Inspired by Dan Berards scanning tunneling microscope (STM),i´m trying to build one,too.Progress,fails and achievements will be shown here
just to show you that i´m still working on the STM, a quick picture of the control board. i´m very busy with my thesis, but more updates will come soon
After finishing the vibration isolation system i built the tunneling current preamp. The common approach to this is a transimpedance amplifier using an ultra low input bias current op amp and a 100M Ohm feedback resistor. The reason for this is, that the tunneling current usually is in the order of a few nA. Using a 100M Ohm feedback resistor, one gets a amplification of 100E6, resulting in a output voltage of 100mV for a tunneling current of 1nA. When using a op amp with a input bias current of the same order of magnitute, there will be huge errors in your measurement. Therefor, you want to use a op amp with, ideally, no input bias current. I have chosen the ADA4530. This op amp offers a guaranteed input bias current of only 20fA (!!!!!!!!!!!!!) at ambient temperature, low voltage noise and very low current noise. So it seems to be the right choice for this project. I implemented a guard ring around the inverting input of the op amp (grey area).
The guard ring is held at the same voltage as the inverting input (-->GND) and there is no solder mask on the guard ring. Since the guard ring has the same potential as the inverting input, no current will flow from the inverting input to the guard ring or vice versa. Any current approaching the guard ring from the outer region of the guard ring will be absorbed by the guard rings GND potential and wouldnt leake into the inverting input. On this way, one can minimize/get rid of currents leaking into the inverting input. These currents would cause a measurement error. The whole preamplifier is located on a small PCB that will be located next to the STM scanning tip (check Dan Berards STM, i basically stole his attempt because i find it very flexible). I bypassed the Input voltage heavily using 3 diffently sized caps. I´m hoping that this will offer nice bypassing and decoupling, since the power supply will be located somewhere away from the preamp.
The PCB was fabricated by OSH park and i´m very impressed by the quality of the boards!
Since OSH park sends you three boards, i built a second preamp using a 100K Ohm feedback resistor to test the amp (since i dont have a 1nA current source) using the measurement current of a multimeter when measuring resistors on a large resistor range. Multimeters typically send a few µAs when measuring large resistors. So i attached a multimeter to the input of the tunneling preamp and measured the output voltage. The multimeter i used outputs 0.43µA when set to the largest resistor range. So you would expect Vout to be Vout = -Rfb * Itunneling (with Rfb 100K Ohm and Itunneling 0.43µA) = -43mV. I didnt documented the test, but the preamp worked as it is supposed to. I havent done dynamic measurements though, just the static current coming from the DMM.
Since i live above a store which needs a lot of air conditioners for cooling purposes, i tryed to implement a fairly elaborated vibration isolation system because sometimes one can feel the building vibrating due to the air conditioner compressors. It is based on a oscillator with a very low resonance frequency. It consists of a housing constructed out of 3cm x 3cm ITEM aluminium profiles and its outer dimensions are 46cm x 46cm x 70cm. Attachted to this ITEM housing are 4 large suspension springs. Those springs are connected to a frame out of steel/aluminium which carries the heavy weights that assure a low resonance frequency of the suspension system. This weight carrier frame has a size of 30cm x 30cm and a height of approx. 12cm. The heavy weights are two aluminium blocks of approx. same size as the carrier frame. I bought one on ebay, the other one is a waste part of a workshop. Together, the two aluminium blocks weight around 20kg. Between the frame and the first aluminium block, there is a layer of armaflex for decoupling purposes. On top of the first aluminium block, there is a 6mm thick sorbothan sheet, covered by the second aluminium block. Sorbothan is a very good vibration isolator, commonly used to decouple speakers from the floor. On top of the second aluminium block is where the sample and the scan head will be located later on. I hope that this setup will give me a good enough vibration isolation, but right now i´m still very confident ;). Up to now the oscillator is not damped. I will implement a damping mechanism comparable to the one Dan Berard is using, a eddy current brake. To achieve this, i´m going to place a few stronk magnets below the weight carrying frame somewhen later.
Another big issue for operation a STM are acoustically induced vibrations (i.e. music, people speaking, the fan of your computer etc.). Therefor i bought acoustic foam. The foam is glued onto thin wooden boards that fit exactly to the "side walls" of the ITEM housing. Between the foam and the wooden boards i sandwiched a layer of very thick (25µm) aluminium foil. When the wooden boards are screwed to the ITEM housing, this will assure a nice electromagnetic (em) shielding. On this way i solved both the acoustic vibration problem and em shielding (or at least i hope so ;) ).
On the video you can see the whole setup on my desk after i excited the oscillator. The system shows a low resonance frequency of around 1Hz.