Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

STM control: Help needed pls advise ....

Hi all,

a brief introduction to this STM control would be using 2 AO channels to control piezo-crystal movements in the XY plane to raster a surface.

a third piezo in the Z axis has its voltage controlled by a analog feedback circuitry so we employ an AI channel to read the voltage.

a topographic data is hence obtained and viewed either as intensity or a 3D representation.

However, as one can see in my n00b (newbie) efforts, the raster is very much restricted to point-by-point acquisition which is terribly wrong (as much as possible no sudden movements must be made. )
Some of the more ideal scanning patterns include

  • left to right, down one row, right to left, down another row, left to right...
    -->  V
    V <--
    --> V
    i think you guys get the gist
  • left to right, right to left, down to next row ...
    -->
    <--
    V
    one can either choose to ignore the right to left part or plot a symmetrical mirror image (such an implementation is good at act as a control for noise)
  • one can also visualise the X voltage as a sine wave with Y as a step function with smooth (not sudden) increment of voltage

would my current coding allow for such raster patterns? how might I go about doing so?
any advice would be much appreciated and many thanks in advance!
Download All
0 Kudos
Message 1 of 5
(3,662 Views)
I'm curious and fascinated by your problem, but I'm also a little confused (I too am a n00b!).  Could you briefly explain the STM control and provide a little more information into what exactly you're trying to do? 

Also, you mentioned analog output channels, but I don't see any in the code you've attached in the jpeg...
0 Kudos
Message 2 of 5
(3,638 Views)
hi there.

you can see the local variables being controlled by the inputs on the top part of the block diagram. as we have not calibrated the system, the range of movement is somewhat proportional to the change in voltage applied to each crystal.

therefore
if one thinks of a 64 by 64 pixel picture and would like the X and Y piezos to scan through 5V worth of distance each, each pixel corresponds to 5/64 V of displacement



swngiam would like offer a brief into to STM:

STM or scanning tunneling microscopy is a non-optical magnification technique which at best gives atomic resolution electrically conductive surface.

by quantum theory, a a small magnitude (think nanoAmperes) of "tunneling current" tunnels through a very small physical gap between the surface-of-study and an atomically sharp tip when a potential differnece is applied between them.

a feedback logic enables the tip to respond to changes in surface topography when you raster the tip in the XY position; by maintaining a constant tunneling current. when distance increase, tunneling current decreases, voltage across Z-piezo is increased to maintain the original tunneling current

since the tunneling current changes (exponentially) with tip-surface distance, if one plots out the Z-piezo's voltage (Analog In, controlled by the feedback mechanism) one gets 3D surface data.


HOWEVER, the problem with my VI is its movement are all sudden as the transition from point to point is not smooth.

the danger is that the change in displacement (for eg in the X direction) is too fast for the feedback to "safely" avoid large topographic changes. if the tip comes into physical contact with the surface henceforth, it would 'crash', rendering the tip un-useable.

i have employed the DAQ express VI to feed in my voltage data and read my voltage data; sure tell-tale signs of newbie-ness here.
0 Kudos
Message 3 of 5
(3,632 Views)

To my experience the hart of a program for STM is the line scan (trace and retrace). You should look for examples on simultaneous analog out + analog in at some clocked rate. Once you have a single line working correctly you may build your raster line after line and display it in an intensity graph.. Of course you may undersample to reduce probe "jumping" (for exaple write and read 5k values/lines and store one in ten ( perhaps after digital filtering the 5k vector)).

Note 1: If you plot two consecutive scans of the same line in an x-y plot you will appreciate, by similarity, if your probe is tracing the surface correctly: if this is reasonably true  all the electronics, including feedback, is working ok.

Note 2: there are good, free, packages available for STM data elaboration: a WEB search before deciding the format for saving data on disk may save you coding  time later on.

0 Kudos
Message 4 of 5
(3,582 Views)
To avoid the chopiness that you will get when you single step the piezo using the for loop with "large" voltage steps, you may want to try using a waveform with a linear voltage ramp. It will be much quicker (can easily get khz line traces up to the limiting piezo resonance) and the voltage increment can be (depending on how you setup the waveform) extremely small so it won't be choppy motion. Just simultaneously acquire the tunneling current signal during the motion and your set.

Message Edited by SurfaceSurfer on 05-05-2007 02:11 AM

0 Kudos
Message 5 of 5
(3,489 Views)