Create high-speed control systems for MEMS micro-shutter testing with NI LabVIEW FPGA

The James Webb Space Telescope (JWST) is NASA's next-generation large telescope. It is more ambitious than the predecessor Hubble Space Telescope, which NASA will place at a stable Lagrange point about 100 kilometers from Earth. The telescope will be the cornerstone of NASA's understanding of the universe and the theory of the Big Bang.

A near-infrared spectrometer developed by NASA and developed by the European Space Agency (ESA) is the main instrument on the telescope. It observes thousands of distant galaxies to detect galaxies formed early in the universe. In order to measure a large number of weak celestial bodies, the instrument must simultaneously observe multiple celestial bodies in previously unknown locations.

使用NI LabVIEW FPGA创建高速控制系统以用于MEMS微快门测试

To observe celestial bodies at these locations, NASA developed a micro-shutter array, a 171x365 array of 100x200 micron shutters that can be turned on under random access control conditions. The 2x2 matrix of four micro-shutter arrays contains approximately 250,000 programmable shutters, allowing the infrared spectrometer to simultaneously measure more than 100 faint stars, proportionally improving the efficiency of major scientific facilities. The system is critical to the development of micro-shutter arrays and is an important indicator of flight qualification for this major international mission array.

What is a micro shutter?
The micro-shutter is a 100x200 micron rectangular switch that uses on or off to block or allow light to pass. The shutter is bent with silicon nitride as a shaft, magnetically driven by means of a magnetic coating, and electrostatically latched by an electronic connection.

When we started this project, manufacturing shutter arrays was a new and complex process that was still in development. NASA manufactures shutter arrays with 365 columns and 171 rows, each containing a total of more than 62,000 shutters. We mounted the shutter array on the substrate and attached the array to the grid so that we can address each shutter with rows and columns. To open the shutter, we let the magnet pass the front of the array and apply a high voltage across the rows and columns of the shutter. The magnetic field will open the shutter and the static charge at the intersection of the row and column will keep the shutter open.

Each shutter array is manufactured to test some aspect of the overall design. Testing this device allows us to further define the manufacturing process. Using the NI PCI-7344 four-axis stepper motor controller and the NI MID-7604 power motor driver, we developed software for controlling vacuum chambers, shutter control instruments, cameras, and other equipment to evaluate array performance.

Tests on the system have shown that an uncontrolled shutter closure limits shutter performance. In this uncontrolled method, the shutter can be closed by cutting off the power of the shutter row or column, but in any case, it affects the shutter of the shutter, significantly shortening the life of the shutter.

The development team decided to use the moving magnet to close the shutter in a synchronized manner so that the magnetic field can buffer the effect of the shutter closing. The test room, completed in 2005, includes a new synchronous on-off feature.

Micro-shutter control system In different shutter designs, the micro-shutter must be able to work reliably for more than 100,000 cycles. Unlike long-term testing, the shutter cycle in the new test chamber is very fast. Since the motor speed is as high as 240 rpm, the magnet connected to the motor using a centrifugal cable moves back and forth in front of the shutter array four times per second. During magnet movement, the control system needs to latch or release 365 columns of the micro-shutter array. To understand the required accuracy and speed, each column of the shutter array can be imagined as a 1 foot wide slat in a 30 foot long fence. The magnet is like a jet that flies over 700mph and is only 3 feet away.

In order to control the shutter, we need to communicate with the control magnet and a custom high voltage shift register. The new system also requires fast communication and testing and functional verification of the various operations of the 584 chip. The system must meet all of these requirements and need to be safe in the event of a failure. A test takes several days and turns all 62,000 shutters on and off 240 times in one minute. If the system is out of sync, it will cause damage to the shutter within a few minutes.

To meet these needs, we can design and manufacture custom chips or use the LabVIEW FPGA Module. We chose a PXI chassis and a controller with a PXI-7813R reconfigurable I/O module and shutter control using the LabVIEW FPGA Module.

Control Design The entire system consists of a master computer that controls the test room, a field programmable gate array (FPGA) main program running on the PXI controller, and FPGA software running on the PXI-7813R. Using the FPGA's main interface, engineers can calibrate the system and execute manual control commands, create and download bitmaps written to the array, and self-test and diagnose other functions of the 584 chips.
The FPGA software reads the position of the magnet from a quadrature encoder or an absolute encoder. The codec algorithm is placed in a 40MHz single cycle to ensure it does not miss any steps. After filtering to eliminate jitter, the position value is written to the first in first out memory (FIFO). Another cycle on the FPGA reads the data from the FIFO and determines the shutter operation based on the current position of the magnet. The state machine communicates with 584 chips and uses the protocol to turn on or off the required rows and columns.

When the FIFO overflows, the state machine that controls the micro-shutter will not run fast enough. The software will send a synchronization error signal to the host computer so that the system will shut down.

This algorithm works well and has become the basis for shutter array control experiments. As engineers develop new algorithms to improve shutter operations, we can easily add or change algorithms in the state machine block diagram.

The LabVIEW FPGA Module and PXI-7813R save hundreds of man-hours and thousands of dollars in overhead for custom chip development. At the same time, we were able to modify the control algorithm at a lower cost to improve testing, find shutter problems, and further develop NASA MEMS micro-shutter arrays.

Mini PC Core I7

Intel Nuc I7,Mini Pc I7,Intel Nuc 10 I7,Mini Cpu I7

Guangdong Elieken Electronic Technology Co.,Ltd. , https://www.elieken.com