Sub micron stepping test.

[mjkzz]

Using method and python software posted here:

https://www.photomacrography.net/forum/ ... hp?t=36871

I just did a sub micro stepping test -- stepping 0.3125um. Euipment used:

continuous LED light to have consistent exposure, inconsistent exposure could cause trouble for alignment process.

Nikon D5200
Nikon 200mm f/4 as tube lens
Nikon 50x ELWD objective,
surplus precision rail
rotary stage to fine tune the camera so that sensor plane is as parallel to the motion plane as possible
one linear micro stage to fine tune focus

micro stepping is set to 1/128, with 1mm pitch screw and 0.9 degree motor, this means there are 16 micro steps for 0.3125um

Living in a high rise building seems troublesome, I have to do this after 2:00am so there is no people coming in and out of elevator, no trains running (though they are about 1km away) and it is cool enough to close all windows to prevent air flow.

With 50x, stepping 0.3125um means 15.625um / 4um (sensor size is 24000/6000) = 3.90625 pixels shift for each subsequent images, no big deal for Zerene.

I still think I need a lens support, too bad I can not get Thor-lab stuff, but the result is OK -- this is 0.3125um step!!!



Setup

[perdu34]

Really lovely looking set up and I would really like to know more so I have a few of quick questions so I can understand it a bit better:

What are your axis labels? All your y-axis values seem to be off by a factor of 10, unless I'm misunderstanding it.

How are you measuring the travel? Is it simply the 3 pixel shift?

How are you controlling for motor backlash? I can't actually see the coupling to the stage.

Have you thought about setting up on a slab of granite or concrete? My old Leica SP5 confocal was set up on a cushioned granite block to stop room vibrations. It wasn't perfect but it worked 90% of the time.

[mjkzz]

hi perdu34,

the Y axis is from what Zerene reports, I believe it is fraction of width (or height , but if pixel is square, it is the same), ie, if full width is 1, then 0.005 means 0.005 of width.

I am not sure why you say it is off by factor of 10, please read that post and Rik has full explanation of what these numbers mean.

The surplus rail is a high precision one, made in Germany, it has repeatability of 0.5um. I did have to replace the motor mount to use a stepper motor instead of servo motor as I only have step motor driver. So I do expect some backlash, but it should be minimum and I did move the rail over start position, so when it starts stacking, it wlll move to start postion, thus eliminates effect of backlash.

all of these sit on top of a large granite blcok measuing 40x30x12 cm, I believe its weight is 25kg

[Lou Jost]

Very nice! Congratulations on achieving this. I hope I can do the same some day. I am going to try to do it with WeMacro's Micromate which is made to turn a microscope's fine-focusing dial:
http://www.wemacro.com/?product=micromate

[rjlittlefield]

mjkzz, thanks for running this test. The results look very good.

perdu34, let me walk through an analysis of mjkzz's numbers as I understand them.

Nikon D5200 has a sensor that is 23.50 mm wide (ref https://www.imaging-resource.com/PRODS/ ... 200DAT.HTM)

mjkzz's graph shows a shift of about 0.025 Zerene Stacker xoffset units between steps 1 and 37.

0.025 * 23.50 = 0.5875 is the corresponding shift in mm at the camera sensor.

Optical magnification is 50X, so then 0.5875 / 50 = 0.01175 is the shift in mm at the rail.

There are 36 steps between #1 and #37, so 0.01175 / 36 = 0.000326 is the shift in mm, per step, at the rail.

0.000326 mm = 0.326 microns, which is a good match to the nominal 0.3125 microns per step that mjkzz calculates from screw pitch and motor drive specs.

mjkzz, does this match your thinking?

--Rik

[mjkzz]

Thanks, Rik, exactly.

[perdu34]

Thanks Rik, that helps loads