You are all invited to join this discussion. I thought it best to move it into its very own thread, so as not to totally disrupt wirerat’s original thread while Woodwiz and I hash the thing out. I’ll recap in 3 separate posts here. Please note! This is not a discussion about the pros and cons of “toeing out” a fence. The central issue is whether a toed out fence creates a tapered cut (however slight). I say “No,” he says “Yes, most definitely.”
OK, here’s the post I made that got the whole thing rolling.
Jamie says: “I think, though, that you and I have our differences about the whole “offset creates tapered cut” thing… I still think if you have a featherboard in front (operator’s side) of the blade holding the stock to the near/front end of the fence, you will not get a tapered cut. [Afterthought: featherboard or hand pressure]
I s’pose some day I should offset my fence and test that theory. Your drawing assumes that the board moves away from the fence at the entry of the cut. If you stabilize the stock at the entry of the blade, as should be done, and your pushing force is parallel with the blade, the cut will be straight.”
The drawing to which I refer is Woodwiz’s below, he has stated: “Here’s a picture of why you don’t want toe-out. It’s exaggerated, but shows how the wood follows the line of cut established by the blade.”
forestgirl — you can take the girl out of the forest, but you can’t take the forest out of the girl 😉
Another proud member of the “I Rocked With ToolDoc Club” …. :>)
Replies
Following up on my response above, here is a sketch of the scenario I described. For those joining the conversation late, I do not believe you need to toe out your fence. This illustration is simply to show that a toed-out fence does not produce a tapered cut.
View Image
Edited 4/20/2005 1:16 am ET by forestgirl
Whew! Now, here's Michael's last response:
"if you start out with the board parallel to the cut line, then it will follow the cut line." Why would you start out with the board anywhere but parallel with the cut line?
ponder the first drawing I provided. Is it not a correct representation of how we rip wood
No, it is definitely not, and that may be the key difference here. When you rip lumber, you start out with the wood tight to the infeed part of the fence, hence necessarily parallel to it. When you feed the wood, you keep the wood tight against the fence, especially at the point even with the front of the saw blade, and again, necessarily moving parallel to the fence. In your drawings, the wood is well away from the fence at the point where it needs to be tightest to it to maintain a consistent cut width.
Now if the blade and the fence are not parallel to each other, then the cut line can't be parallel to the fence because the cut line always wants to follow the line of the blade.
I maintain that the drawing you provided is not representative of a toed-out fence, but rather representative of a blade that's cock-eyed, and an imaginary woodworker feeding in the direction of the blade.
It's not a matter of which one is cockeyed. It only matters that they are not parallel with each other. Your own drawings clearly show the wood is not moving parallel to the fence, but rather at an angle to it. This will produce a relatively straight cut, because you are starting out with the cut line more or less parallel to the edge of the board. I used to cut plywood on a table saw without a fence a lot. But you lose precision that way. The purpose of keeping the wood tight to the fence is to control the width of the piece, and when the wood pulls away from the fence you lose that control, and your dimensions start to wander.
If you changed your drawing to show a starting position with the front corner of the wood tight to the fence at the point where the blade starts to cut as well as the rest of the way to the front of the fence, you would see that the cut line of the blade would go on a slant through the lumber.
WE all seem to agree that when you have a rip fence toed in, you get a bind. If you have a very slight toe-in, the blade will cut sideways enough to compensate for the toe-in, but if you have more than just a very few thousandths, the wood will bind. This is because the blade and the line of cut are not parallel to the fence. The blade wants to cut outwards in relation to the fence, and since it can't, the wood tries to move sideways into the fence. The fence, if it is a good one, isn't going to move either, so the wood binds up pretty quickly, with predictable results.
With the fence (or blade) toed out, is is the exact same principle, only in reverse. The wood wants to pull away from the fence instead of riding into it, and there is nothing to restrain it other than a featherboard or the saw operator. If there are more than a very few thousandths of toe-out, the wood can't be kept from moving away from the fence, just like it can't be kept from binding in the reverse case..
Your drawings clearly show that the wood can't move parallel to the fence, and at the same time parallel to a saw blade that is not parallel to the fence.
This might be a key question: If a piece of wood binds when the fence and blade are toed in relative to each other, what keep the exact same principle from applying to make the wood move away from the fence when they are toed out from each other?
I know I'm right, because I have physically demonstrated this any number of times. And I know you are an intelligent and avid woodworker. The interesting challenge is to show you what I know to be demonstrable truth without actually showing you on a saw. Maybe I'll try one more drawing if I can do it without repeating the same approach. I'll have to think a bit about it.
It might be worthy of note that "back in the day" when things weren't nearly as precise (Jet-lock fence, anyone") we used to determine whether the fence was parallel to the blade by looknig for an equal number of "upcut" blade marks as of "downcut" marks because this indicated that the blade was not rubbing more at the front or the back.
Now I did as you asked, and looked at your drawings and your argument, and I hope, answered them. I'd appreciate it if you would, in turn, make a real effort to understand my points before you try to refute them. You might also want to take a deep breath, and sit and ponder why so many people with substanial industrial experience agree with me, and why they disagree with you. I know that's argument from authority, and a logical fallacy, but it does carry some practical import, nonetheless.
I'm not fighting with you, and this isn't about winning or who's right. It's a fairly important safety issue, and it's about what is right. If I can convince you, I can probably convince some other folks at the same time, and those would both be good things.
Michael R
"You have to look for possibilities where there are none" Krenov
Yippeee, now I get to write some. For those of you who are bored, that's OK, you can check out. Michael will want me to respond, I'm sure. :-)
"When you rip lumber, you start out with the wood tight to the infeed part of the fence, hence necessarily parallel to it." Exactly, but then again different than I think you're meaning. At the beginning of a rip cut, we have pressure against the wood just in front of where it enters the blade, agreed? OK, that means the board is parallel to the fence at that time, and over a relatively short distance. However....as the board moves into the blade, we continue the pressure against that very specific point (I'll call it Point A), and that is where the width of the cut is established.
