Table saw set up, how accurate is accurate?

[The Bear]

Hi

I'm starting to put together my table saw I bought second hand a couple of months ago. I'm following the woodwhisper vids. First job is to align the mitre slots to the blade. I've used the adjustable set square method he describes, rotating the teeth to measure from the same tooth front and back. I've then screwed a dial gauge to my mitre gauge and done the same. It reads 2/100 to 3/00 of a mill out. Is this acceptable in general terms? Its hard to move the cast iron top with any degree of accuracy and this is the closest I've got it so far.
I want to use it to build cupboards, tables etc and be able to mitre accurately using the EB3.
I also don't want to have to redo it in a few weeks/months time because that will mean taking the cast iron wings and the rails off and having to redo the fence.

So is it accurate enough?

Mark

 

[wizer]

Acceptable.

 

[maltrout512]

If I am right with what you posted then that's a *** paper thickness. Don't see a problem with that.

 

[head clansman]

hi

if you, can adjust that any closer go for it , :lol: :wink: if not no problems. hc

 

[wizer]

Silver Regular is an Ultra Fine-Weight Paper and is extremely thin. According to Rizla, the Silver Regular paper is 20 micrometres thin as opposed to the thickness of a human hair, which is 50-100 micrometres thick and has a weight per unit area of 12.5 g/m². A normal thick paper is 16-21 g/m².

(2/100) millimetre = 20 micrometres

So yeh. About a *** paper

:lol:

 

[newt]

Although its highly desirable to have the mitre slot parallel to the blade, it does not directly influence the cutting angle. The cutting angle is controlled by the angle set on the mitre gauge.

 

[wizer]

The way that Marc sets his TS up in that video is to get the mitre slot as accurate as possible and then he sets the fence to the slot. Hence making everything true to each other.

 

[The Bear]

Thanks for the replies, sounds fairly accurate when compared to a *** paper.

I would have thought it would be more difficult to cut accurate mitres if the blade and mitre track were not parallel as I wouldn't be able to use the graduations on the mitre and would have to make adjustments to take into account the offset. Wouldn't I???

Will continue putting it together tomorrow as I have just found out I need a 16mm spanner, which I don't have.

Mark[/quote]

 

[wizer]

I don't know how you set those mitre gauges up, but I assume you can adjust 'zero' so that the graduations are bang on.

 

[Digit]

What I do wizer is this. Having eventually got to the same *** paper as Bear I use a set of plastic drafting angles and set to the blade. Saves a lot of trouble, having said that for mitres I use a sled and cut to the left and to the right of the blade alternately.

Roy.

 

[newt]

Thanks for the replies, sounds fairly accurate when compared to a *** paper.

I would have thought it would be more difficult to cut accurate mitres if the blade and mitre track were not parallel as I wouldn't be able to use the graduations on the mitre and would have to make adjustments to take into account the offset. Wouldn't I???

Will continue putting it together tomorrow as I have just found out I need a 16mm spanner, which I don't have.

Mark

[/quote]

The error of the track to the blade will not have any impact on the angle you set on the mitre gauge. Difficult to imagine but just draw it out on some paper.

this of course is exaggerated but it shows what I mean. I should stress that having the slot parallel is highly desirable for other reasons, quality of cut etc. The red line is the blade.

 

[BigShot]

Newt

If the blade presented a vertical cutting edge like a bandsaw the diagram would be spot on (subject to the exaggeration of course) but since the blade in a tablesaw presents a curved cutting edge, it strikes me that the track being parallel to the blade is a factor in the squareness of the cut.

In your diagram the cut would be square when viewed from above as in the sketch, but actually concave to some degree when viewed from one of the sides perendicular to the plan view.

I've done my own sketch - started as an elevation and morphed into a poor isometric (didn't fancy starting again) for clarity.

See what I mean?

Obviously this assumes a rather large angle between blade and track but I believe the same would happen (to a lesser degree) with a tighter angle and may be an issue if cutting wood for butt joints, mitres and the likes - potentially leaving a small gap at the back of the joint when the face is in full contact.

Your thoughts?

