* GadgetBuilder.com *     © 2013 by John Moran

Edge Finder Adjunct          

Last Modified:

Edge Finder Adjunct

Click to Enlarge

The Starrett edge finder gets considerable use with my mill because it's easy to use and, for me, repeats better than a wiggler. A minor wrinkle when using the edge finder is that when it kicks, the spindle is actually 0.100" from the edge. One must then account for this 0.100" offset by adding it to the desired dimension or offset the DRO zero appropriately.

The gadget seen in the picture above is an adjunct for use with the edge finder to provide an edge 0.100" from the work edge it is held against. This places the spindle directly above the work edge without requiring an offset calculation or re-zeroing the DRO. Clearly this isn't a necessary piece of gear - just a convenience to reduce errors and, in some situations, speed up the process. (Another tool for the lazy man....)

Build Notes

Edge Finder Adjunct - Parts Initially I thought of making it from one piece but this would require a very careful approach to final dimensions since it needs to be within a tenth or two in order to be useful. While possible, it seemed easier to make it in two parts so if I overshot slightly I could recover easily. A mic reading tenths or better is needed; I used a Harbor Freight electronic mic which seemed adequate.

The two part design ideally has one part with a raised section 0.1000" above the other; a simple flat is then attached to this raised section to provide two flats 0.1000" apart. Size is arbitrary -- about 0.7" wide seemed reasonable. I had a scrap 3/4 x 1/4 x 3" so it was convenient to mill a 3/8" wide step across one end to a thickness of 0.110" and then mill the higher 7/16" wide section next to it 0.210" thick, i.e. about 0.100" difference in thickness. Prior to milling the material was ground on both sides with the surface grinder to ensure the 3/4" wide sides were parallel. The work was then gripped vertically and a 1/16" side cutting slitting saw was run along the division between the two heights to smooth the junction and also undercut the 0.110" thick part next to the junction to 0.095" (approximately). This undercut is to allow the surface grinder wheel to cover the area that needs to be flat without contacting the side of the wheel. The machined section was separated from the stock via the bandsaw.

The flats on this part were then ground to about 1 thou thicker than the target thickness (i.e. 0.101" and 0.201"). This part is small so a 1/16" flat scrap was used adjacent while grinding to ensure the work didn't move on the magnetic chuck. Unfortunately, my surface grinder needs new motor bearings so vibration leaves a hammered looking finish. I finished these surfaces by lapping them on fine carbide paper against a flat surface to remove the hammer look. This needs to be done carefully since any tilt added will reduce accuracy. Actual thickness isn't critical since it is the difference between the thicker and thinner sections that is critical. Overshooting by a tenth or two on either flat surface is accommodated by lapping the other surface to achieve the desired 0.1000" difference in thickness.

The second part is simply a flat piece of steel. I used 1/8" gauge plate and ground both sides because it had surface rust. Two 4-40x1/4" screws hold the parts together. Debur carefully on all edges and around the screw holes - even slight burrs will affect accuracy. Similarly, clean the mating surfaces carefully prior to assembly.

User Notes

Verification requires locating the edge of a part in the normal way, offsetting. Then using the adjunct, verify that locating with it matches the offset result of the original method. This will typically require multiple trials, possibly with averaging of results. I find my edge finder typically repeats +/- 0.5 thou - there is some difference in edge finders and also in how each of us use them. If you have a Moore edge finder that can be used as a standard of comparison to figure out how accurate your edge finder is for you and also perhaps to calibrate this gadget. Unfortunately, the Moore unit is very expensive and also slower to use else we'd all be using them ;-)

If you have a comment on this site or its contents, click here scroll down and click again.

Valid XHTML 1.0 Transitional