Here's the critical difference: The stock does not have to move parallel to the fence the entire length of the fence. If that were true, then short fences such as they use in Europe, or as found on the Unisaw, would not work. The stock would just shoot off in the direction of the big gap where the short fence ended.
The reason that such an event does not occur is because of the pressure at Point A, just in front of the blade, which keeps the cut referenced to that part of the fence. Even with a toed-out fence or a short fence, the stock will continue in a straight line, wether the fence fades away gradually or just disappears. As you said a couple of times, the cut line will follow the blade.
"In your drawings, the wood is well away from the fence at the point where it needs to be tightest to it to maintain a consistent cut width." Au contraire, mon ami. The cut is made at the front of the blade, and that's the only distance that counts. Again, if that were not the case, a short-fence for ripping would not work.
"When you feed the wood, you keep the wood tight against the fence, especially at the point even with the front of the saw blade." I would edit that to say "just in front of the saw blade" (otherwise, you're putting pressure on the blade), and note also that there is no pressure per se after that point, as there is no way to provide any pressure, unless you can slip your fingers between the blade and the stock that's been cut and push toward the fence <g>. [There's the short fence thing again; how did they ever get those to work?!] If there is pressure, it would be because the fence is toed in and needds adjustment.
Now comes the complicated stuff.
Statement A: "...if the blade and the fence are not parallel to each other, then the cut line can't be parallel to the fence because the cut line always wants to follow the line of the blade." Yes, we want the cut line to follow the line of the blade. As long as the pressure is kept consistent at Point A, establishing the width of the cut where the stock enters the blade, the cut piece will be straight, not tapered.
Statement B: "It's not a matter of which one is cockeyed. It only matters that they are not parallel with each other." Sometimes yes, sometimes no. Here's that pesky Point A again. If you are only putting pressure at Point A toward the fence, it probably doesn't matter. But if you are somehow magically pressing the board onto the fence the entire length, all the way through the cut, even at the back of the blade, yes you will get a tapered board. Physically, though, that's not likely to happen.
[hmmmm, gotta take a breath]forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Sorry, I don't have any way to address this to Woodwiz. I think the essence of the confusion here is the idea that the stock has to ride along the fence all the way, parallel to the fence, however it's phrased and thinking, therefore, that the blade has to be parallel to the fence to get that straight cut.
Another way to think of it is: The cut has to be parallel to the side of the board being cut. Since, with correct pressure against the fence at Point A, the entry edge of the blade is always the same distance from the side of the board, the resulting cut will be straight.
forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Edited 4/20/2005 2:22 am ET by forestgirl
Edited 4/20/2005 12:45 pm ET by forestgirl
See my post this date in the original thread.
Forestgirl. you are basically correct but everybody just hasn't thought through the physics of the matter. Short fence is all you need as long as it is long enough proximally.
Just DON'T toe IN the fence, or the board will bind on exit and be a problem.
Bones
Most of the European tablesaws have a rip fence that can be shortened or lengthened to accommodate this technique, though typically I have heard its explained usage in the shorter position for those times when you are ripping a board that is under tension. Do you have problems with this technique with wobble at the end of cut?
I see y'all started without me. I'm two hours ahead of you - had to sleep.
Before I start on a different approach to presenting my case, I need you to answer the main question from my last post, which is in essence:
We all agree that excessive toe in causes the wood to become wedged between the blade and the fence, resulting in binding and potential kickback. What keeps the exact same principle from applying in the other direction when the blade and fence are toed out in relationship to each other, causing the wood to be wedged away from the fence?
It would appear that you have forgotten or aren't familiar with some of the basic definitions and principles of geometry, and that those will have to be established and agreed on before I can convince you of the truth. If this one question doesn't help convince you, I'll try a different approach. I will try to establish one principle at a time in sort of a socratic dialogue, and will only move on to the next when we agree on it. Once the two or three principles are agreed upon, I will present one deduction at a time based on those principles for discussion, and those will lead both of us to an inescapable conclusion. That should keep things short, focused, and to the point. Is that OK with you? I'm not near as bright as Socrates, but I can try......
Just to recap, this is what I am trying to prove to your satisfaction:
1. Excessive toe out is almost as bad as excessive toe in, just not quite as dangerous. The same principle that causes boards to bind when blade and fence are toed in, in relationship to each other, causes the wood to be pulled away from the fence when blade and fence are toed out in relationship with each other, resulting in tapered cuts.
2. Since carbide saw blades have teeth that overhang the sides of the plate, they can cut sideways a little bit to compensate for minor toe-in OR toe-out, but in doing so, they have to cut with the back of the blade, which is uindesirable and potentially dangerous. The overhang on the carbide blades I measured is about .010 to ,012" so the saw blade can compensate for quite a bit of misalignment, in OR out, before these effects become a noticeable problem, but it is still better for the saw blade to be centered in, and parallel to, the kerf so that the blade is not cutting on the rise.
Earlier in this fandango you wrote:
I'm completely convinced that the whole toe-out idea is an archaic, useless habit that should be given an honorable funeral and disposed of.
That shows me you have an intuitive and practical understanding of what I am trying to get across, but that we don't agree on how one gets there. Agreed?
We might as well start with those two questions - the one about generally agreeing on the conclusion you stated above, and the question about the same principle applying on toe-out as well as toe-in.
This may be very boring to most folks, but I think that if I can convince you, I will have come up with a presentation that will convince just about anybody, and can establish for all that fences should indeed be set as near parallel to the blade as possible. Thanks for being a good sport.Michael R
"You have to look for possibilities where there are none" Krenov
Sheez, what a long discussion of a relatively simple subject, perhaps I'm missing something. Let me know, because I believe this can be explained in one short paragraph, just need to do some English rehearsal.