 

[newt]

Newt

If the blade presented a vertical cutting edge like a bandsaw the diagram would be spot on (subject to the exaggeration of course) but since the blade in a tablesaw presents a curved cutting edge, it strikes me that the track being parallel to the blade is a factor in the squareness of the cut.

In your diagram the cut would be square when viewed from above as in the sketch, but actually concave to some degree when viewed from one of the sides perendicular to the plan view.

I've done my own sketch - started as an elevation and morphed into a poor isometric (didn't fancy starting again) for clarity.


See what I mean?

Obviously this assumes a rather large angle between blade and track but I believe the same would happen (to a lesser degree) with a tighter angle and may be an issue if cutting wood for butt joints, mitres and the likes - potentially leaving a small gap at the back of the joint when the face is in full contact.

Your thoughts?

You are quite correct I was only considering the issue in the horizontal plane. It was just that spending a lot of time adjusting the top to within a few thou of parallel would be a waste of time in terms of mitre accuracy. It would be of academic interest to calculate the curve which I think would be function of blade diameter and mitre slot error. However I do not have the brain capacity to do the sums. Just out of interest, when I have used my mitre gauge at angle to set it back to 90 I just remove from the slot and adjust with a set square the fence face to the bar that engages in the slot, always spot on. edit There is of course a technique that uses the curve of the blade to hollow out wood by moving the piece over the blade at an angle, but it does look a bit dangerous to me.

 

[BigShot]

Upon reflection I think the isometric element makes it less clear so I did a quick edit to produce this cross section which better illustrates it.

It shows a pretty terrible skew between the blade and tracks that just wasn't there in the first image.

I agree that what I've shown won't affect mire accuracy in so far as measured at any point up that curve the cut angle would be constant (45° for example) - even if there might be a gap when two cut ends are putted against eachother.

I'm actually of the school of thought that says we don't need close tolerance metalworking accuracy in woodwork - but I think some inaccuracies are worth ironing out to a reasonable degree.

It would be interesting to see just how much an effect angles like this would have on real world joints - though I think they would be VERY small considering we're more realistically looking at less than a degree out rather than the huge errors we've both drawn for illustration.

I'm not sure if I have the brain power for the sums, but I certainly don't have the sums to hand nor the time to do them! AutoCAD may well make it easy to do though... I'll have a play later and see what I can come up with.

 

[frugal]

The error of the track to the blade will not have any impact on the angle you set on the mitre gauge. Difficult to imagine but just draw it out on some paper.

But doesn't that have the issue that as the sled is pushed along the mitre track the saw wil start it's cut, but then the piece to the right of the saw blade will start to be pushed against the side of the blade and have a risk of binding?

 

[BigShot]

On second thoughts - AutoCAD makes it so fast it's hardly worth not doing it.

If the angle between a 250mm diameter blade and mitre track is 0.25° the distance between the blade and track measured at the front and back of the blade (assuming the blade could be raised so its axis was coplanar with the table surface) would differ by 1.09mm (43 thousandths of an inch).
With the blade rasied to cut with the top of the blade 2mm above the wood the error on a 90° cross-cut would be as follows...

Wood thickness = 50mm
Offset between top and bottom of the cut = 1.30mm

Wood thickness = 25mm
Offset between top and bottom of the cut = 0.91mm

Wood thickness = 12.5mm
Offset between top and bottom of the cut = 0.59mm

I assume 1mm difference between the front and back of the blade is something most would correct (I certainly would).
A 0.25mm (9 thousands of an inch) difference between the front and back of the blade reduces the angle to 0.0573°

Surprisingly that still leaves an offset of 1.03mm between the top and bottom of the cut on a 50mm lump of wood.
25mm and 12.5mm wood have offsets of 0.72 and 0.47mm respectively.
(again all assuming the blade is 2mm above the top of the wood for each cut thickness)



EDITED 24/11/2009 15:34hrs
Added the blade size above and to add this note.
I've specified a 90° cross cut in my numbers above, but now I think about it I'm not sure that was necessary. I can't see any reason that a mitre cut would produce a different offset error. Measured perpendicular to the cut face I believe the errors would be the same regardless of the mitre angle.
I'm afraid I don't have the time to figure out how to draw that up in AutoCAD to check that theory though.

