Edge Finder

[Janderso]

I use an edge finder similar to the ones below almost exclusively for finding edges or centers of my part.s. I do so because, with care you can locate an edge to within +/- .0001" and because the edge finder locates and edge relative of the spindle axis, regardless of any runout you may have. I use the 1/2" diameter one most as it seems to be more sensitive.
View attachment 381764

Used properly, you approach the edge slowly and as you do you will see any wobble in the tip reduced until it is spinning true to the spindle. As you move a bit further, the tip will jump to the side. This is the point of first contact with the edge. If you note your position on the dial or the DRO, you can lift the edge finder and continue moving 1/2 the diameter of the tip and your spindle will be exactly over the edge of the work.

If you are working from the dials, you will need to know what your backlash is to determine what the dial reading will be if moving in the opposite direction (away from the edge). Backlash can be determined by locating the opposite edges of a block or hole of known size. The difference between the measured and actual size will be the backlash. Keep in mind that backlash can vary with position, especially on a worn machine.

A DRO simplifies the process since backlash isn't of concern. The position is determined by the DRO reading. Even on a worn machine, it is possible to determine position to better than a thousandth.

I have a hard time using the double ended ones.
In an ER40 collet, they get stuck.
I use edge finders all the time, single .200" spindle.

 

[ErichKeane]

I have a hard time using the double ended ones.
In an ER40 collet, they get stuck.
I use edge finders all the time, single .200" spindle.

Same problem in R8! If you get the edgefinder in too far, you have to push it upward, which is difficult.

 

[RJSakowski]

I run my edge finder in an end mill holder.

It is easy io run an edge finder past the point of first contact. I use it by coming up relatively fast to get and idea of the edge. I will zero the DRO and back off the edge then I approach the edge at a rate around .0002"/sec. With the jog function of the Tormach set to .0001"/step, this is fairly consistent. When I am able to repeat the position to .0001", I consider it good. Patience and practice are the key words.

When I have checked this method for absolute accuracy in the past, I have been good to within a few tenths. This is better than the backlash/lost motion on the Tormach can deliver.

 

[stupoty]

Fun with tools

Mr pete / tublecain does good vids on a lot of tool use like edge finders.

Ox tools has interesting info on picking up edges etc.

 

[Ken226]

When you machine a part using a dimensioned technical drawing, as I almost always do, the features location will always be defined by its position relative to a reference datum.

The edge finder let's you accurately reference your mill to that datum, then move the specified distance from the datum to the feature.

As an example, here is a simple screw hole pattern I drew up in CAD a moment ago.

example:


Here, the reference datum for the holes is the lower left corner, ignoring the fillet.

Once you've picked up the left edge with the edge finder, zero the X axis dial/DRO. Then pick up the bottom edge and zero your Y axis dial/DRO.

From there, the the spindle can be positioned for each hole quickly, accurately and repeatedly.

From your datum zero, you can run to each hole with the center drill, then again with the pilot drill, again with the tap drill, then with the tap.

Only one tool change per tool, with everything done using numbers, so your accuracy isn't dependent on the ole mk1 eyeball.

On the above drawing, the big hole is positioned 3.64" from the X axis datum and 2.00" from the Y axis datum.






Here's is another example. Most of this part was made on my PM935 manual mill. The technique I mentioned above was used for all of the holes, and for the ball detents under each index number around the curved front face.

They were cut with a 1/8" end mill, with each detent's center positioned on the curved edge and referenced from the left, upper corner for X and Y zero.

 

[projectnut]

Keep in mind in order for the edge finder to be accurate it has to be run in the proper speed range. I have several older Starrett finders that are recommended to be run between 300 rpm and 900 rpm. There is a caution on the package not to run them over 1,000 rpm in that they may fly apart. I also have a few no name ones from Shars that are recommended to be run between 300 rpm and 600 rpm.

 

[markba633csi]

The poor man's way: I mount a piece of 1/2" drill rod in the spindle then creep up to the part edge with say, a .005 feeler gauge in between. When the feeler starts getting pinched I stop. Then I remove the rod, and I know I'm 0.255" away from the edge. If I put in a 1/2" endmill I know I'm 0.005" away from touching the part, assuming the endmill is on size and hasn't been reground.
Backlash awareness essential, as always.
Sometimes I'll use a piece of paper of known thickness instead of the feeler gauge
-Mark

 

[ErichKeane]

The poor man's way: I mount a piece of 1/2" drill rod in the spindle then creep up to the part edge with say, a .005 feeler gauge in between. When the feeler starts getting pinched I stop. Then I remove the rod, and I know I'm 0.255" away from the edge. If I put in a 1/2" endmill I know I'm 0.005" away from touching the part, assuming the endmill is on size and hasn't been reground.
Backlash awareness essential, as always.
Sometimes I'll use a piece of paper of known thickness instead of the feeler gauge
-Mark

I've seen a bunch of the youtubers do that with a piece of paper (particularly for the Z axis, since there is no edge finder in that direction!), which is roughly 3 thou.

 

[Ken226]

I've seen a bunch of the youtubers do that with a piece of paper (particularly for the Z axis, since there is no edge finder in that direction!), which is roughly 3 thou.

I use the paper trick to touch off Z on round stock, or for the occasion I'm working with a tall part and long drill, and don't have the minimum 4" needed to fit my Z gauge in.

For most stuff, my Edge Technology Z touch of gauge works fantastic. There are cheaper versions out there though.



It's super easy to use. Calibrate it on the mill table with 1-2-3 blocks, set it on the part, and run the table up untill the end mill starts depressing the plunger. When the needle points to zero, your 4.000" above the part.

 

[Janderso]

When you machine a part using a dimensioned technical drawing, as I almost always do, the features location will always be defined by its position relative to a reference datum.

The edge finder let's you accurately reference your mill to that datum, then move the specified distance from the datum to the feature.

As an example, here is a simple screw hole pattern I drew up in CAD a moment ago.

example:


Here, the reference datum for the holes is the lower left corner, ignoring the fillet.

Once you've picked up the left edge with the edge finder, zero the X axis dial/DRO. Then pick up the bottom edge and zero your Y axis dial/DRO.

From there, the the spindle can be positioned for each hole quickly, accurately and repeatedly.

From your datum zero, you can run to each hole with the center drill, then again with the pilot drill, again with the tap drill, then with the tap.

Only one tool change per tool, with everything done using numbers, so your accuracy isn't dependent on the ole mk1 eyeball.

On the above drawing, the big hole is positioned 3.64" from the X axis datum and 2.00" from the Y axis datum.






Here's is another example. Most of this part was made on my PM935 manual mill. The technique I mentioned above was used for all of the holes, and for the ball detents under each index number around the curved front face.

They were cut with a 1/8" end mill, with each detent's center positioned on the curved edge and referenced from the left, upper corner for X and Y zero.

How nice it must be to have the ability and training to "draw something up in CAD".
Way over my head. I work off my doodling.
That explains a few things.