Dan Gelbart's centering microscope

avivz · Dec 3, 2021

In his latest video "Workshop Tour" Dan presents his centering microscope, which seems to be as useful as his laser centering device (starting at min 3:15):

As the general idea is understandable, does anyone knows the practical implementation details or has someone built such a device?

Dan gives a clue that it is based on a car rearview wireless camera with the lens being replaced. However, in order to have such device that allows you to locate features down to 1um, what mechanical tolerances are required? which lens, how to align it? any hints are welcome.

BugRobotics · Dec 3, 2021

I built one about 7 years ago for my Gorton I22 mill. Works well. I used a red laser instead of a backup camera. Link below to the laser I used.

2pcs Focusable 3-5V 650nm 5mW Red Laser DOT Diode Module 12x35mm w/ Driver in 712012253557 | eBay.

I didn't have any adjustment for the laser angle and just set it 20° to the spindle. I designed it to fit nicely around my quill so no adjustment there either. I used a coin cell battery for the power and it turns on when it starts to rotate due to centripetal force. A switch would have worked just as well.

These guys started making an off the shelf setup: Centerquic Laser Centering Device – CENTERQUIC

mhajicek · Dec 3, 2021

I got an optical center scope, but found I couldn't reasonably get my eye on it in the machine. Thought about dong a digital one, made from a cheap USB microscope.

Antarctica · Dec 3, 2021

I don't think tolerance matters in his implementation - it just needs to be adjustable. You put the thing chuck and spin it. Adjust the x-y on the work until the diameter of the 'dot' you're looking at (in his test case) is minimized - you know that the spindle is centered. Then go back and adjust the x-y of the camera in the spindle. You won't know which whay is which at first, but if the error circle gets larger, you need to go the opposite way on the adjustment you just made.

iterate.

Is really pretty simple and brilliant if I am understanding it correctly.

Antarctica · Dec 3, 2021

Never heard of this guy before but just watched the rest of the video you posted as well as the 1st in his 18 part series. Neat stuff!

CarbideBob · Dec 3, 2021

avivz said:
. However, in order to have such device that allows you to locate features down to 1um, .

Build tolerance of no matter since it rotates.
Optics not that hard as it is just mag and standoff.
1 micron here is no way on God's green earth as you would be breaking some basic rules of physics and optics.
There is this whole deal with wavelength of light and a lens and sensor.
I'll buy a tenth but do sort of doubt that.
I know this guy highly respected but sometimes over the top a bit.
Bob

dian · Dec 4, 2021

do i understand correctly? you move the quill until you "focus" the rotating camera on the work. with that magnification you can do 1µ i guess. but what happens if you move the quill do do the operation? even that grinder wont keep 1µ not to mention a lesser machne.

Joe Gwinn · Dec 4, 2021

CarbideBob said:
Build tolerance of no matter since it rotates.
Optics not that hard as it is just mag and standoff.
1 micron here is no way on God's green earth as you would be breaking some basic rules of physics and optics.
There is this whole deal with wavelength of light and a lens and sensor.
I'll buy a tenth but do sort of doubt that.
I know this guy highly respected but sometimes over the top a bit.

Gelbert's whole approach is to achieve precision far exceeding the precision with which the components were made. This is exactly how we were able to get from two rocks and a stick precision levels to fractions of a wavelength of light.

As for the possibility of getting to tenths of a thousandth of an inch, note that what he is doing does not require resolving features that size, it requires determining the centroid of a Gaussian spot of light. Even if the gaussian spot is wider than a tenth, one can find the center of that spot with far better precision that the spot size.

Search

Largest Manufacturing Technology Community on the Web

Search

Dan Gelbart's centering microscope

avivz

Cast Iron

BugRobotics

Stainless

mhajicek

Diamond

Antarctica

Hot Rolled

Antarctica

Hot Rolled

CarbideBob

Diamond

dian

Titanium

Joe Gwinn

Stainless

Similar threads