Digital Angle Gauge

[Father Haskell]

That depends on the job.

You want as much accuracy as possible when cutting parts for jigs or fixtures. Other than that, allowing yourself fair tolerances speeds work.

[Tom Veatch]

I certainly won't argue against that point.

But agonizing over the Tilt Box deflecting the blade something on the order of 0.025° (inverse sine of .002/4.5) when the accuracy of the box is such that the tolerance on the measured angle is almost 10 times that amount? It just seems to me that agony could be put to better use elsewhere.

Tom Veatch Wichita, KS USA

[Upscale]

A little transfixed about this aren't you? It's wood. From minute to minute it's affected by expansion and contraction.

I've cut what I'd consider to be really tight tolerances in wood, but the most I've ever needed has been 1/64" exactness. Maybe you don't consider that very exact, but it's worked well for me as I suspect with most people.

[Father Haskell]

I just use whatever square or bevel is handiest. Dim the overhead lights, set a lamp on the saw table, and try to kill the peep of light between the blade and the gauge. According to Starrett's engineers, if you have good eyes, that puts you within 50 millionths of an inch of spot-on.

[Lew Hodgett]

Think a custom cut triangle qualifies as a bevel, doesn't it?

The "line of light" technique is quite accurate, but my preference is to compare freshly cut scrap piece against the complementary angle on the table as opposed to the gauge angle against the blade.

If there is an error, it creates a "line of light" you could drive a Mack truck thru.

(OK, a Peterbilt for the Texans)

Lew

[Father Haskell]

Be sure to use a violet light, since it has the shortest wavelength still visible to human eyes.

Press both halves of the joint together and measure the dihedral along both pieces. That'll multiply the error lots more.