Design and Construction
Version 46 (Andrew Starr, 11/20/2012 03:00 am)
1 | 9 | Andrew Starr | h1. Guiding principles of this STM design |
---|---|---|---|
2 | 1 | Andrew Starr | |
3 | 13 | Andrew Starr | * Minimise the effects of external vibrations by minimising size and mass |
4 | 13 | Andrew Starr | * Minimise the effects of temperature variation by matching material coefficients of expansion as closely as possible |
5 | 13 | Andrew Starr | * Minimise the effects of electronic noise by locating the sensitive circuitry as close as possible to the signal source |
6 | 13 | Andrew Starr | * Maximise instrument flexibility and ease of use by doing as much of the signal processing and control as possible in the digital domain |
7 | 13 | Andrew Starr | |
8 | 13 | Andrew Starr | Also |
9 | 13 | Andrew Starr | |
10 | 13 | Andrew Starr | * Use open source tools for all development |
11 | 13 | Andrew Starr | * Source all specialised materials and components from readily-available online sources |
12 | 13 | Andrew Starr | |
13 | 9 | Andrew Starr | h1. Main building blocks of the STM |
14 | 4 | Andrew Starr | |
15 | 4 | Andrew Starr | h2. High frequency vibration isolator |
16 | 4 | Andrew Starr | |
17 | 14 | Andrew Starr | The STM relies on the precise control of the probe-sample gap. Any external vibration has the potential to disrupt this, so efforts must be made to isolate the instrument. Low-frequency vibration isolation is generally done with large pneumatic or elastomer damping systems. I'm hoping to be able to avoid these by making the instrument as compact and stiff as possible, thereby maximising its resonance frequency (and therefore its response to low frequency vibrations). To isolate from high frequencies, I have chosen the 'stacked plate' approach, as described and characterised in *"Low- and high-frequency vibration isolation for scanning probe microscopy" (A I Oliva, M Aguilar and Victor Sosa in _Measurement Science and Technology_, 9 (1998) 383-390)* |
18 | 14 | Andrew Starr | |
19 | 18 | Andrew Starr | The drawings for the isolator stack plates can be found {{dmsff(15,here.)}} The plates are each separated with 3 pieces of 5mm long, 5mm diameter pieces of viton rubber as described in the paper. I used an off-the-shelf viton o-ring from a local supplier and cut pieces off, securing with cyanoacrylate glue: |
20 | 5 | Andrew Starr | |
21 | 20 | Andrew Starr | !Base_disc_with_viton_spacers_ODE.jpg! |
22 | 21 | Andrew Starr | |
23 | 21 | Andrew Starr | The assembled stack: |
24 | 1 | Andrew Starr | |
25 | 1 | Andrew Starr | !Assembled_HF_vibration_isolator_ODE.jpg! |
26 | 24 | Andrew Starr | |
27 | 24 | Andrew Starr | I haven't yet characterised the attenuation of the isolator stack - more on this after I've done some testing. |
28 | 22 | Andrew Starr | |
29 | 42 | Andrew Starr | h2. Base plate |
30 | 42 | Andrew Starr | |
31 | 42 | Andrew Starr | The isolator stack will rest on a tufnol base plate, which will provide connection points for the wires coming off the instrument. These connection points will run back to the control module via JST SR-series connectors on adapter PCBs in 3 recesses on the base plate. A shallow rectangular 'trench' in the plate runs around the base of the isolator stack. This is a mating recess for a steel fabric-lined hood that will cut down acoustic and temperature disturbances, and provide a mounting port for a USB microscope (used in the initial coarse approach phase of a measurement). |
32 | 42 | Andrew Starr | |
33 | 43 | Andrew Starr | The drawing file can be found {{dmsf(30,here.)}} |
34 | 42 | Andrew Starr | |
35 | 42 | Andrew Starr | !https://opendesignengine.net/attachments/159/STM_-_Base_plate.jpg! |
36 | 42 | Andrew Starr | |
37 | 25 | Andrew Starr | h2. Piezotube |
38 | 1 | Andrew Starr | |
39 | 28 | Andrew Starr | The piezotube was purchased from EBL Products Inc. The material is EBL2 (PZT-5A). The dimensions and electrode configuration are as below: |
40 | 27 | Andrew Starr | |
41 | 26 | Andrew Starr | !Piezotube_diagram.jpg! |
