Lab Journal
Version 7 (Andrew Starr, 10/19/2012 11:20 pm)
| 1 | 1 | Andrew Starr | h1. Lab Journal |
|---|---|---|---|
| 2 | 2 | Andrew Starr | |
| 3 | 2 | Andrew Starr | h2. 20 Oct, 2012 - Tunnelling Current Amplifier simulation |
| 4 | 2 | Andrew Starr | |
| 5 | 7 | Andrew Starr | The tunnelling current amplifier is an example of a _transimpedance_ amplifier: it converts a current input into a voltage output. This amplifier is a critical circuit in the microscope, so I have taken advantage of available electronic simulation software (TINA-TI version 9, free to download from Texas Instruments) to test and tune the design before committing to a PCB. I have a choice between a simple, 'classic' transimpedance amplifier and a more complex - but possibly better performing - 'composite' amplifier. |
| 6 | 2 | Andrew Starr | |
| 7 | 2 | Andrew Starr | h3. Bandwidth vs noise |
| 8 | 2 | Andrew Starr | |
| 9 | 3 | Andrew Starr | In a transimpedance amplifier there is an unavoidable tradeoff between the bandwidth of the amplifier response and its output noise. The higher the bandwidth of the amplifier, the faster the response, but also the higher the output noise will be. These two parameters have the following effects on the microscope performance: |
| 10 | 3 | Andrew Starr | |
| 11 | 3 | Andrew Starr | # Bandwidth affects response time of the amplifier to a change in tunnelling current, and therefore affects the rate at which the sample can be scanned. |
| 12 | 3 | Andrew Starr | # Noise affects the vertical resolution of the tunnelling current, and therefore limits the vertical resolution of the sample data. If the noise is bad enough it may also cause problems in maintaining the tip-sample tunnelling distance. |
| 13 | 4 | Andrew Starr | |
| 14 | 7 | Andrew Starr | The design of the tunnelling current amplifier is guided by the following questions: |
| 15 | 6 | Andrew Starr | |
| 16 | 6 | Andrew Starr | * *What is the highest acceptable level of output noise?* |
| 17 | 7 | Andrew Starr | * *What is the maximum bandwidth from each design that satisfies the output noise requirement?* |
| 18 | 1 | Andrew Starr | * *Does the composite amplifier offer a bandwidth advantage that justifies its greater complexity for this application?* |
| 19 | 1 | Andrew Starr | |
| 20 | 7 | Andrew Starr | h3. Output noise |
| 21 | 7 | Andrew Starr | |
| 22 | 7 | Andrew Starr | |
| 23 | 4 | Andrew Starr | h3. Simple vs composite amplifier |
| 24 | 5 | Andrew Starr | |
| 25 | 6 | Andrew Starr | The 'classic' transimpedance amplifier design for small currents such as photodiodes etc is an op-amp with negative feedback via a single large value resistor, with a compensation capacitor to prevent oscillation due to noise gain: |
| 26 | 6 | Andrew Starr | |
| 27 | 1 | Andrew Starr | !https://opendesignengine.net/attachments/173/Simple_transimpedance_amp.jpg! |
| 28 | 6 | Andrew Starr | |
| 29 | 6 | Andrew Starr | R1 is the feedback resistor, C2 is the compensation capacitance (including the parasitic capacitance of the resistor itself). C1 represents the parasitic capacitance of the current source IG1. The output of the circuit is nominally is 1V/nA. |
| 30 | 6 | Andrew Starr | |
| 31 | 7 | Andrew Starr | Assuming 10pF capacitance for the tunnelling signal input, and using only the parasitic capacitance of the feedback resistor for compensation (around 0.3pF for a leaded resistor), the 3dB bandwidth is about 520 Hz - not great, but possibly usable. The output noise looks to be around 200uVrms. |