Michael Juan
I'm Michael, I will be working on the litho-stepper
Last updated
I'm Michael, I will be working on the litho-stepper
Last updated
Focus for this week is to create a plan to quantify the errors in the current litho stepper. Work on my tormach 440 cnc machine to have "in house" machining capability.
Preliminary Readings: ISO 230 Geometric accuracy of machines operating under no-load or quasi-static conditions, Foundations of Mechanical Accuracy, Precision Machine Design.
Create a project proposal
Created project proposal, received input from litho-stepper team and edited the proposal by added target values for mechanical accuracy of proposed nano-positioner.
Welded a steel coolant tank using TIG (tungsten inert gas) welding. Leak tested the coolant tank by filling it with water and waiting to see if the tank leaked. There were several pinhole leaks.
Looked at the cad files for this open source piezo nano-positioner. https://www.sciencedirect.com/science/article/pii/S2468067222000621
Created spreadsheet of potential tooling required for CNC milling nano-positioner parts.
Started on creating the documentation for measuring the mechanical accuracy of Stepper V2.
Problem 1
Coolant tank was not watertight. suspected cause was contamination from inadequate surface prep as well as skill issues with TIG (tungsten inert gas) welding. An attempt was made to braze the locations of leaks but a second leak test was not attempted due to time.
Proposed Solutions
Grind out areas that have leaks, weld the leak locations again.
Buy a coolant tank.
Problem 2
Did not finish documentation on test cases due to time overrun with coolant tank task.
Proposed Solution
Carry over task to the next week.
Problem 3
Unsure about who reviews my gitbook updates as well as github project tracker usage.
Proposed Solution
Ask at the next meeting.
Measurement:
Measure the mechanical accuracy of Stepper V2.
Machining:
Order tooling.
Order material.
Tram the head of the CNC machine.
PI tune CNC machine spindle motor.
Nano Positioner:
Create a plan for machining mechanical parts of the nanopositioner.
Work on CAM (computer aided manufacturing) for one loose tolerance part.
Carry Over Tasks:
Create and document plan to measure the mechanical accuracy
It is important to note that I am following axis orientation of the Stepper GUI, which is not consistent with industry norms.
Link to testing results spreadsheet: https://docs.google.com/spreadsheets/d/1kp33Uu0bnELoj7gj8I_tnA5ZHgbyOgtRBmU556TN2yw/edit?gid=457045106#gid=457045106
Link to test procedure document:
https://docs.google.com/document/d/1GBycb5NIzfImCTrVROQhtj5_gETUNBXJAFKQSsnJRk8/edit?usp=sharing
Created a SOP for checking the parallelism of wafers.
The highest point and the lowest point were 7 microns apart. This could be due to residue from the cleaving the wafer, manufacturing tolerances, or small amounts of contamination that was not cleaned off through washing with acetone and isopropanol.
Did an informal test on contamination. Sharpie marks are around 2.5 micrometers, and finger smudges are measurable under 0.5 micrometer.
Tested the backlash on the x and y axis of stepper at 10 positions each.
The fixes for the X and Y axis by Carson resulted in less that 10 micron backlash.
The backlash on the Z axis was so bad that it over-traveled my indicator. (will test it after a proposed fix)
Tested the step accuracy of the stepper in 10 micron, 5 micron, 2.5 micron, and 1 micron increments.
Steps were consistently inconsistent. (see linked spreadsheet)
At smaller steps such as 1 micron there would often be no movement even after a command.
Did initial testing on squareness of axes. Initial testing with a 2 micron indicator and a granite square suggests that x and y axis squareness is not a issue.
Did initial testing of how parallel the vacuum held wafer was to the axes.
I did not bring my course indicators because I assumed the error would be in the micron range. The error was higher than what my indicators could measure.
Initial testing suggests that the vacuum is bending the wafer.
Performed manual PI tuning for CNC machine spindle motor.
