gave it up because I was tired of the fence, and didn't want to spend $$ on a new one, and the lock for the height was so soft I was going crazy when making dadoes in close-fitting item. But that was after several years, so no complaints here.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
Since I just measured the runout and got fairly high numbers ( about six thousandths of an inch just above the flange on the blade) I'm thinking that harmonic flutter is not the only factor here. I've used a lot of unisaws, and this one just feels and sounds rougher. I'll go ahead and replace the belts, but I get the feeling while using the saw (which is backed up my my readings) that something else is out of true. vincent
If you used a blade to check run out you may have an issue as not all blades are dead flat. Try it with several differnet blades and see what happens.
Doug
"about six thousandths of an inch just above the flange on the blade" I guess this is one case where we can hope you're wrong? That would translate to quite a bit of difference at the rim, eh?forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
yes, I got a measurement of 14 thousandths of an inch runout at the rim of the blade. And I know that everyone says you can't trust the blades to be accurate, but I'm using a very nice ridge carbide TS2000 blade which FWW measured as having about .oo3" of runout in their blade test. Also, it's quite new and shows NO sign of runout or wobble on the other tablesaws that I have used. Even if my numbers are off by .003 at the rim of the blade, it's still cause for great concern!I'm seriously considering just spending the $150 to buy a new arbor w/flange and belts. From what John White said about bearings, I don't think that mine are the culprits... and I'm a bit weary to replace them without the proper experience and equipment. I got the saw for $600 and it's in great shape otherwise... so i really wouldn't mind if I had to put a little more money into it. (it would still cost less than half as much as a new Unisaw)
Well, good luck with your efforts to settle that saw down! I should be able to fire my Unisaw up this weekend, maybe even Friday. Hope the new mobile base can handle our crappy, unlevel cement floor, LOL.
forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
Vince,
If you haven't already done so; take the blade off and measure the runout at the arbor flange. If this is where you have .006 runout, you have a problem.
I have had good results with this procedure, you may want to try it. With the blade removed, wrap a small, square piece of hardwood with 180 grit metalcut paper (not garnet paper, it's too messy). Start the saw and polish the arbor flange, holding the paper FLAT onto it. Recheck your runout and repeat if necessary to the point of diminishing return.
New saws, I've heard, have a tolerance of .0015 runout at the arbor. I have a 1950 Unisaw that had just under .003 when I got it and has less than .001 now.
After you get the runout issue solved, there may be other issues that are giving you excessive noise; belts would be my first point of attack.
Best of luck with it!
Mack"WISH IN ONE HAND, S--T IN THE OTHER AND SEE WHICH FILLS UP FIRST"
The flange on my saw fits loosely and does not spin with the arbor unless the blade is on and tightened down... I can slide it on and off the arbor easily when the blade has been removed. It's only when it it tightened up against the bearing that it stays put. I wonder if this is a potential problem? I think i can screw just the nut down to the flange to get it to hold still and then take a measurement. I'll give this a shot and tell you what reading I get. I agree, I would like to replace cheaper items first... unless they are not broken. vincent
Arbor runout was .006 to .007
Some of this i would attribute to the rather rough surface of the arbor flange. I took the measurements from the surface of the flange where it touches the blade, since this is the part that determines the runout in the blade. vincent
Try putting the dial indicator on the arbor flange, that will eliminates any discrepancies with the blade. If the run out is greater than .005 you should have the flange machined flat. I would also throw new bearings in. You may have a little more down time on the machine, but based on the vintage, its time for bearings. When you take the belts off spin the blade, if it continues to spin more than a couple of revolutions, the bearings probably dont have any grease left in them.
Best of luck,
Paul
The saw has a strange older style arbor and flange which are two separate pieces. I can slide the inner flange right off the arbor easily. The flange had some deep grooves in it which make it impossible to get a dial reading that represents the runout applied to the blade... so I've just been measuring runout on the blade plate just above the flange. I did true up the flange by running the saw without a blade and holding a flat sanding block up to the flange while it was spinning. This polished out most of the grooves and helped a LOT with the amount of runout I was measuring. Somewhere down the line I'm just going to rebuild it with new bearings, spacers, and a new arbor.thanks for your ideas.Vincent
Try the simple, cheap stuff first.
Clean everything. Polish and feel for burrs or inclusions.
You might have more than one problem.
