Nanopositioner - WIP

This project builds on previous work done at HackerFab, as well as concepts described in the following paper.

Based on those designs, we decided to go with a similar mechanical approach, focusing on simplicity, compactness, and the ability to iterate quickly.

What it is

This is a compact nanopositioner designed for precise motion control using a piezo actuator. The system is intended for applications like STM, optics alignment, and general nanometer-scale positioning tasks.

The mechanical design is based on linear guides using ceramic ball bearings (G5 Si₃N₄), which provide smoother and more stable motion over time compared to stainless steel balls (which tend to wear and degrade motion quality).

The current design uses:

• 2.778 mm ceramic balls

• stainless steel shafts (preferably hardened), Ø1.5 mm, length 24 mm

• thin positioning plates (~0.25 mm) to constrain ball movement

The sliding interface is based on AlN (amorphous substrate) bonded to magnets and magnetic stainless steel (grade 430). Initially, motion is not perfectly smooth, but after a short “run-in” period (self-polishing), the system becomes significantly more stable.

Vacuum-first design

From the beginning, this project is designed with vacuum compatibility in mind. That directly influences material and component choices.

We intentionally select materials that are:

• low outgassing

• mechanically stable under vacuum

• compatible with UHV-like environments (or at least not terrible in rough vacuum)

This is why we use:

• ceramic balls instead of lubricated bearings

• metal structural components instead of polymers where possible

• minimal use of adhesives (and only vacuum-compatible ones like Torr Seal where needed)

• materials like AlN for sliding interfaces

This constraint makes the design harder, but avoids having to redesign everything later for vacuum operation.

Piezo & Mechanics

The actuator used is a PK2FMP2 piezo, mounted using end hemispheres and end cups. This mounting approach helps maintain alignment and reduces unwanted mechanical stress.

One critical aspect of this design is maintaining strict axial alignment, especially along the Z-axis.

Any deviation from the Z-axis introduces lateral forces, which:

• mechanically expand the structure

• couple motion into unwanted directions

• generate significant vibrations

This is one of the main limiting factors for stability and precision, and we are actively working on improving this.

Electronics (WIP)

We are currently redesigning the electronics with a focus on reducing cost and simplifying the system.

The first version of the board used Apex PA441 amplifiers, which cost around $25 per unit. This was originally chosen to support a wide voltage range (+100V / -50V).

However, after initial testing and switching to a different piezo, such a large range is no longer necessary. The current piezo operates up to +60V, which allows us to:

• simplify the driver design

• reduce cost significantly

• improve overall efficiency

We have already designed a new power supply providing +60V / -30V, which better matches the requirements of the current setup. At the moment, we are waiting for components to complete the new amplifier stage.

The system is designed with opto-isolation (optocouplers) to separate control electronics from the high-voltage section, improving robustness and reducing noise coupling.

Current Status

This is an early prototype, but the mechanical concept is already working and gives promising results in terms of stability and resolution.

We are currently working on:

• improving electronics (cost + design)

• refining mechanical alignment (especially Z-axis stability)

• integrating an encoder (MT6835) for feedback and control