 

[newt]

On second thoughts - AutoCAD makes it so fast it's hardly worth not doing it.

If the angle between the blade and mitre track is 0.25° the distance between the blade and track measured at the front and back of the blade (assuming the blade could be raised so its axis was coplanar with the table surface) would differ by 1.09mm (43 thousandths of an inch).
With the blade rasied to cut with the top of the blade 2mm above the wood the error on a 90° cross-cut would be as follows...

Wood thickness = 50mm
Offset between top and bottom of the cut = 1.30mm

Wood thickness = 25mm
Offset between top and bottom of the cut = 0.91mm

Wood thickness = 12.5mm
Offset between top and bottom of the cut = 0.59mm

I assume 1mm difference between the front and back of the blade is something most would correct (I certainly would).
A 0.25mm (9 thousands of an inch) difference between the front and back of the blade reduces the angle to 0.0573°

Surprisingly that still leaves an offset of 1.03mm between the top and bottom of the cut on a 50mm lump of wood.
25mm and 12.5mm wood have offsets of 0.72 and 0.47mm respectively.
(again all assuming the blade is 2mm above the top of the wood for each cut thickness)

Very interesting thanks for that, what diameter blade did you use?. I assume if the blade is raised above the surface by a typical amount for the thickness being cut then the offset error reduces. Might try a practical experiment. But as we have said highly desirable to be parallel for a number of reasons, but not your mitre angle. Many folk also put a set square on the face of the mitre gauge and then reference against the face of the blade, this will produce an error unless exactly parallel.

 

[BigShot]

You know, I thought I'd included blade diameter in the post - apparently not.
I did it for a 250mm diameter blade, mainly because that's what I've got on my saw.
I'll edit my previous post to include that detail for reference and clarity for future readers.

I'm not sure if I made it clear in the post, but I made those theoretical cuts with an adjusted blade each time...
So the 50mm wood was cut with a blade raised 52mm above the table surface, the 25mm wood with a blade raised 27mm above the table surface and 14.5mm for the 12.5mm wood.

If the blade was run at full height for every cut the offset error would indeed be smaller (with the theoretical maximum of 125mm above the table surface - physically impossible due to the arbor - the error would be 0.14mm for a 50mm bit of wood (0.03mm for 25mm and 0.01mm for 12.5mm) with the same 0.25mm [9 thou] blade allignment error as in the latter numbers of my previous post).
That brings about the pretty bad safety aspects of a full height blade though - I'd rather take the time to adjust than risk losing a hand for the sake of +/- 0.5mm per cut.

Agreed that the blade allignment is not an issue for mitre angle - but I'd qualify that by saying it IS an isse for a good mitre joint. With that impossible theoretical best I just gave for a 50mm cut in the previous paragraph, each mitre joint (say in a picture frame) would have a 0.28mm gap at the back of each joint leading to either a frame that isn't flat, or one with visible gaps at the corners.

That does make me wonder though, is there any benefit to having a V shaped joint like that filled with glue, or is a snug joint between parallel faces and a thin layer of glue stronger?

 

[TrimTheKing]

The error of the track to the blade will not have any impact on the angle you set on the mitre gauge. Difficult to imagine but just draw it out on some paper.

But doesn't that have the issue that as the sled is pushed along the mitre track the saw wil start it's cut, but then the piece to the right of the saw blade will start to be pushed against the side of the blade and have a risk of binding?

Bingo!

Ignore all the rest of the theories (as true as they are), this is the fundamental issue that will affect your cut!

 

[BigShot]

Frugal, Mark, aye, there could well be binding in that case (though it would only actually bind if you were cutting enough material off - at best it would be a rubbish cut though).

Of course, you can cut coves on a TS by pushing the wood over the blade at an angle somewhat like shown in newt's diagram (though that was obviously an exaggerated example) http://woodgears.ca/cove/index.html and get an OK cut... how that compares to the binding in a mitre cut is beyond me though.

All this CAD, theory and I've still not actually got my workshop set up. My TS has had its run-in and I've set it up as accurately as I can at the moment, but the blade still hasn't tasted wood yet.

The theory does make not having time to get my workshop fitted out a lot easier to bear though.