Author : Scott Schmidt and Brian Cox, Aerotech, Inc.
16 August 2011
From defect detection and repair to surface mapping and profiling, applications requiring a tight relationship between sensor output and location in 2- or 3-dimensional space are many and growing. Only by polling sensor output at the correct position can data be tied to key spatial information. While some methods exist to tie part position to sensor output, these often have non-idealities that corrupt the required synchronization or inhibit accurate data collection. In all cases, automated motion is assumed to be moving either the part under test or the sensor itself.One method to tie sensor output to the position of a mechanical system is to stop the automated motion, wait for it to settle to an acceptable level, and then poll the sensor. This approach wastes time, which is a costly disadvantage in production. Also, systems with poorly tuned servo feedback loops, as well as those subject to external vibrations or other disturbances, might never settle to an acceptable value. If the output is triggered during this period of mechanical instability, measurement quality and synchronization to part position will not be optimum. To overcome this uncertainty and instability, the machine designer could be forced to specify a tighter tolerance than otherwise required, thereby increasing system cost.A second method uses tracking hardware or software, external to the system’s motion controller, to monitor position of the axes. External tracking hardware often is capable of tracking only a single axis which is not a viable option for multi-axis systems. Also, external tracking hardware often is incapable of high tracking rates, resulting in slower processing. Custom software can be used to track axes, but adds complexity to the system, limits tracking speed, and can delay the trigger due to software execution time. Hardware and software triggering solutions are often custom-designed by the user, requiring significant design time and expense while hampering overall system integration.A third method triggers the sensor based on time, and presents three problems. First, the user is tied to a time base that can be difficult to maintain, and is typically asynchronous to the part/sensor motion (the critical parameter is part position, not time). Second, this does not allow for any errors in motion. For instance, velocity regulation (which is difficult to quantify) becomes important, and any variation of velocity can cause errors in trigger location. Third, accuracy of the time base and frequencies of inputs must be chosen carefully to prevent machine counts from being missed by the motion feedback loop, leading to lower maximum speeds.Position synchronized outputA solution lies in the realm of laser processing systems. Fast shuttering of laser pulses is required for precision welding and marking applications in the laser market, closely tying the motion control system to laser output. Position Synchronized Output (PSO) uses innovative circuitry to deliver actual encoder-based positional information to the laser in real time. Because the signals are transmitted with nanosecond latency, the laser is shuttered/fired at correct positions to achieve accurate welds, cuts, or marks on the target substrate. Low latency is achieved because position information is sent by dedicated circuits. Because PSO can be configured for up to three axes of motion, the triggering output pulse can be dependent on a vector position in three-dimensional space, and not tied to one moving axis. Furthermore, encoder signals on which triggering is based are corrected with calibration tables generated at the actual system work point and applied in real time.PSO as a test and measurement solutionPrecise and accurate positional information tied to sensor data collection enhances almost all test and measurement routines and allows post-processing features such as:* Retrace to a defect location for repairs.* Post-measurement contour- or feature-mapping.* CMM or modeling of a scanned surface.* Collection of I/O states, system velocity, and dynamic error information at the instant of PSO firing.The most common implementation of PSO in a testing application is to trigger the output pulse on a fixed distance. This allows triggering a single pulse, or multiple pulses, at constant, pre-specified intervals along the travel, and is typically used to strobe the system’s sensor to begin collecting data.Array-based triggering allows the user to specify trigger points that are unequally spaced along the travel. This style of PSO firing can be used to trigger imaging sensors at precise positions for mapping or contouring operations, or when working with irregular part geometries.Another method known as “windowing” allows an input to be set when the axis is precisely within a specified position window. PSO provides up to two windows that can be used to specify two distinct areas of interest in one axis or for two-dimensional windowing. This can be coupled with fixed distance triggering to mask the triggering to certain areas of travel, and is useful when direction reversals are required during large area scan routines.PSO can also trigger asynchronously, allowing it to work as a function generator to output an arbitrary frequency with programmable duty cycle to third-party devices.ConclusionAdding PSO to an automation controller offers flexibility in interfacing with external devices and completing difficult measurements. High-speed triggering options provide precision and speed unmatched by traditional solutions, while the data capture/update features provide rapid data collection and status updates.References1. Cox, B. (2006). Precise Triggering of External Events Based on Axis Position, Design News Motion Control and Automation Supplement, Issue 10, 7/16/2006.
Print this page | E-mail this page
This isn't a paywall. It's a Freewall. We don't want to get in the way of what you came here for, so this will only take a few seconds.
Register Now