MIT Microvision and Testbed

MIT Microvision and Testbed


The MIT Computer Microvision and Testbed is a system for in situ motion analysis of MEMS devices. The Microvision system has the ability to characterize MEMS devices by measuring their motion in the x, y, and z planes on the order of nanometers (nm). One example use is to find the resonant frequency of a device from a Bode plot generated by sweeping the stimulation frequency. Interferometry has also been demonstrated on this instrument using a MIRAU lens and a MATLAB script to calculate the curl.

The Mircrovision system includes the computer ( with the Microvision software, a Zeis Axioplan2 microscope, a CCD camera, a PI E-662 LVPZT controller for the microscope lens, and a joystick contolled stage. The system is enhanced by an additional testbed to supply external voltages and waveforms. The testbed includes a 34401A Digital Multimeter, two 33120A Function/Arbitrary Waveform Generators, an E3631A 80W Triple Output Power Supply, and an E3647A 60W Dual Power Supply, all from Agilent. These instruments can be controlled by National Instrument's LabView for Linux installed on the second computer (memslab2).

Usually, the Microvision system is used to test a packaged device. However, the stage is equipped with a vacuum controlled mount for testing unpackaged dies. There are four Micromanipulator Model 525 MT magnetic maual manipulators available along with Micromanipulator 7A-M probe tips (fine-tip, flexible tip and shaft).

Here are recent (December 2003) instructions for using the Microvision system.

 ... to top

Usage Policy

Currently, there is no specific usage policy laid out. Qualifies Users are allowed to access the machine as long as no one else appears to be using it. Please respect others members of the lab by cleaning up after yourself and not leaving your device set up if you won't be testing it for a period of time.

As part of the DARPA sponsored Matisse Project, the Microvision system is connected to the SuperNET to allow for remote use. An increase in remote users may require a more defined usage policy.

 ... to top


Name Office Extension Email
Chiung-Cheng Lo HH A212 Link Link

 ... to top

Qualified Users List

Name Office Extension Email
Gary Fedder
Mike McIlrath
Matte Zeleznik

 ... to top

Account Setup

To use the Microvision system, you must have an account created on cm-1 through Gripe. In addition, you can ask for your own directory to be created in the cm-1 /home/ directory to save data on the local hard drive. When you log into the machine, you'll go to your AFS ECE home directory. Initially, your shell for cm-1 will be set to /usr/local/bin/bash. You must change this to /bin/tcsh using the command 'chsh'. In your AFS ECE .login or .login.local file, you must add the two following lines...

  • setenv MICRO /usr/local/cm
  • set path=/usr/local/cm/bin (or add this to your list of paths)
If you do these things correctly, you ought to see the following lines when you log in...

Running common the system cshrc: include microvision setup
end of microvision setup

Another check is to type 'echo "$MICRO"' to see that it returns the path.

A few people have experienced problems with the windows environment. If the 'startx' commands fails, work with Gripe to fix the problem.

Access to LabView would require a separate account on memslab2. A testbed VI is still in development.

 ... to top

Standard Operating Procedures

Command: The command to start the Microvision software on cm-1 is 'cmacq'. (The command for the windows environment is 'startx'.)

User's Guide: A somewhat outdated user's guide is available at The user interface has changed some since then. It should be noted that one should SKIP the 'Sample Image' button and instead use 'Watch Motion'. There is some conflict in using both.

Tutorial: Another tutorial is available at The first part of the tutorial discusses the design of the resonator; the second part discusses the testing using the Microvision system.

The document micro_demo_script.pdf may serve as a tutorial on using the MIRAU system. This was used as a demo for a Matisse project DARPA presentation.

Koehler Illumination: Aligning for Koehler Illumination: The LED must be aligned for Koehler illumination whenever it is replaced. The procedure is explained in the document micro_koehler.pdf. Note: The standard operating voltage for the LED appears to be 2.85V. If it is necessary to test the LED using an external power supply, limit the applied voltage to near this value.

PZT Controller:The PZT controller can be adjusted by zeroing. There is a small hole labeled 'Zero' with a screw inside. With the cables disconnected, the voltage can be adjusted to shift the value.

Objectives: The objectives should be handled with care, especially when removing them from the PZT. A piece of rubber (one may be found in the Microvision toolbox) may help in loosening an objective.

LED: The LED is replaced by soldering a new LED onto the LED circuit board. This circuit board is located within the LED housing on the back of the microscope. Great care must be taken in doing this. Aligning the LED is easier if it is soldered in place flat on the board.

 ... to top

Current Maintenance

  • The PZT controller ought to be recalibrated in the near future. Regular use since the last calibration has caused the voltage to drift. It is currently maxed out on the zeroing but the motion is limited by the voltage range.

  • The LED is puttting out a limited amount of power. Currently, the optical power is not sufficient for using the MIRAU lens.
  • The Piezo controller was replaced in 3/03

 ... to top

Application Notes

A few memos (working documents) avialable from MIT at These should be used for reference and are not considered publications.

 ... to top

Useful Links

Instructions in pdf form (John Neumann, CMU December 2003).

See above for links to user guides, tutorials, and memos.

The Matisse project page is at

 ... to top


Dr. Hasnain Lakdawala (PhD, 2002) developed the phase unwrapping code for use with the MIRAU lens.

John Ramsey (MS, 2002) worked with the Matise Project during the 2001-2002 academic year.

John Neuman set up much of this help page in 2002

 ... to top

Last updated May 2002.

overview | projects | people | publications | intranet | resources         © 1998-2009  Carnegie Mellon