I'm with Forest on this one, it's good practice to have slight toe out.
About cutting a taper, well, I cut coves with a fence set up to 60 degrees toe-out on my table saw and they are all parallel and straight?
When you toe out, all you do is widen the kerf......
Also, when you toe out, you have less chances of cutting with the back of the blade, compared to perfect alignment, due to internal stresses in the lumber, specially kiln dried. Besides binding, when toed in, this is the whole idea, to widen the kerf slightly with toe out and avoid kick back, for us bad guys who have their guards and protection gathering dust.
Damkit! I knew this was going to happen. "I'm with Forest on this one, it's good practice to have slight toe out." Not!
I will repeat the first paragraph of the first post:
As I stated several times, I do not "believe in" a toed-out fence, but that's not the question here. (See "tapered cut" above).forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Sorry Forest, when things are self explanatory, and real easy, my speed reading sometimes gets me into trouble, as I miss unimportant detail.
I'm still with you though, why are my coves so nice and straight, square and true with a toed out fence?
Anyway, do what makes you happy, my fence will always be toed out 1/32" over it's length. In fact it's written in my saw's owners manual, so someone agrees with me at least....
If you want a real short Math geometry explanation let me know, I'll even include the calculations.
If you don't believe in a toed out fence, why bother arguing the point about tapered cuts anyway, it has no relevance to you then? (True, but just jokin)
Edited 4/20/2005 12:23 pm ET by Jellyrug
I got some 'rthmetic that might shed some light on the question of "excessive toe-in or toe-out.
As I wrote above, carbide saw blades have about .010" overhang. I believe the side of the tooth can cut this full amount.
Assuming this is so, then a saw blade could work without binding if it were as much as .010" out of alignment with the cut, toed in or out. This amounts to about .001" per inch of blade or fence. The alignment is relative.
Assuming the blade is parallel to the miter slots as is should be, and that the table is 28" deep front to back, this would allow the fence to be toed in or out approximately .028" or almost 1/32" over the depth of the table before problems started showing up. This is probably why some people find 1/32" of toe out acceptable. What isn't acceptable to me is having the back of the blade cutting on the rise, and the only way to avoid that is to have blade, saw and wood movement all parallel to each other.
This is a little off point, I know, but does tend to define limits.
Back to the two main questions, (I hope.)Michael R
"You have to look for possibilities where there are none" Krenov
> As I wrote above, carbide saw blades have about .010" overhang. I believe the side of the tooth can cut this full amount.
> This is a little off point, I know, but does tend to define limits.
This is far from being off point, in fact it's the very crux of the misunderstanding.
If the toe out (or in) is slight enough that the material encounters the rising teeth rather than the flat side of the blade, it will be cut again at that point, widening the kerf but allowing it to remain parallel to the fence. If, in a toe out, it just barely touches the side of the blade before it gets to the rising teeth, it may be possible with enough pressure to bend the part being cut off and keep the other part against the fence. Again, the rising teeth will do some cutting. How much bending is possible will depend on the size and springyness of the part being cut off.
However, if the material is stiff enough and the toe out great enough, then it will be guided by the flat side of the blade before it reaches the rising teeth, and a tapered cut will result. In the case of a toe out, this will force the material away from the fence. In the case of a toe in, it will jam the material between the fence and the side of the blade until friction against the blade becomes greater than friction against the fence. At that point, the whole might and fury of the table saw is applied to flinging the workpiece backwards. That's why we stand to one side, not right behind the workpiece.
So, you're both right. Which way it goes depends on how much toe out there is, and the size and properties of the material being cut off.
-- J.S.
Edited 4/20/2005 6:03 pm ET by JOHN_SPRUNG
Edited 4/20/2005 6:06 pm ET by JOHN_SPRUNG
Hi Michael, welcome back. My turn to catch up on the time thing!
I agree that we're not "fighting" (I would never do that, right? too funny Only person I fight with is my hubby, go figure). We're having a lively discussion. I'm just hoping we can keep the discussion on the central point which is, "Does a toed-out fence cause a tapered cut?" This is the historical question that came up a month or so ago, and re-appeared in a more recent discussion. I have been saying no (given good technique) and you have been saying yes. If, in the meantime, we cover your "toed-out is dangerous" idea, that's fine.
Yes, agree that the fence should be parallel to the blade. The only thing I disagree about is whether toeing out (to be disdained) causes a tapered cut because the board would drift away from the blade.
So, having gotten part way through the 1st cup o' java, I'll take a look at your explanation above......
Leseee, yep I agree that the semi-Socratic approach is fine, and (LOL) I agree it might bore our buddies. They can check out if they want.
So, now we get to Question #2 "the question about the same principle applying on toe-out as well as toe-in" as posed here: "What keeps the exact same principle from applying in the other direction when the blade and fence are toed out in relationship to each other, causing the wood to be wedged away from the fence?" The "what" is the pressure we provide against aforementioned Point A, pressure applied to hold the stock against the fence just in front of the blade. By holding the board against the fence at that spot, the operator prevents rotation of the board outward toward the fence. The board will travel straight along the line of the blade, irregardless the fence fading away from the blade. (The other force being applied to the stock is the feeding force, which obviously must be correct, parallel with the blade. It's a minor actor with respect to this discussion, but should be mentioned.)
If the fence is toed in through misalignment, the fence creates dangerous pressure toward the blade that is only exacerbated by the pressure at Point A.
OK, your turn, my friend. I think we can both be dubbed as "good sports."
BTW, your explanation of how carbide blades, with their overhanging teeth, can accomodate a slight misalignment of the fence is excellent. I will, with your permission, use it in future discussions.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Sorry, I wrote that last post while you were writing yours.
Thanks for your response. I think this will work better. At least with shorter posts (we both have work to do,I think.)
