The leading supplier of cutting edge technology

Identifying and Controlling Vibration

Introduction

For the best measurement performance on the NewView™ 6000, environmental vibrations must be minimized. These vibrations can be transmitted to the instrument itself, the test part or in both. ZYGO’s MetroPro™ software is supplied with simple and useful tools for quantifying vibration. With these tools and some common sense troubleshooting approaches, it is possible to identify sources of vibration and isolate them to produce the ‘quietest’ measurement environment possible.

What does vibration do to my data?

Figure 1.

Figure 1 – Effects of vibration on the fringe pattern at top are shown in the middle image. The bottom image shows the same fringe pattern with the vibration removed

Depending on the sample surface, vibration effects can range from nearly undetectable to extremely visible and can be detrimental to data integrity. Figure 1 shows a SiC optical flat measured with 4 fringes of tilt. The first image shows the orientation of the fringes on the camera for this demonstration. The second shows the measured surface when a mechanical vibration is applied near the sample and the third shows a surface measured by the same non vibrating system.

It is quite obvious that when the mechanical vibrations are applied, a surface ripple occurs at twice the fringe pattern – i.e. there are 8 ripples seen for the 4 fringes of tilt.

Common Sources of Vibration

While there are countless possible sources of problematic vibration, there is a small set that should first be investigated when a vibration problem is encountered:

  • Improperly installed or leveled Isolation system
  • Poorly routed or tightly stretched instrument cabling
  • Proximity to air handling discharges or returns
  • Proximity to mechanical pumps or other large machinery
  • Acoustic coupling to the sample or fixture

Vibrations stemming from these common sources are often quite large and sometimes can even be visually observed on the fringe monitor. If vibration of the fringe pattern is visible on the fringe monitor, the following quick checks can help to eliminate them quickly.

Basic Vibration Troubleshooting

The first check should be the instrument isolation system. When installed and configured properly, an isolation system will effectively eliminate even very strong mechanical vibrations. The isolation system should be supplied with the recommended air pressure and all the isolator legs should be floating – even one non-floating leg is enough to transmit vibrations to the interferometer. It is also important to ensure that the legs of the isolation system are properly contacting the floor. Most isolator legs have adjustable legs or leveling feet that can be set to accommodate a floor that is not perfectly flat.

If the isolation system appears to be correctly installed, then the instrument cabling should be checked next. Good routing of instrument cables, and sample stage cables in particular, is a very easy way to avoid imparting vibrations from the environment to the system or sample. Including sufficient slack or service loops in the cable routing allows the cables to absorb the vibrations they experience, rather than transmit them into the measurementsystem.

If correcting the isolation system installation or the cable routing, then the MetroPro™ environment test may be a very useful tool for diagnosing the problem.

MetroPro™ Environment Test

Beginning with version 8.0.1, the capability to perform an environment test is included in every MetroPro™ microscope application. The MetroPro™ environment test uses a narrowband filtered light source and a tilted optical flat to determine vibration frequencies and relative magnitudes. The result of the test is a plot of vibration amplitude vs. frequency.

The large peaks at 30, 60, 90 and 120 Hz indicate the presence of a vibration problem. For this environment, a 30 Hz vibration source was placed in contact with the isolation system, and the air to the isolation system was intentionally disconnected.

Figure 2.

Figure 2 - Resulting plot in a poor installation environment



Note that even though the vibration source was at 30Hz, there are large undesirable vibrations at harmonics of this frequency – keep in mind that when trying to isolate a vibration problem, that the actual vibration source may not be at exactly the same frequencies that you detect!

To get maximum information from the environment test, one should use the fastest camera mode available on the instrument. When the measurement is performed, the resulting plot ranges from 0Hz to a maximum of ½ the camera frame rate. A table of high speed camera modes and environmental test ranges is included in the following Table.

Table 1 - Camera Modes and Environment Test Limits

Environment Test Procedure

The procedure for performing an environment test is quite simple:

  1. Open an Environment Test window in MetroPro™ and ensure that it has an Environment Test Button, Test Duration Control, and an Amplitude vs. Frequency Plot. Refer to OMP-0347 for help creating these objects if they are not present.
  2. Select and use the “focus” filter in the standard optical filter tray located on the side of the NewView™ Head.
  3. Place a SiC or other optical grade flat on the stage.
  4. Select an objective / image zoom / camera mode combination that allows enough light for good fringe visibility and maximizes the frequency range over which data will be collected.
  5. Tilt the stage so that 5-20 vertical fringes are visible on the fringe monitor; 7-10 fringes is ideal.
  6. Choose either a long or short environment test using the “Test Duration” control. A long test acquires twice the camera frames as a short test. Typically a long test is preferred except when using the 1k camera.
  7. Press the “Environment Test” button to start the test.

When the test completes, the plots in the environment test window will be populated with data. The rule of thumb is that magnitudes should be less than 1nm RMS. It is acceptable for larger values below 5-10Hz as the NewView™ is relatively insensitive to these frequencies. High magnitude peaks in the Amplitude vs. Frequency plot indicate possible vibration problems with the most sensitive region located at ½ the camera rate

Figure 3.

Figure 3 - A typical "good" installation environment test

– large spikes here will have the largest effects on measurement data. By performing the tests under varying environmental conditions such as with potentially troubling equipment turned on and off or with airflow modifications, environment test feature can be used to identify and isolate vibration sources. After correcting vibration problems, a typical good measurement environment will look similar to the plot in figure 3.

Acoustic Coupling

Occasionally, vibrations are not driven through mechanical coupling, but rather, though acoustic coupling. This type of vibration source can be driven by sample design, fixture design, and tool location just as a few examples. Using the MetroPro™ environment test, these can be diagnosed, in just as with pure mechanical vibration. Removing the fixture from the setup and locking the part directly to the sample stage if possible is a very easy way to verify that the fixture is not acting as an antenna for vibrations. If the acoustic source cannot be removed, acoustic enclosures around the instrument or the test area have been successfully employed by ZYGO and many customers.

Conclusion

Most vibration problems encountered by users can be traced to a small number of causes that can be diagnosed using the MetroPro™ environment test capability. With this tool and some simple troubleshooting methods, most vibration problems can be efficiently eliminated to produce the best possible measurement environment.

Click here for more information