Problem 1
Did not get to test positional repeatability of axes. repeatability of limit inductive sensors, and hysteresis of limit sensors.
Proposed Solutions
Perform test on Tuseday 02/04/2025
Problem 2
Testing by manually typing g-code is slow. Even though most of the time spent was manually adusting the dial indicator there is speedups fr
Proposed Solutions
Create a gcode script if testing is going to be an ongoing thing.
Measurement:
Test the positional repeatability of axes, limit inductive sensors, and hysteresis of limit sensors.
Converts notes from testing to repeatable SOPs.
Test longer travel distance accuracy of axes.
Machining:
Order tooling. (carry over)
Order material. (carry over)
Tram the head of the CNC machine. (carry over)
Nano Positioner:
Work on CAM for the top plate of open source nanopositioner. Because there are no tolerances in the paper create a dimensioned drawing with best guess tolerances.
Work on creating a simpler way of testing piezo nano positioning. (Sanity check)
Inaccuracies of small steps (<10 micron) seems to be from motors lacking torque to micro step. friction/binding in the system.
Mechanical accuracy of z axis is so bad that its functionally unusable. (a fix has been proposed and is currently being implemented by Carson )
The axis orientation of the stepper do not follow industry norms. This should be updated to prevent confusion.
even though the surface of the wafer is almost atomically flat. the bottom and top layer are not necessarily parallel.
The 3d vacuum wafer holder is tilted significantly.
Important Notes
The stepper components and axis orientation were changed between the last measurement and the tests performed on 2/8/2025.
Worked on documentation on test procedures for measuring stepper.
finished positional repeatability measurements of x and y axis.
finished repeatability testing of homing switches.
Attempted CAM (Computer Aided Manufacturing) on open source piezo nano-positioner.
There were some DFM (Design for manufacturing) issues that makes the design unnecessarily complicated.
Worked on a modified design for a single axis nanopositioner, (modified from https://www.sciencedirect.com/science/article/pii/S2468067222000621).
read papers on friction drive nanopositioning.
3d printed the stl of nanopositioner. (printing it feb 10 morning)
Problem 1
Did not order some tooling and material for nanopositioner.
Proposed Solutions
add items to purchase sheet before tuesday.
Problem 2
The parts specified on nanopositioner have micron level tolerances. Therefore the nano positioner when stacked on top of each other might not be anywhere close volumetrically to nanometer accuracy.
Proposed Solutions
talk to team about it. Should not be a big problem.
Problem 3
Have big due date for major courses. Have to set up things on thursday for design school career fair.
Proposed Solution
Let leads know on tuesday meeting.
Measurements:
Check with team on proposed positional accuracy of redesigned nano-positioner.
Check with team on proposed redesign of piezo nano-positioner.
Machining:
Add spotting drill, drills, taps, chamfer mills,ball endmill, roughing endmill, and collets to purchase sheet.
Nano Positioner:
Add pre-ground aluminum bar to purchase sheet
Machine bottom plate part when end mills arrive.
Will be a lot of work.
Probably going to use sacrificial workholding instead of making custom workholding.
Worked on CAD for redesigned nano-positioner
Worked on test procedures for positional accuracy.
ran test cuts on scrap aluminum
Problem 1
Tooling for machining did not arrive.
Proposed Solutions
Machine the week of feb 17
Problem 2
Did not complete CAD for nanopositioner to a suitable state to machine.
Proposed Solutions
Work on cad before presentation date.
Problem 3
NanoPositioner linear rail out of stock
Proposed Solutions
Design around an alternative THK bearing slide.
General
Prepare presentation.
Measurements Finish up documentation for positional repeatability, step accuracy. and backlash measurement testing.
Bought an LVDT probe that "should" be able to measure double digit nanometers across very short distances. If time allows will redo some measurements.
NanoPositioner
should be able to make all the parts for one axis of nanopositioner IF tooling arrives.