OK, here is one point where we differ in perception:
Assuming the board is straight, and the fence is straight: If the board touches the fence at any two points, say at the front of the fence and the point just in front of the blade, it is by definition parallel to the fence. Two points determine a line, and since the board and fence both touch at the same two points, they are by definition on the same line. The length of the line is infinite. The actual length of board and fence doesn't have much to do with it, as long as they are of finite length.
This is axiomatic: it derives from the fundamentsl principles of geometry. Are we in agreement so far on just this statement?
I will make a second assertion that only works if we agree in the first one: If the edge of the board and the face of the fence are on the same line, any movement of the board along and in contact with the fence has to be, again, by definition, parallel to the fence. Any movement not parallel to the fence will push the board into or away from the fence.
Agreed so far? On just these two points. If not, then what is the alternative?
Michael R
"You have to look for possibilities where there are none" Krenov
Edited 4/20/2005 1:00 pm ET by Woodwiz
Hi again, I'm back.
"Two points determine a line, and since the board and fence both touch at the same two points, they are by definition on the same line." Yes, they are, at the point in time when you lay the board down on the saw table. We generally snug the board up to the fence.
After that point in time, however, another factor enters the equation, the force at Point A, illustrated above. The pressure at Point A is the only lateral pressure that is applied to the board once the ripping operation begins. That means there is no longer a "second point" in your line, and there is nothing to keep the board parallel with the fence. The board follows the path of least resistance, which is the cut line of the blade, with the pressure at Point A establishing the width of that cut.
The fact that the fence is pretty short on the entry side of the blade keeps things sane in the real world. I'll grant you, if the entry end of the fence was 3' long, you would run into a problem with a toed out fence, but only because the length of the fence on the entry side is so much longer than that on the other side of the entry point of the blade.
When I look at it in a really exaggerated way, I can see where you might think the outgoing cut would be tapered, but in a real-life situation, the pressure at Point A overrides everything else and keeps the cut parallel to the edge of the board.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Ok, now wer'e getting down to the fundamental differences. You got a little away from my question, but that's OK.
My PM 66 has almost 18 inches in front of the fence. If you have a unifence or some sliders and slide the fence back, you have approximately 36 inches of fence in front of the blade. It doesn't matter at all for the moment what's behind the blade.
With even a foot of fence in front of the blade, you can't rotate the wood away from the fence to follow the line of the blade without pulling it away from the fence at the point where it enters the blade. Your earlier drawing illustrates that perfectly. After the cut starts,the wood touches the fence only at the front of the fence, and pulls away as you get farther back. You can't keep the wood against the fence at the point if entry and still have it follow the (angled) cut line without bending the wood, or forcing the blade to cut a little bit sideways.
Do you see this, or would it help if I make a drawing? I'm still trying to establish one point at a time and not develop the argument further until we have agreement on each point. One flows from another.
Michael R
"You have to look for possibilities where there are none" Krenov
Edited 4/20/2005 2:11 pm ET by Woodwiz
Woodwiz and forestgirl-The runaway verbiage on most of these posts has probably addled my mind, but hasn't Jellyrug given a simple answer to the problem you are worrying to death?If you toe the rip fence away from the blade and keep the edge of the workpiece against the fence (why would you want to do anything BUT keep it against the fence?), then you've duplicated the setup for cove cutting. The blade height is set for ripping rather than coving, so the top to the cove is above the board, but the fence-blade-board geometry is the same. The cut will be straight, and the kerf will be wider (it's the bottom part of a cove), as Jellyrug has pointed out.I don't understand what all the mysto-fog about pressure points and fulcrums and fence lengths has to do with the question any more than the gyroscopic properties of the spinning blade, so I will leave it to you two to work that out.
The point of the exercise is to find a way to convince an unbeliever, then simplify it into a concise, clear argument that will convince anybody. We know it won't interest everybody, so if it doesn't, just wait til we're done, and see whether you agree with the results.
To answer your question, you can't cove cut full height in one pass. When you cove cut, you are forcing the blade to cut with the side if the teeth, and you can only raise the blade about the height of a tooth in one pass. This is because the wood hits the plate of the blade when you raise the blade too much.
When you toe the fence in or out, first the blade is forced to cut a little bit sideways, widening the kerf as you pointed out. This is bad because it forces the blade to cut on the rise in the back. This is one of the prime causes of kickback, and also gives you a poor surface. I hadn't thought of it being out of square, but it does that too. Thanks for the observation.
When you toe the blade in or out more than about 1/32" over the length of the saw table, the blade can no longer cut sideways enough to compensate for the angle, and the plate of the blade either forces the workpiece into the fence (toe-in) or away from the fence (toe out), resulting either in a bind or a tapered cut.
That's all the dicussion is about. All the verbiage is about finding a way to make FG see this, or for her to convince me. I think we're getting close, and having a good time in the process.Michael R
"You have to look for possibilities where there are none" Krenov
Michael-"When you cove cut, you are forcing the blade to cut with the side if the teeth, and you can only raise the blade about the height of a tooth in one pass. This is because the wood hits the plate of the blade when you raise the blade too much."I mentioned cove cutting as a quick way to reference familiar geometry. But even with a small toe-out (you mentioned 1/32" at the rear of the fence)) you're still cutting a cove--just a very narrow one. With your small toe out, the set of the blade (or it's brazed carbide equivalent, overhang) will cut a wide enough kerf that the stock won't hit the plate before reaching the uprising teeth at the rear. So the cut will be straight.I agree with you that an exaggerated toe-out will cause the board to collide with the plate, moving it away from the fence and ultimately resulting in a curved cut. That's dangerous on more than one count.If brevity and clarity is your goal, I think a sketch showing several snapshots of a board progressing through the cut would be all that you'd need.
I pretty much agree with you , but it's not you that I'm trying to convince.
Even a slight toe in or toe out causes the blade to cut on the rise in back, and that's not good.
