Dial indicator set up for table saw?

[keith ouellette]

I have my $9 dial indicator with 1" of travel. I also have a steel miter bar with bearing springs and three set screws for alignment. I drilled some holes in a scrap piece of angle iron to attach to the miter bar and the dial indicator.

How important is it for that set up (meaning the angle iron arm attached to the dial) to be square with the miter bar/slot?

I was thinking because I was going to be able to slide the whole mechanism from front to back the angle will stay the same and thus yield the same measurement. I think I learned that in geometry class long ago.

Is this correct?

[Dan Friedrichs]

You are correct. The dial doesn't need to be square to the bar, nor does the bar need to be perpendicular to the slot.

[Duncan Horner]

Yep, As far as I can work out. Actually, the further off 90 degrees you are the more sensitive your gauge will be, although the absolute readings (miter slot to blade) might be numerically off. But it's the relative measurements that matter.

[Tom Esh]

...
I was thinking because I was going to be able to slide the whole mechanism from front to back the angle will stay the same and thus yield the same measurement.

Correct. The angle is not critical unless it's way off square. For example at 45 degrees, you would be measuring only 1/2 the difference, making it that much harder to resolve small amounts.

[Bob Wingard]

Yep .. for what you are doing, the actual number is unimportant .. you just want to get the reading to be the same at two points of contact. When your needle moves, say 0.015", it will, in actuality be a bit more, because of the non-perpendicularity of the setup. If you were using it asa a test fixture, you would want it to be a perfect 90 deg, but then you would probably want an indicator costing more than $9 too. Go ahead and use it .. it will do exactly what you want of it.


<<<__ Bøb __>>>

[Duncan Horner]

Correct. The angle is not critical unless it's way off square. For example at 45 degrees, you would be measuring only 1/2 the difference, making it that much harder to resolve small amounts.

I believe that it's the opposite.

You're using a longer stroke on the dial to dead the differential if you measure on a diagonal, making it easier to resolve small differences.

Think of a 3-4-5 triangle. Where the blade is on the "3" side. The actual measurement to the blade is 4 but if you measure on the hypotenuse, you show 5. Meaning you show more dial travel for the same distance, hence if you show a deviation of .0015 on the dial, the actual deviation will be slightly *smaller*, you've used the longer measurement to exaggerate your dial reading = more sensitivity.

[Tom Esh]

I believe that it's the opposite...

I believe you're correct. As Homer would say - "DOH!" :o

[Bob Wingard]

I'm pretty sure he had it right the first time .. the greatest sensitivity is along a line connecting the shortest distance between two points (1 = the point of contact / 2= the center of the indicator body).

<<<__ Bøb __>>>

[keith ouellette]

great. Thanks guys. I got into the shop late tonight and didn't get to put everything together yet.

[Duncan Horner]

I'm pretty sure he had it right the first time .. the greatest sensitivity is along a line connecting the shortest distance between two points (1 = the point of contact / 2= the center of the indicator body).

<<<__ Bøb __>>>


The shortest distance guarantees the shortest dial travel for a given deviation. Work it out on graph paper, you'll see it. If you do it at 45 degrees, you get 1.414 times the dial travel for the same measurement, increased sensitivity.

Then again, we're agonizing over a theoretical point which raises the precision by several degrees over what we actually need here :)