Optical Blade Alignment Checking

[MPeppard]

Here is a fixture I have been playing around with for checking the alignment of the blade in the holder.  It's a bit over kill and not something that anyone really needs, just experimenting trying to come up with a way to also measure how far off-angle the edge is from the spline. 

It works by having an optical reference point in-line with the center axis of the SVM-45's handle using an USB microscope.  If the blade's edge is centered in the jig correctly it will intersect with the reference point.
The reference point is set using a 12mm calibration rod with a 1mm post at the center and drawing a circle around it with the viewing software.

After the edge is centered, moving the focal plane up the blade you can check for off-angle by the symmetry of the blade's sides from the distance to the reference point.  Here are some examples with one blade being very close to being perfectly aligned and one with the edge centered but off-angled a little.

Just centering the edge doesn't really fix everything if the blade is off-angle, the blade also needs to be parallel with the axis otherwise the bevel grind widths can be uneven.  In reality the edge would need to be off-angle quite a lot before it would start to make a noticeable difference.

[Fernando]

I can only say that it surprises me and it seems great for those of us who like precision.

[Ken S]

Wootz (Vadim at knifeGrinders Au) uses a simple wooden block tool to measure Projection. He demonstrates it as part of several of his you tubes. If this block tool was tuned such that the knife jig is paid in parallel, a blade edge could fit into a sharp line in the stop block. If the jig correctly matched the thickness of the knife blade, turning the jig and blade over, the reversed side of the blade should also fit into that sharp line. Misalignment could be corrected with shims.

I use a standardized 139mm Projection. The block tool would not have to be adjustable. It could be fixed at 139mm, thus eliminating one required measurement are potential error.

Ken