We agree on the consequences of excessive toe, in or out.Michael R
"You have to look for possibilities where there are none" Krenov
Michael/f_g.I haven't read every word in this thread, but I think you've both got valid points.You don't want the cut to bind between the blade and the fence a/c kickback (= slight toe out)You don't want the saw blade to contact the kerf at the back of the cut a/c rough saw cut edges and potential for kickback if the workpiece moves (=fence parallel)The lesser of two evils here is the saw cut edge being scored by the teeth at the back, in my opinion, so I sway towards having a slight toe out for safety reasons, regardless of cut geometry.Having said that, to throw my two cents worth into the pot, will the fence being toe out create a tapered cut? Not if it's slightly toed out - 10 thou toe out over a 10" blade = 0.06 degree - virtually parallel. I set the rip fence at about third of that and it doesn't make a practical difference.If, on the other hand, the fence is wildly out of square and feeds the off side of the cut onto the saw plate, then you'll either start deflecting the cut or bending the saw blade/arbor, none of which are really what you'd want.Cheers,eddie
As a separate aside on rip cuts (hence the separate post)I was trained in trade training on our legal machines (European-style setup in US parlance).We set the rip fence short to avoid the potential of binding.To rip (edit) narrow pieces {edit - no less than 48" but less than 96"), we make a cut half way along the length, and, with the saw running, hold the piece hard against the table, back the piece out quickly and then flip it end for end and finish the cut - as you've then got a safe hold of the offcut and the workpiece things won't fly across the workshop. [edit for pieces less than 48", use the fence as normal]But, having said this, remember there's a riving knife there to help avoid the piece coming into contact with the back teeth of the blade.We do use a long fence, but only if you're ripping a panel that's as long as it is wide. To cut a 12" by 12" panel, use the sliding fence, but a 3' by 3' panel can be cut either with the slider or the fence.Cheers (again)eddie
Edited 4/20/2005 6:57 pm ET by eddie (aust)
Ahhhhh, everyone's been busy while I was gone. Cool.
This is the essence of it for me. The "real world" thing. A couple of points have been made that didn't occur to me at the beginning of the debate. One is that the teeth might be cutting on the rise if a sufficiently toed out fence is used. And, correspondingly, the kerf would be wider than if the fence was parallel.
Whether or not either of these occur must depend on how toed-out the fence is, all technique being equal and good. You guys are willing to do the math, I have to admit I'm not, so have at it. Jellyrug, bless his heart, came up with .005 wider than the actual thickness of the blade/teeth.
Michael's concern is "Even a slight toe in or toe out causes the blade to cut on the rise in back, and that's not good." I would agree that in theory, that's not good, but in reality, a few thousandths of toe-out doesn't seem to hurt anything.
I will concede that if the fence is drastically toed out (canted, really), there's going to be a weird cut, because you can't get the stock set up between the fence and the right part of the blade, and the front part of the stock will be angled. However, in the workaday world, I suspect that, although we snug the stock to the fence when we set it down on the table, the moment we apply that sideways force at Point A on a toed-out fence and push the stock the 1/2" or whatever it takes to get it started on the cut, the board becomes parallel with the blade rather than the fence. It would be an almost microscopic difference with a fence that's toed out just slightly. The width of the kerf would easily accomodate that microscopic shift toward the blade.
I very well might go home and purposely toe out my fence, rip a little wood and take some measurements!
Well, I'm supposed to be working. My boss (me) is getting restless.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Forerst,
I know you hate it when I get off the main subject and become non-obedient. You scold me with blue, red and bold black ink and some exclamation marks as well.
But on the "related" subject of toe out, do you have the fellow from Freud's email, so he can tell us what they recommend, here in the forum?
> I very well might go home and purposely toe out my fence, rip a little wood and take some measurements!
Just clamp a board with a shim for the near end to your fence. That way you can easily and reliably return to your original conditions after the tests.
-- J.S.
Thanks for pointing that out, John. Way better idea. I was in no shape to turn on a table saw last night, under the influence of antihistimines and barely able to think!forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Okay,Lemme see if I have this right: Wiz an FG agree, but they can't agree on why it is that they agree. Huh?Without demeaning your exercise of civil discourse, please allow me recount the observation of my departed grandfather, to wit:"A man convinced against his will,
is of the same opinion still."Have fun,
Tom
"When you toe the blade in or out more than about 1/32" over the length of the saw table, the blade can no longer cut sideways enough to compensate for the angle, and the plate of the blade either forces the workpiece into the fence (toe-in) or away from the fence (toe out), resulting either in a bind or a tapered cut."
Michael,
When considering extremes, you are correct, and yes, you will cut on the rise side, of the back of the blade.
The blade is actually not cutting sideways though, it is still cutting straight, as the cut will follow the line of the fence, which is also the line of motion. The workpiece will not be in touch with the plate of the blade, as the teeth of the blade are wider and they will remove material, prior to touching the side of the blade.
You are correct, in an extreme situation, the quality of cut will deteriorate.
This is a very hypothetical discussion though, as toe out works, both in practice and theory.
If you do some quick calculations, these are the numbers.
On a table saw, with a 27" long table, if you toe out 1/32"(measured at the back end of the table) and you cut 1-1/2" lumber, with a 10" blade, the kerf in theory, by calculation, will be 0.005" wider than the thickness of the blade teeth projection to the sides. This gives you a perfect clearance of 0.0025" on each side of the blade, which gives a better cut and prevents all kinds of problems. There will be no problem in keeping the work piece against the fence and the cut will be straight and true. The cut will take the form of a miniature cove, but too small to measure, if we are cutting wood. As lumber is not an Engineering uniform material, it has internal stresses and differences in density, without any clearance during a cut, the chances of touching the sides of the blade, pushing it away from the fence and cutting undulations, are more.
Why don't we get the saw blade guru's "Freud" into this discussion and see what they say. I'm sure they have spent more time on this subject than all of us.
