0

As part of a current-to-voltage preamplifier that I'm working on improving, I'm trying to model the current source of an scanning tunneling microscope (STM). The way it works is described below (based on my understanding), with the block diagram shown below. Given a material sample, it is biased at some voltage level. The STM has a special tip, and a current passes through the tip from the biased sample, with the STM using that current measurement as a way to view the surface of the sample. The tip starts at X=0, and it moves up/down until the computer measures a specific current. Once it sees that current, it moves on to X=1 (or whatever the next point is) until it reaches the end. Think of it as scanning in a grid-like motion.

schematic

simulate this circuit – Schematic created using CircuitLab

This block diagram may be more complicated than I'm making it, but this shows the rough idea of how it works based on my understanding. That tip current is converted into a voltage for measurement later on. In LTspice, I was simulating the current as an ideal current source, but how do I model the parasitic capacitance between the tip and the sample? Is there a standard value that is used to model the parasitic capacitance, or does it vary?

BestQualityVacuum
  • 456
  • 7
  • 32

1 Answers1

1

Capacitance of a parallel_plate capacitor is

  • C = E0 * Er * Area/Distance

You can insert values for area and distance.

The Er of air is 1, as is vacuum.

Eo is apbut 8.89e-12 farad/meter.

=============================

You might get better accuracy by using formula for round_surface (your needle) to a plane.

analogsystemsrf
  • 33,703
  • 2
  • 18
  • 46
  • Thanks for your reply. Yes, in this case, the tip can be thought of as a round plate. So, one plate is round, and one plate is square. – BestQualityVacuum Oct 20 '20 at 17:57
  • Yeah, it never really occurred to me to model capacitance like this until you mentioned it. In this case, the tip point is magnitudes smaller than the sample surface, so in the formula, we would only consider the area that is covered by the tip? – BestQualityVacuum Oct 20 '20 at 18:05
  • It depends. If the radius of the tip surface is very large compared to the distance between the tip and the plate then you can model it as a parallel-plate capacitor. If not then you need to worry about fringing from the side of the tip, which depends on its shape. – Elliot Alderson Oct 20 '20 at 19:16
  • @ElliotAlderson The numbers I'm using has a fixed distance of 1 angstrom between the tip and surface. But, as you said, it depends on whether the tip is sharp, or worst case, like a cylinder. The diameter is quite large compared to the distance resulting in large capacitances. – BestQualityVacuum Oct 21 '20 at 21:44