How accurate are your cuts? How close is close enough?

[LCHIEN]

I always check my work with my calipers which I have pretty good confidence in to .001"

In making the miter indicator late last night,

I crosscut (miter saw) the aluminum, marked 1.0"; cut and measured 1.009" an error of .009 or 9 mils.

I ripped (table saw) the width of the pointer to a desired 0.335 using 2 test cuts and the micro adjuster and got 0.338" or a error of .003" or 3 mils.

I simply set the rip fence to the scale indicator and ruler on the rail for a desired 3/4 inch or 0.750 inches and ripped the width of the pointer to and was less careful because it was not critical, did not make a test cut and I got 0.745 or an error of .005" or 5 mils.

Given that 16 mils is 1/64th I routinely get better than 1/128th inch on parts that are under 6".

Am I overly nit picky or does everyone try and cut to this accuracy when making jigs and fixtures and small items?

[Carlos]

I tuned the TS to under 1/256th in every direction.

[cwsmith]

Honestly, I don't really know! I'm not 'machining' metal, I'm cutting wood. I want the angle's to be precise of course, and I measure and mark the length or width of the stock I want to end up with after I feed it through the BT3100. After the cut I use the same measuring tape (long cuts), or a stainless steel ruler if it's under 3-Ft to double check. But of course, that's just a matter of fractions, and maybe 1/32 or even 1/16th in many cases is sufficient. I measure, I set the fence, double check the distance between the fence and blade tooth and then feed the stock. Variations of the cut would not make that a finite process, but it would get me accurate within a 1/32 I hope.

On my RAS, I have a steel tape across the outfeed of my fence and use a home-made measuring stop (picture below). With that I can cut any number of pieces to length with relative precision. Using a measuring tape they all come out the same length that I've set on my indicator. However, time effects the moisture content and therefore the length of my 'stop' and I have to change it every so often. Before doing the cut, I just run the stop up to the blade and check the indicator on the fence measure, if it's '0' then I'm good, if it's less, then it's time to cut a new stop, screw it in place, adjust the mark to '0' and then cut the stop to length.

To that point, 'precision' in wood is sort of illusive, at least when it comes to measuring in thousands. It expands and contracts too much.



Another factor is the blade itself. How much is that going to flex during a cut. Soon after I bought my CMS, I noticed I couldn't cut a precise compound angle. Try as I might, the cut never came up accurate. Doing some exploring, I decided to change out the thin-kerf blade for a standard 1/8th kerf. That made a big difference, the flex was gone. I addressed this with Ridgid support, and never got an answer.

For something like the angle indicator, I think if you want absolute precision you'd have to check that thickness every few months as the season changes; however, how much change to the angle setting would occur if that small block shrank or swelled by a 64th of an inch? Probably not enough to care about, and if it did, then make the block out of some material that temperature or moisture wouldn't effect.

CWS

[twistsol]

I have a metric Incra TS/LS fence on my saw, and once zeroed I can dial in to the 1/10 mm or 1/254th of an inch which for me is completely irrelevant. Consistency and repeatability are far more important than precision. My next project will be bookcases for my office that will be 8" deep for paperbacks on one side of the room and 12" deep for other books on the other side. Plus or minus half an inch or more doesn't really matter. What is important is that all the 8" case sides are exactly the same width regardless of the actual measurement. That said, 1mm or .04 inch is good enough for my purposes.

[leehljp]

1. In general for wood I cut for 1/32 or 1/64. if a fine fitting cut, I will mark and cut in 1/64th increments. Wood moves more than that on tables and book shelves.

2. I have not started into segmented bowl turning but I know that in angles, perfection is a must or a turned bowl will reveal imperfections. Lengths of segments require perfection also. And I know some who can do just that.

3. This is not in the same category as the OP question, but I turn pens and measure with calipers to the tune of .001" regularly. And I measure sizing in increments of .001 above and below that when fitting wood pen parts. On a few occasions, I have pulled out my micrometer and measured metal parts on the pen to the tune of .0001.