Edited 4/20/2005 6:26 pm ET by Jellyrug
Edited 4/20/2005 6:36 pm ET by Jellyrug
Jellyrug-A very minor quibble with your analysis:- There is a cutting region at the front (and rear) of the blade. The size of the region is a function of blade size, board thickness, height of blade above the board, and probably the gravitational pull of the operator.- If the fence is toed out, the actual cut will be widened: no toe-out, no widening; some toe-out. some widening; lots of toe-out, lots of widening.- Once the leading edge of the cut passes the cutting region, the offcut side will drift toward the blade plate, following the angle of the toe-out.- But you, yourself, don't toe the fence out enough to cause contact of the board with the blade plate. Never the less, the offcut side is closer to the plate than the half-width of the cutting region by the time the leading edge of the board reaches the rear cutting region.- Hence the rising rear teeth of the blade will contact the offcut side of the board. - There is thus a greater-than-zero potential for kickback when a fence is toed out even slightly. That is the quibble.There is much to be said for the European short rip fence.
Edited 4/20/2005 7:15 pm ET by Donald C. Brown
Donald,
Theoretically, I agree with everything you have said.
When we are talking about lumber and 0.0025" differences, really it becomes irrelevant.
The few kick backs I have had, were all while cutting Kiln dried lumber and the kerf closed on me at exit from the blade. (No splitter)
In theory, a wider kerf should help in this instance, but again the tolerances are so small and we are working with "living" wood, so it's irrelevant.
About touching the plate, you are correct, as with most carbide tooth blades you would have to toe out quite a bit.
We're seeing the same thing, but coming to slightly different conclusions.
I think the kerf will be about .010" wider. since that is roughly the amount of misalignment in 10 inches. (1/3 of the .031 in 30 inches) but that's a minor matter.
True the kerf will be straight if you keep the wood against the fence, and if there isn't too much toe. I've said that several times. By sideways, I meant the blade will be cutting with the side of the teeth instead of the tops.
But how can it be good to have a blade cutting on the rise? It makes a much rougher cut, requiring more cleanup, and most imoportantly provides an environment prone to kickback.
I didn't figure the amount of cove, but yes, it must be very slight.
You do agree, don't you, that the same degree of toe-in that will cause a board to bind, when reversed, will cause a board to pull away from the fence? Extreme seems to be anything over 1/32" either way, and that's really easy to get with a jetlock fence. Not so easy with a Bies.
Also, don't you agree that just as the saw can compensate for up to 1/32" of toe-out, that it can, in exactly the same manner, compensate for the same amount of toe-in? That's why most (not all) people who do this for a living prefer to set their fences as close to dead parallel as they can.
Michael R
"You have to look for possibilities where there are none" Krenov
"I think the kerf will be about .010" wider. since that is roughly the amount of misalignment in 10 inches. (1/3 of the .031 in 30 inches) but that's a minor matter."
LOL
That 0.010" is when you have 5" of blade sticking out above the table, cutting 1-1/2 lumber? My saw won't allow more than 3-1/8" and I won't adjust the blade as high as it will go, when cutting 1-1/2" lumber. This discussion is about a minor matter, but it's becoming interesting to follow the major debate.
Really, the debate is a hypothetical discussion, as you won't find the level of precision in any lumber, make a substantial difference to the minor amount of toe out we are talking about. I believe most fences out there are not this accurate anyway and a lot of woodworkers do just fine with an old contractors saw.
In the end, it depends on your type of blade, as some may need more clearance around the kerf than others.
If we were in a CNC machine shop, cutting steel, this could be a major discussion.
Edited 4/20/2005 8:16 pm ET by Jellyrug
That is when you have 5" of blade sticking out above the table, cutting 1-1/2 lumber?
I'm talking about the almost 10" length of the blade, not the height, since the misalignment is along the length of the fence.
Really, the debate is a hypothetical discussion, as you won't find the level of precision in any lumber, make a substantial difference to the minor amount of toe out we are talking about
It's maybe not so hypothetical if your saw binds up and kicks back, or if you're ripping a 14 foot piece of expensive hardwood and your cut comes out 1/8" off by the time you get to the end of the cut.
It might not be so hypothetical if you make a living working wood and have to spend extra time cleaning up edges.
It's perhaps not so hypothetical if you just spent $100 for one of those blades that is supposed to give perfect cuts, and you are getting saw marks in your wood due to misalignment.
I've seen lots of fences that are 1/16" out from front to back. All you have to do is to have a sheet of plywood hang up when you are crosscutting it, and your fence can get tweaked pretty good. I've had those old craftsman and jet lock type fences go out over 1/4" from one hangup.
No saw is going to stay perfectly set up for long if you use it much, but the closer it can be to optimal, the better, in my experience. If you set up dead on, you can have some wobble either way and still come out pretty good. If you set up with even 1/32" toe out, just a slight increase in toe out starts giving you problems with taper.Michael R
"You have to look for possibilities where there are none" Krenov
"I'm talking about the almost 10" length of the blade, not the height, since the misalignment is along the length of the fence."
Michael,
To have almost 10" of blade in contact with the wood, means you have to have almost 1/2 the blade stick out above the table of your saw. First, your trunion won't allow you to do this, but lets say you have one of those special design saws....
It means you now have to remove the throat plate, as 1/2 the arbor and 1/2 of the arbor nut, the blade stiffeners as well as washers, will all stick out above the table.
Toe-out now becomes a real minor issue, as you have to deal with all the above components, becoming part of the milling exercise and contend with 5 different dadoes added to your board after ripping.
Don't take this too serious, I'm just jokin.
