By Jayson Wilkinson
National Instruments Corp.
Austin, Tex.
Edited by Leland Teschler
|
A light source, photodetector, and an actuator are the basic components required for aligning fiberoptic cables. Motion control and data acquisition can take place using instrumentation modules that operate from a PXI bus, basically an industrialized version of the well-known PCI bus that has been augmented with additional synchronization and control signals. |
|
The NI 8176 RT is one example of a controller aimed at real-time applications. A 1.2-GHz processor runs LabVIEW real-time control tasks and can update control loops at up to a 40-kHz rate. |
The alignment of optical components such as lasers and fiber-optic cable generally involves both coarse and fine-alignment steps. The coarsealignment step is designed to save time by quickly bringing the components close enough to perform a more accurate fine alignment.
The usual approach for the coarsealignment step is to use vision systems as a source of position feedback. This is because vision systems can capture a large field of data and quickly get two components visually aligned.
But it's not practical to hit accuracies greater than a few microns using vision alone. There must be a finealignment process to get the nanometer resolutions that many applications demand.
Fine alignment of optical components most commonly involves performing synchronized motion and optical power measurement. Traditionally, the fine-alignment process has used some type of scanning path, such as a raster or spiral-scan, combined with optical power measurements. The motion controller moves to a position, stops, records the position, makes an optical power measurement, then continues this sequence through the rest of the scan path to locate the peak power. The point of peak power corresponds to the position at which the optical components are best aligned.
The downside of this method is that you need to stop for each measurement. The repeated stop/start causes extra wear and stretches out cycle times. Fortunately, it is possible to eliminate the need for this sort of motion through sophisticated search algorithms and new technology for synchronizing motion and measurements.
One development in this category is a special real-time systems integration (RTSI) bus. RTSI, from National Instruments Corp., simplifies the task of synchronizing and correlating position versus optical power. RTSI is a special bus designed to enable high-speed synchronization and triggering between two different PC plug-in boards without external wiring.
When doing a synchronization application, one board must act as the master by generating the trigger pulse. The other board acts as the slave by receiving the trigger pulse and performing the requested action.
In the case of controlling stage position from optical power meter data, RTSI might use the data-acquisition board as the master and the motion controller as a slave. The data acquisition can be programmed to make an optical power measurement at the moment RTSI passes a trigger to the motion controller to capture the position. The chief advantage of the RTSI approach is that the motion-control and data-acquisition hardware can communicate independent of the operating system and perform as if they were one board.
For PCI-format boards, the RTSI interface takes the form of an internal 34-pin connector. Signals pass between boards via ribbon cable. But PXI boards (basically industrialized versions of PCI boards) require no such cabling because a built-in trigger bus handles RTSI functions.
Finally, the programming for such applications can take place through use of real-time facilities in LabVIEW. Routines within motion libraries for LabVIEW handle direct reads and writes from RTSI to synchronize motion. Similarly, a breakpoint output feature facilitates the programming of measurements by sending trigger that execute when a preprogrammed position is reached. All these tasks take place through high-level software commands.