Nanopositioner
Introduction
Highly accurate positioning systems are integral for semiconductor manufacturing. Right now the HackerFab uses a custom version of the Open Micro-Manipulator with additional support for rotation, heating, and vacuum.
Goals
Positional Resolution
<1um
Global Positional Accuracy (e.g. repeatability of 0,0 after restart)
~5um after homing.
Rotational Resolution
1/12 deg resolution
Global Rotational Accuracy
1/2 deg resolution
Heating Range
50 - 200 degrees Celsius
Heating Accuracy
0.5 degree Celsius
Vacuum Strength
50mN of holding force
CMU Progress
2026 - Spring
Goal Status
Positional Resolution
<1um
~1um resolution but highly affected by vibration
Global Positional Accuracy (e.g. repeatability of 0,0 after restart)
~5um after homing.
Not scientifically measured.
Rotational Resolution
1/12 deg resolution
<= 0.01 deg
Global Rotational Accuracy
1/2 deg resolution
Not scientifically measured.
Heating Range
50 - 200 degrees Celsius
Tested up to 150 degrees. Not externally validated.
Heating Accuracy
0.5 degree Celsius
0.5 degree accuracy up to 150 degrees (higher not tested). Not externally validated.
Vacuum Strength
50mN of holding force
58nM of holding force
Summary
During Spring of 2026 the CMU nanopositioner team was able to produce a working integrated system that contained all general features including 3 DOF movement, rotation, heating, and vacuum. However, there are still 3 main issues with the current implementation:
Vibration resilience
Packaging and Setup
Firmware reliability
Vibration Resilience
Due to the nature of moving at sub-micron levels the micro-manipulator is extremely sensitive to vibrations. The above gif was taken at a 10x object lens where each line is ~2um wide. As you can see the movement is significantly larger than the goal of 100nm movement accuracy.
We have many theories about what could be causing it but have not done enough scientifically rigorous testing. We were able to greatly improve the vibrations with tennis balls but it was not enough.
To analyze the vibrations and we used an opencv flow detector; however, we are limited by the fps of the recording camera.
Packaging And Setup
The current micro manipulator took over a month to be built and is very susceptible to things getting unplugged/twisted/broken/etc. The system should be more robust to handling and easier to put together
Firmware Reliability
Right now there are various firmware reliability issues which could be improved. They do not stop the device from functioning but make it harder to use and less reliable. So far we have found the following issues:
Axis "freezing". Sometimes one of the axes will stop responding to commands
Bad nano-positioner homing. Sometimes after homing a axis it will not respond to commands
This might be related to the above "freezing" issue
Improper rotation homing. Sometimes if the rotation stage is really out of alignment it will fail homing
This should be resolved but still to be fully tested
Last updated