My experience is mostly in larger, thinner industrial saw blades. While all blades have their own harmonic frequencies we usually see blade wobble at higher speeds. What we do see as a machine slows down, we classify as a different kind of wobble. (Not everyone does it this way.) We would call what you are seeing machine wobble not blade wobble. (This might be splitting hairs.)
First, check your flange and your nut. They can’t be too smooth. Feel them with your thumb; rub a cotton ball over them. I have known guys that felt them with their tongue arguing that the tongue is the most sensitive.
If you have a bit of dirt measuring 0.0005” of an inch a half inch from the center of a ten inch blade you can roughly figure it will want to push the edge of the blade out of line by about 0.05” Roughly, roughly. As you tighten it will be forced into the blade and the nut or collar. The dirt will compress. The actual blade will be harder to distort than this simple mathematical model. Still a little dirt in the wrong place can make a big difference on the edge of the blade.
Tom
I have cleaned all these parts carefully and checked for burrs. They don't seem all that wonderfully smooth to begin with... but they are clean and there are not any burrs that I can detect. vincent
Have you tried emery cloth on a flat surface or a stone to polish them flat?
<!----><!----> <!---->
Might not be it. Some of the guys are pretty sure it is belts.
Tom
I just tried the trick where you glue some 220 sandpaper on a board, take off the blade and turn on the saw. Then you polish the flange while the saw is running with the sandpaper... like a lathe. This actually made a significant difference. It's not perfect, but I suspect the belts still play a part in things as well. I haven't taken the measurements yet, but I could tell a real difference right away just in the appearance and feel of the saw, and in the cuts it made.
Vince,
It sounds like you are on the right track. Try this as well; with the blade off, stand a nickel on edge on the table top. Start the saw, let it run a bit and shut it down. I would think that if belts were causing enough vibration to transmit it to the arbor the nickel would fall over. If it stands, I would try the sand paper trick on the arbor flange again. Be careful you are FLAT against the flange and NOT removing metal from the shaft itself.
On my old Uni' the flange is either an integral part of the arbor or a very tight press fit because it does not move---at all.
I understand yours basically has two washers and a nut. Is that standard for Unisaws of your vintage??
Regards,
Mack"WISH IN ONE HAND, S--T IN THE OTHER AND SEE WHICH FILLS UP FIRST"
I don't know if that's standard for the vintage. That was something I was hoping to find out here on knots, but it seems like this thread has pretty much lost it's momentum. I'll have to post a new thread and ask folks about that. It didn't pass the nickel test every time, so I went ahead and ordered new belts from delta, as well as a new outer flange and nut. If things are still rough after these are replaced, I'll go ahead and get a new arbor and bearings. I have located a friend with the tools and skills to replace these components correctly (he's a mechanic AND a woodworker) thanks for your helpvincent
There is no need to have ordered the belts from Delta. You will be over paying a stock belt. The cogged belts from Mcmaster would be fine and a lot cheaper. Unisaw belts are 28-5/8" but a 28" or 29" cogged belt would be around $7 each. $4 each for regular belts. They make a wedge belts wich is deeper than a regular belt and transmits more power. Also is the cogged wedge which is better for smaller diameter pulleys.
I've been repairing machines for 25 years and spent 8 years as a tech at an independent dealer. We did a lot of serive and warranty repairs on Delta machinery. I've done every repair you can do on a Unisaw.
Did you try running one belt?? Most folks don't realize you only need one belt. The three belt thing goes way back and it's from the days of overkill. PM went from three belts to two belts. The engineers know they only need one belt but it's a marketing thing. If they went to less belts the public would scream. Further proof of the one belt thing is all the Euro sliders I've worked have one belt. I'm talking 7.5 to 9 hp motors with bigger blades than a Unisaw.
Edited 5/15/2008 12:40 pm ET by RickL
Edited 5/15/2008 1:01 pm ET by RickL
Edited 5/15/2008 1:02 pm ET by RickL
I already ordered the delta belts, since I didn't know the size of the unisaw belt or how to determine the size... (i should have just asked) But I'm sure they'll be fine, and they were about 8.50 per belt. I was worried about getting a different size because there isn't that much adjustment to move the motor up and down... but next time I'll try your suggestion.
Sounds like you really know what you are doing. Also the discussion is getting pretty machine specific and I have never rebuilt one. So I think I'll back out and wish you well.
Let me know if there is any way I can help.
Tom
This forum post is now archived. Commenting has been disabled