Accuracy question

[tfrayne]

After aligning my Mk V 520, I felt good about the results. However, last night I needed to use the table saw to cut exactly 1/8" off of both sides of a workpiece, such that the piece was 1/4" shorter but my laminations still centered.

One one side, I was just ever so slightly off (nothing critical) and ended up with an offpiece about as thin as a piece of paper. Initially, I was thrilled that the saw was able to accomplish this. Then I measured it with a pair of digital calipers.

The piece was 25" long. One side measured 0.0135". The other side measured 0.0090"

That makes my fence off square to the blade by 0.045" over 25".

Question - What is the standard for being square? Is this difference worth looking into or should I be happy with it?

[twistsol]

First, check your math. 0.0135 - 0.0090 = 0.0045 not 0.045. Second, that is extremely accurate for woodworking in my opinion. I doubt any amount of fiddling with it would improve the results and you would likely end up further off.

[tfrayne]

OK - Sounds good. Promise me you won't tell Ms. Boler, my 5th grade math teacher!

[mountainbreeze]

Someone correct me if I am wrong.

Ideally, the distance from the leading edge of the blade to the fence during a cut remains constant, even if the fence were not parallel to the blade. If the distance from the trailing edge of the blade to the fence is a bit greater, wouldn't the cut still yield a board that has parallel edges? If this is true and the board had parallel edges to begin with, then the variance in the thickness of the cutoff would be because the distance from the leading edge of the blade to the fence changed during the cut. This is easily accomplished on a Shopsmith due to the flex in the table, (especially if the main table is not tied to the auxiliary table) by slight changes in side-to-side pressure during the cut.

[dusty]

Someone correct me if I am wrong.

Ideally, the distance from the leading edge of the blade to the fence during a cut remains constant, even if the fence were not parallel to the blade. If the distance from the trailing edge of the blade to the fence is a bit greater, wouldn't the cut still yield a board that has parallel edges? If this is true and the board had parallel edges to begin with, then the variance in the thickness of the cutoff would be because the distance from the leading edge of the blade to the fence changed during the cut. This is easily accomplished on a Shopsmith due to the flex in the table, (especially if the main table is not tied to the auxiliary table) by slight changes in side-to-side pressure during the cut.

When I need to make precision cuts such as described here, I consider it an absolute necessity to tie the main table to the extension tables and to ascertain that all the locks are secure especially the main table and rip fence.

Curious - why was it so necessary to be so accurate.

[dusty]

After aligning my Mk V 520, I felt good about the results. However, last night I needed to use the table saw to cut exactly 1/8" off of both sides of a workpiece, such that the piece was 1/4" shorter but my laminations still centered.

One one side, I was just ever so slightly off (nothing critical) and ended up with an offpiece about as thin as a piece of paper. Initially, I was thrilled that the saw was able to accomplish this. Then I measured it with a pair of digital calipers.

The piece was 25" long. One side measured 0.0135". The other side measured 0.0090"

That makes my fence off square to the blade by 0.045" over 25".

Question - What is the standard for being square? Is this difference worth looking into or should I be happy with it?

OFF SQUARE??? or asymmetric.

[tfrayne]

Good point. I will do some experimenting with the table vs table + extension table. I suppose the "5 cut method" used on crosscut sleds could just as easily be done with the rip fence. I got off on this tangent by an interesting but unintentional shaving.

I am currently building a crosscut sled for my Mk V. Before I lock anything down, I want to go through my alignment again. Just to be sure.

[dusty]

Building a cross cut sled. Very important and well worth the effort. Just make sure your table is square to the miter bars and parallel to one another. Take your time...it will payoff in the end.

Be aware that as you push the sled through the cut that by the time the work piece has cleared the saw blade the miter tracks have begun to disengage with the table top (the miter bars/tracks). This results in some instability and thus accuracy of the cut. Too deep a cross table can work against you.

[RFGuy]

After aligning my Mk V 520, I felt good about the results. However, last night I needed to use the table saw to cut exactly 1/8" off of both sides of a workpiece, such that the piece was 1/4" shorter but my laminations still centered.

One one side, I was just ever so slightly off (nothing critical) and ended up with an offpiece about as thin as a piece of paper. Initially, I was thrilled that the saw was able to accomplish this. Then I measured it with a pair of digital calipers.

The piece was 25" long. One side measured 0.0135". The other side measured 0.0090"

That makes my fence off square to the blade by 0.045" over 25".

Question - What is the standard for being square? Is this difference worth looking into or should I be happy with it?

Just so that I am clear...you performed alignments on your Mark V and then did a crosscut or rip cut of a board? After this you measured the offcut (shaving) to determine its thickness? Is this correct? If so, I guess I would have to ask why check the offcut? There is no guarantee that the board you started with had square ends, so the offcut will combine a multitude of errors. Also, no guarantee that your Mark V is perfectly aligned and if you used the miter gauge on it, no guarantee that the miter gauge is perfectly aligned either.

I assume you have a good quality square that you used to perform alignment of the Mark V (including aligning the miter gauge). After making this cut, I would have checked the end of the board that was cut and make sure a square cut was obtained by eyeballing it with your good quality square. I would ignore any crosscut or shaving that came off of this cut. You could make successive crosscuts and measure the width of them as you describe, but honestly I think you will be disappointed in the repeatability of this even though the cut is likely going to be "square enough" for woodworking purposes. JMO.

P.S. With a crosscut sled it is relatively easy to perform the 5 cut method to check for alignment if so desired. Also, as Dusty pointed out make sure you tied the main and extension tables together for rigidity using the connector tubes. Otherwise the main table can rather easily be gently pulled out of alignment during operation due to main table flexing.

[JPG]

When contemplating using the 5 ecut method, one must consider the purpose(determining the amount of correction to obtain square cuts with a fence on a cut off sled. Other scenarios are not likely relevant.

I cannot conceive using 5 cuts to align a rip fence.

For checking a miter gauge squareness, a straight edged workpiece against the miter gauge fence, then a single cut, then stand the two ends just created with the straight edges parallel with each other. Observe the amount the straight edges misalign as the distance from the cut increases.(0 = square)