On a more serious note, I use a 3hp 10" cabinet saw, set up with 1/32" toe out. When I purchased this saw almost two years ago, I purchased an el-cheapo Oldham Signature blade with it, ($45) as my first project was a workbench and I did not want to use an expensive blade. I have since purchased a couple of Freuds and a few Forrest blades and honestly after a gazillion board foot of lumber through the Oldham, I still see no difference in cut, compared to blades costing four times as much. I glue-up 8/4 directly from the Oldham combination blade and you won't see any glue line.
So far I had one kick-back, due to 12/4 Hard Maple closing the kerf at exit of the blade and that was manageable.
So perhaps you have a point, but I did what is written in my saw manual and obviously it gives satisfactory results. I don't believe in fixing something which ain't broken.
People like Freud do endless research on how to get performance from their blades, why not consult with them rather than this lengthy debate about cutting tapers etc.?
"People like Freud do endless research on how to get performance from their blades, why not consult with them rather than this lengthy debate about cutting tapers etc.?"
Actually, the original question had nothing to do with getting the best performance from a blade. I'd imagine Charles has seen this discussion and not felt a need to contribute, but I can email him and simply ask if Freud has any recommendation about fence setting.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
"Actually, the original question had nothing to do with getting the best performance from a blade. I'd imagine Charles has seen this discussion and not felt a need to contribute, but I can email him and simply ask if Freud has any recommendation about fence setting."
You had an open invitation to everyone to join this discussion...............
My sincere apologies for not posting what you wanted to read, but "toe out", cutting "tapers" due to "toe out", boards being pushed away from the fence or into the fence, due to "toe in" and "toe out", the "quality of cut", "performance from a blade", "kick backs" are all very closely related to your subject matter.
Perhaps next time post some very strict guidelines, forcing participants to conform to a very narrow field of discussion.
Good, luck with the allergies, hope you get better soon.
You better watch out. You could get kicked off Knots altogether.
Jelly, I wasn't scolding you, just making a point. You asked a question, I answered it. After all, I just "blessed your heart" in a post above, how could you be so mean? :-)forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Yes, how could you be so mean, Jellyrug? Being mean, deviating from instructions - you're not following protocol at all, my friend.
Mark,
I'll have to think about this for a while......
Let's see here, I have someone blessing my, heart, another who wishes to kick me off the forum and accuses me of being mean.
OK, I've got it.... This is all the fault of my Table saw manufacturer, as he wrote that I must toe out my fence by 1/32"
Or then... perhaps it's me, after all, all woodworkers are nice smart people. I guess I have to change my ways then....
Do you know a good shrink in CA?
"Do you know a good shrink in CA?"
There must be a lot of them!
Charles is on his way, but he's not a shrink!forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Jamie,I doubt if I can top the diatribes on the first page of this thread and I didn't have the time to read all of the posts but I'll take a stab at this. In a purely theoretical world where the saw and wood are both perfect and the saw is perfectly aligned and the fence exactly parallel to the blade you could rip the perfect wood and never have any issues. Since this situation is impossible it is always better to err on the side of caution and toe the out fence slightly because trying to set the fence parallel is impossible to accomplish with complete accuracy and toe in is BAD. The amount of toe should be proportionate to the accuracy of the method you use for the setup but should be as small as possible. I use a dial indicator and aim for a few thousandths of an inch but if you are using a tape rule you will have to set as close as you can by eye. Did I cover all the issues?
Charles M
Freud America, Inc.
"Did I cover all the issues?" Well, I was hoping for world peace, but I'll let you skate on it this time!
Thanks for pitching in Charles. Sounds like (from your post and John's) the saw companies suggest toe-out just be sure we don't toe-in. Ahhhh, isn't litigation fun.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
"Did I cover all the issues?" Well, I was hoping for world peace, but I'll let you skate on it this time!
Thanks for pitching in Charles. Sounds like (from your post and John's) the saw companies suggest toe-out just be sure we don't toe-in. Ahhhh, isn't litigation fun.
I think as long as you touch the fence in only one point (before the blade starts cutting) and the sawblade is more than 1/2 into the wood, the blade becomes the second fence. As long as the touch-down point on the fence stays the same you have a straight cut.
I think. Not sure for anything yet.
Forestgirl, your position seems to me wrong because you begin by assuming a falsehood.
Specifically, you start by assuming that (1) the wood is snug against the fence, and (2) the fence is toed out, meaning not parallel to the blade. Thus, unless I'm missing something, your assumption is that the edge of the wood is not parallel to the blade.
So far so good. But from that setup, you somehow assume that when you push the wood into the blade the cut will be parallel to the edge of the wood!
Isn't this wrong? If the fence is toed out then the cut will start out being tapered. And it will continue to be tapered both because the fence is not parallel to the blade and, as you correctly note, the blade itself is keeping the cut in a straight line.
If you draw the diagrams correctly, they will show the wood oriented toward the blade at a tapered angle at the beginning, thus resolving the argument.
Just for fun, here's a representation of Point A.
View Imageforestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Note that your drawing shows fence, blade, and board aligned parallel to each other, as far as I can tell. This might be of import later.
Now, what about my two little questions?Michael R
"You have to look for possibilities where there are none" Krenov
Yep, I made that drawing the way we normally set things up. It was just an illustration of Point A, for those who care. See my first illustration, with the featherboard, for an illustration of Point A (unlabelled) with a toed-out fence.
forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Edited 4/20/2005 1:42 pm ET by forestgirl
BTW, I'll share the picture-thing, but in an email so as not to get this thread confused.
