Recently got a Powermatic PM2000 cabinet saw. Yesterday I set the miter slot to blade and miter slot to fence parallel adjustments using a dial indicator. I was able to get both to within .002". So far so good. Today, I've place the blade at 45 deg., and the measurement from the front of the blade to the rear is +.016". I understand this indicates the table is sloping at the front side & needs to be shimmed. OK, I get that. HOWEVER, if I loosen the 2 bolts on the front side of the table & lift, the error INCREASES. If I loosen the 2 bolts on the REAR SIDE of the table & lift, the error ALSO INCREASES. How an this be??? Any suggestions appreciated.
First suggestion is that a range of .002" to .016" from 90 to 45 degrees is more than likely well within the tolerance specifications of the machine.
Secondly, it is not only the table that can cause the change. The machined tolerances of the trunions, trunion brackets, plane of the cabinet flange, change in the moment arm created by the motor, the blade if that is what you are checking off of. as well as the flatness of the table, how parrallel the table bottom bosses are to the top, etc.
During the assembly process the manufacturer sets the alignment at 90 to as close to zero as possible. The tolerance stack up of all the other parts mentioned above is what determines what that becomes at
The manufacture then rolls it over to 45 and checks the alignment to the slot and has a tolerance that qualifies the saw as a good unit. Often there is nothing you can do unless you want to loose the close setting at 90. to improve the 45.
Frank-thanks for the reply. I'm using a good quality blade (Forrest woodworker II) which I have measured & which has minimal runout. Plus I am measuring at the same point on the blade front & back (rotating the blade). I understand what you mean about tolerances stacking up. I also understand the issue may not be purely the top is uniformly lower in the front; could be one corner, etc. However a .016 variation is more than 1/64 of an inch. This amount would make it virtually impossible to make anything like a tight edge miter of any length (say, for a speaker cabinet) with the blade tilted. I expect more from a $2000 saw.
Once again, not necessarily. You might get a surface that is ever so slightly more rough than perfection. Probably take a profilometer to measure it. You might get a touch of burn, although I doubt it. But the wood travels past the whole blade. Doesn't mean that any part of your cut is .016" off to any other part of the same board.
However, as you might know, a plane is established by three points only and your table and cabinet are four. You may improve by shimming one corner, if the bulk of the "problem" is either parrallel of the plane of the bosses to the top of the table or plane of the top plate on the cabinet. It is all trial and error. Or if bevel cuts are a way of life for you, you might cheat your 90 setting the other way a little.
And you might call Powermatic to see if they have any tricks. They will probably tell you it is within specs. but may give you some ideas.
However, if it were mine and I were within .016" with a dead on 90, I'd lock it down and cut wood.
I would at this point suggest to first take Frank's advice and start cutting some wood. While exacting tolerances are a plus, sometimes we cannot take full advantages of these fine settings because of the quality of the material that we cut and or out technique can be sub par to the machine tolerances. First see how the saw cuts at a 45 degree bevel. If you don't see any kerf marks or burning you should be good to go. We often get a bit too wrapped up in using a dial indicator to measure "everything". If you are not happy with the cutting results call Powermatic and get them involved.
Ditto. The proof is in the pudding ... or, as an old friend, full of (resigned) Italian wisdom, oft stated: "What you lose on the bananas, you make on the grapes".
One of the most time wasting tools in the hands of an anal wooddorker is an alignment tool. How do I know? ... just finished my regularly scheduled, once a decade, more or less, alignment of the table saw last week, so the subject be (too) fresh on my mind. :)
When I start chasing my tail, I split the difference ... cuz, what you make on the grapes, you also lose on the banana's.
Boudreaux says c'est la vie.... and make some sawdust.
Thanks for the replies. I guess the net result of the blade being angled ~.016, all other things being equal, would be a kerf wider by that amount. By futzing around, I was able to get a tradeoff error of .008 on 90, .009 on
One reason I'm a bit obsessed with this is about a year ago I bought a Bosch 5412 slide miter saw, specifically to execute a dining room table design I had in mind. Hard to describe in words, but where the members which connect to the legs attach to a center longitudinal member entailed 4
45 degree cuts. No matter how much screwing around I did with that saw, I could not get this joint to not have gaps in the range 3/64". The pieces kept getting shorter & shorter as I attempted to get the joint to be tight. Made a lot of expensive sawdust (but no table) with that one.
After setting the saw this way, I made some test cuts on a piece of 3/4" inch plywood, 4.5" wide by 10.5" long. The board is perfectly square and flat. A 90 degree cut made on each end is perfect. However when I switch to 45 & make an edge miter cut, the angle is perfect, but the trailing edge of the cut protrudes 1 full 1/16" above the square when the leading edge is brought flush with the square on that end. In other words, the board is now
1/16 inch longer on one edge at the miter cut than at the other. Try to make a box with 8 such cuts & you'd produce a spiral. If anyone has further suggestions please share them.