42 | 25 | Andrew Starr | |
43 | 38 | Andrew Starr | I've used 36AWG telfon-coated wire (R36A705-00 from Hobby King) to connect to the electrodes, secured with epoxy glue. The electrical connection is done with Wire Glue conductive adhesive, as I didn't want to risk depolarising the piezotube by heating it with a soldering iron. The Wire Glue contacts were reinforced with a coating of cyanoacrylate. It's a bit ugly but the connections are very secure: |
44 | 30 | Andrew Starr | |
45 | 29 | Andrew Starr | !Piezotube_with_wires_attached.jpg! |
46 | 29 | Andrew Starr | |
47 | 25 | Andrew Starr | h2. Sample mount |
48 | 25 | Andrew Starr | |
49 | 25 | Andrew Starr | h2. Coarse positioner mechanism |
50 | 25 | Andrew Starr | |
51 | 37 | Andrew Starr | The microscope must have a way of bringing the probe to within tunnelling distance (i.e. nanometres) of the sample without the two crashing into each other. This requires a mechanism that possesses the following properties: |
52 | 37 | Andrew Starr | |
53 | 37 | Andrew Starr | * low mechanical vibration |
54 | 37 | Andrew Starr | * low mass |
55 | 37 | Andrew Starr | * nanometer resolution |
56 | 1 | Andrew Starr | |
57 | 38 | Andrew Starr | Two traditional approaches have been stepper motors and 'inchworm'-type piezoelecric devices. Stepper motors are heavy and require lever-reduction systems to achieve the required linear distance resolution. The inchworm devices typically use exotic materials and can be difficult to control. |
58 | 38 | Andrew Starr | |
59 | 38 | Andrew Starr | I have opted to use a Piezo LEGS linear motor from "PiezoMotor.":http://www.piezomotor.se/products/linear/piezo-legs-linear-6n/ This motor weighs only 23g, has nanometer resolution, and PiezoMotor supplies a control module that enables the motor to be controlled via USB. The drive rod is made from alumina, which means it will have a similar coefficient of thermal expansion to the ceramic holder and piezotube. |
60 | 37 | Andrew Starr | |
61 | 25 | Andrew Starr | h2. Probe holder |
62 | 6 | Andrew Starr | |
63 | 4 | Andrew Starr | h2. Piezotube and coarse positioner support |
64 | 6 | Andrew Starr | |
65 | 4 | Andrew Starr | h2. Bias voltage module |
66 | 6 | Andrew Starr | |
67 | 41 | Andrew Starr | This module has 2 functions: |
68 | 41 | Andrew Starr | |
69 | 46 | Andrew Starr | * Provide a low-level bias voltage - typically on the order of 100mV - between the probe tip and sample to generate the tunnelling current |
70 | 46 | Andrew Starr | * Generate (relatively) high voltage pulses - typically several volts - to experiment with atomic lithography |
71 | 41 | Andrew Starr | |
72 | 40 | Andrew Starr | !bias_voltage_sketch.jpg! |
73 | 39 | Andrew Starr | |
74 | 4 | Andrew Starr | h2. Tunnelling current amplifier |
75 | 6 | Andrew Starr | |
76 | 35 | Andrew Starr | A transimpedance amplifier is required to convert the tunnelling current to a voltage signal. I have opted to go with a high-bandwidth dual stage design as described in *A low-noise and wide-band ac boosting current-to-voltage amplifier for scanning tunneling microscopy (Dae-Jeong Kim and Ja-Yong Koo, _Review of Scientific Instruments_, 76, 023703 (2005))*. The design in the original paper uses a OPA111 op-amp as the input amplifier, and a OPA2111 dual op-amp for the ac recovery stage. I have opted to use an AD549 in a TO-99 case, as the TO-99 offers the ability to connect the metal case to ground potential, which should assist in screening out external noise. For the dual op-amp I have opted to use an OPA2209, as it is a much better performing device (lower noise, greater gain bandwidth, greater CMRR etc). The design files (in Kicad format) can be found {{dmsff(14,here.)}} |
77 | 35 | Andrew Starr | |
78 | 36 | Andrew Starr | !tca_schematic.jpg! |
79 | 36 | Andrew Starr | |
80 | 35 | Andrew Starr | The amplifier will mount onto the back of the ceramic support, so the signal wire carrying the tunnelling current can be as short as possible. |
81 | 34 | Andrew Starr | |
82 | 4 | Andrew Starr | h2. Control module |