I can go one more round this morning, then I've got to go! Tick-tock. Work beckons.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
OK, I gotta run. I just emailed you picture-posting info. If you don't receive it, that means your addy isn't up-to-date with Knots. LMK! And, send me an email if you have any ??s.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Just a thought, and I'll leave you to your textbook arguments:If blade and fence aren't parallel, then there's freedom for the workpiece to wander. If it ain't locked up, then it's loose. Reference discussion another thread re "glue line rip." If the work is free to wander a bit, e.g., not firmly against the fence through the blade's line of cut, or if it can pivot about any point of the fence, then WoodWiz' upcut/downcut marks become very evident, and "glue line" is out of the question (assuming blade quality and good condition excludes blade from argument. Ditto arbor.) If misalignment is great enough, then *any* straight line may be out the question. If it's set up dead parallel, the cut will display the blade's best potential and cut the truest possible line. The printed suggestion to deliberately set the fence for toe-out is mfgrs advice to novices so as to avoid liablility lawsuits. I've not cut a few million board feet, but I've cut a few hundred thousand, and have always tuned my fence to be as nearly parallel as possible to the blade. If there's the slightest bit of toe-in, the whole setup gives instant feedback - long before there's trouble. Have had but one sharp kickback in all these years, and that the result of too-tired sloppiness in setup.There's a penny's worth, anyway ...---John
John,Actually my advice was not based on fear of litigation but rather on shear mechanics. Here's a parallel (no pun intended) for you:
The arbor bores of our saw blades can never be "EXACTLY" on size since there is a certain amount of tolerance in everything. Even though we use the most accurate machines to produce the blades, if we attempted to come as close as possible to the stated arbor size there would be some that were actually undersized. Therefore we aim for a size that is slightly over so that on the low end of the tolerance we will be on size and on the high end wwe will be even more oversized. This is to avoid an interference fit. (The actual tolerance on a 5/8" bore is 0.0625" +0.0007/-0.0000 which means the machine is set to make the bore 0.062535")This is logic behind why I recommend toeing the fence. I don't espouse "angling" the fence, merely allowing for variance by setting for a toe out that is equal to half of the tolerance you are able to hold. This will decrease the odds of having a toed in fence which we all agree is bad.Charles M
Freud America, Inc.
Edited 4/22/2005 11:27 am ET by jackplane
Hello, Charles...I greatly appreciate your having taken the time to post those comments - it seems typical of Freud to be paying attention. I recently spec'd a custom precision replacement arbor for my saw at 0.6250 +.0000 / -0.0005, and my Freud blades stand out for bore diameter accuracy (among other equally impressive attributes)- each mounts "just barely", and exactly as wanted. My Freud SD508 has revealed itself as having been machined with remarkable accuracy in all respects, and is now a pleasure to use. Kudos!(BTW - there was no slur intended against Freud in my original remark about liability. I honestly thought that the toe-out instruction had been found in the machine's owner's manual.)I seem to have missed something in the fog. I comprehend that positive clearance, correct or otherwise, in bore vs. arbor diameters must yield circumferential eccentricity about the arbor centerline. However, I've failed to intuit how that affects the potential for binding in a plane at right angles (hopefully!) to the arbor axis. Can you help me understand?Thanks again...,,, John
John,Thanks for the kind comments. I apologize for any confusion. My post was meant to illustrate the principle of the tolerances in fence alignment by using an analogous explanation of the same principle as it applies to saw blade bores. There is no correlation between the fit of the blade on the arbor and the way the material tracks along the fence.Charles M
Freud America, Inc.
Charles..Whew! Thanks - thought I'd rounded the bend for a moment.One techie to another - see if you'd agree with this:
-----------------
Toe in is potentially dangerous. "Dead parallel" is desirable, but if acquired, it's just good luck in accumulated tolerances. Depending upon condition of blade, fence and arbor; type of measuring devices; and a host of other influences, the ability to align within a thousandth or two is in most cases remote. Although any misalignment can be shown to be theoretically deleterious, it's pointless or wasteful to spend the time and money to set up to measure and control infinitesimal variation. (This is wood machining we're talking about!)The deliberate misalignment of about 1/32" toe-out is nominally practicable, as it will cause angular misalignment of only about 0.001" per foot for a 30" fence - less for a longer fence. With more accurate measuring devices, closer is possible, but in any case "If there's going to be error, then be sure it's in the direction of toe-out, and in the least amount reasonably possible within the range of your practices and equipment."
-----------------------
All things considered, I think all else is splitting hairs for no measurable reward.Regards,---John
"The deliberate misalignment of about 1/32" toe-out is nominally practicable, as it will cause angular misalignment of only about 0.001" per foot for a 30" fence - less for a longer fence."
John, agree, just a small quibble with your numbers.
1/32 = 0.03125"
Over 30", this gives you and angle of atan (0.03125/30) = 0.06 degrees.
Per foot of a 30" fence: 0.03125/30 x 12 = 0.0125", or in fractions 1/80"
That's a bit bigger than 0.001"
Sheesh - was I flyin' too low! Just sorta let 30" and 30' be the same number. Idiot (me.) Thanks - you're absolutely correct. How about 0.001/1", or .012/ft ... which just happens to be about 1/80" ...Now I gotta rethink a little - a thou/ft is negligible - 12 times that is getting to be recognizable - as cut marks, if nothing else. I don't agonize over it, as I set up more closely than that - just don't want to be passing on incomplete info when someone asks.I hope that Charles might weigh in here and share with us the normal Freud factory specs for runout of tooth edges - both runout per se, and allowable planar deviation from one tooth to another.There's one of the reasons that I think credible glue line rip is pretty remarkable. In another thread, there was discussion of miniscule scallops from planer or jointer knives. ... yet a rip saw blade offers 20-50 little cutters with the ability to make marks...---John
John,Your assessment (and Jellyrug's math) are on the mark. At 0.001"/12" deviation toward toeing out the saw would be considered aligned. I also aim for closer and would recommend less if one has the capability of holding the tolerance.Freud blades are held to less than 0.003" runout (which is actually a misnomer since we are really referring to planar deviation of the teeth). So in the situation described the likely worst case is a tooth mark that is visible but insignificant to the glue-ability of the joint.Charles M
Freud America, Inc.
Charles, you are both gentleman and scholar. Thanks much for the info!---John
This forum post is now archived. Commenting has been disabled