On a contractors saw, this problem is usually caused by the two trunnions not being parallel and the answer is to shim one side of one of the trunnions. I don't know if the same applies to cabinet saws, but they do have trunnions - perhaps someone else can comment.
I read through the original problem and your replies so far. Sorry I didn't notice the thread earlier. As you have now discovered, the situation isn't quite as tolerable as some have said. There has to be a reason why shimming the table doesn't fix the problem but makes it worse. It doesn't make any sense that this should happen. I'm sure that it wouldn't take long to figure it out if I could come by to look at it but I suspect that you are not in my neighborhood (Boise, ID). So, you'll have to do the legwork.
If the reading at the rear of the blade is higher than the reading at the front, then the rear side of the table needs shimming (bringing up the table will, in effect, lower the blade). If the reading at the front of the blade is higher than the reading at the rear, then the front needs to be shimmed. Pay careful attention to the magnitude and direction of the reading because it's easy to get confused. You might think that you've made it worse, only to discover that you have over corrected and caused larger error in the other direction. And, don't use any paper, wood, or other fiberous material for the shimms. I'm not a big fan of plastic either. I like using aluminum. Thicker pieces can be cut from soda cans, thinner pieces can be had from a roll of foil.
Give me a call on the phone if you want to go through it in real time:
Yeah ..., make some goddamn furniture and quit f*cking with the table saw!
If nothing else, what I did for the 30 years before I had a table saw worth "setting up" ... a plane and a shooting board.
It is absolutely ridiculous to get bogged down in measuring, in thousandths the tools we use, in a medium that is liable to change +/- 1/16th between sundown and sunrise.
And ... beware of those who do nothing else but measure, and make no sawdust whatsoever.
Ed-Thanks for the reply. Seattle's not THAT far from Boise! ;-) Saw is left tilt, I'm using the right miter slot. On 90, both slots measure the same, so are parallel to one another. I reset the 90 to about .003, at which point the 45 is about +.012. To me this says the table is HIGHER in the rear (gauge tip is going up the left tilting blade as it moves front to rear, and thus increasing the gap). I'll try the pop can shims. And thanks for the offer of phone assistance, I may end up taking you up on that!
I understand the "make some sawdust" sentiment, but cutting expensive wood before the saw is all it should be makes no sense. If you bought a brand new sports car which pulled to one side, would the response be "just go for a drive"? 1/16" of error across a 4.5" cut is simply not acceptable. I could better that result cutting by eye with my circular saw.
Ok, so if I understand you correctly, the reading on the indicator is going down (counter-clockwise) as you move from the front to the back. Correct? This would correspond to the table being higher (blade being lower) in the rear. This means that you need to shim in the front.
You can quickly approximate the amount of shim needed by knowing how much the reading changes and how far apart the two readings are. First, correct the reading for cosine errors. Then figure out the ammount of change per inch. Finally, multiply this by the distance between your table bolts.
So, for example, if your blade is tilted to 45 degrees, then the correction factor is going to be 0.7071 (cos(45)). If you measure
0.012" misalignment, then the actual change is 0.0085" (0.012 *
0.7071). If it was measured over a distance of 8", then you have a change of 0.0011" per inch (0.0085 / 8). Then, if your bolts are 20" apart, then you should start with 0.022" (0.0011 * 20) worth of shims. This isn't going to be exact but it's a good place to start.
No problem. Glad to help.
I understand completely. If you wanted to spend your time reworking everything that your tablesaw should have done correctly, then you wouldn't have spent $2000 on a new tablesaw.
Beware of the guy who spends more time making sawdust than furniture. ;-)
You have the scenario exactly, and I must say I'm very impressed, though I am simultaneously saddened by the fact that at one time I actually *did know* all that trigonometry. Now get this: After some experimenting (and BEFORE I read your prediction) I was able to get the error of both 90 & 45 degrees down to ~.003 by inserting a .025" washer under the table at the 2 front table corner bolts. Again, very impressed you were able to predict this so closely with calculations.
Cuts are now PERFECT!
LET'S MAKE SOME SAWDUST!!!!!!!!!
BTW if anyone is interested, I can email PDF's of several articles I have on tablesaw tune-ups which helped me work this out, a video of one of which can be seen at
If the only problem is that the axis of rotation of the arbor is not parallel to the table, then shimming the table relative to the cabinet can fix the problem.
If there is other slop and play in the mechanism, I agree that there isn't much that can be done.
Good quality pieces are designed to handle wood dimension changes. It's expected. That doesn't mean that you can cut your parts to +/- 1/16th and have it come out anywhere close to decent right off the saw.
If the power tools aren't accurately set up it means that the hand tools get more of a workout paring and shooting to get everything accurate.
It's possible that you're cranking the blade too hard against the 45 or 90 degree stops. I might try cuts at 46 and 89 degrees, just to see if the stops are the problem.
Hard to imagine how the error can increase no matter which way you tilt the table.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.