Some data acquisition applications will inevitability run into one of the most common problems of the information age: information overload. Today, modern DAQ systems, such as the highly versatile Crystal Instruments Spider product line, can store significant amounts of vibration data. Now, memory limitations are significant. However, sifting through large amounts of data in order to capture a small transient event is undesirable.
Let us imagine an event-based capture scenario where an event or phenomenon of interest occurs, but environmental limitations, long periods of inactivity between events, or unpredictability of the event would result in excessively large time recordings. Even though data acquisition systems like the Spider-20E have several gigabytes of internal memory, trudging through long-time recordings is a tedious and unnecessary undertaking. Additionally, the sheer size of such a recording could complicate post-processing. Important tasks such as extracting frequency data, performing statistical analysis, or calculating frequency response functions (FRF’s) from the data would be more difficult.
Consider some possible scenarios where this could happen:
Scenario 1 (triggered-rearmed recording): Ground vibration data needs to be recorded near a railway to determine if the surrounding land can be zoned from commercial to residential. Trains travel the tracks at infrequent intervals. Sensors located near the tracks will constantly record the ambient conditions. The compact Spider-20 system will record the sensor data. When trains pass by, the Spider system will record a short period before a train passes, the instance when a train passes, and a portion of time when the train cars are passing. No other data will be saved to the Spider memory, even though trains pass by infrequently.
Scenario 2 (extended-pre-triggered recording): Records vibration and sound levels stemming from a satellite launching from a platform to determine whether the sensitive satellite components would withstand the rigorous launch conditions. The long period of initializing all the launch steps, the blast-off, and a brief amount of time post-launch are captured. During the launch sequence, a Spider system will constantly record sound, acceleration, and other datasets. However, all the equipment operator has to do is start the test. All relevant portions of the launch sequence up until briefly after the solid rocket boosters fire will be seamlessly saved.
Scenario 3: A machine shop runs a variety of parts across multiple machines. However, the shop structure is aging, and the overall vibration levels cannot exceed a specified high-level value. If that elevated level is exceeded, an analyst only needs to view the vibration data at the times closest to the limit being exceeded – the data for an entire day’s operation does not need to be reviewed. Additionally, saving days ‘or weeks’ worth of data is unnecessary.
How is efficient and selective data acquisition possible in such complex and diverse settings? The answer is a powerful feature called Circular Recording, readily available in EDM 10.0 software. What is the difference between circular recording, and a standard time stream recording? A standard recording saves all data for an interval of time. The circular recording takes in the same data as a standard recording, but the live recording only keeps a specified chunk of the data at any given time. Saving that chunk of data is caused by a triggered event. Let’s examine this more closely, using Scenario 1.
Once the sensors are placed near the train tracks, the Spider system will constantly monitor the vibration signals. Typically, the Spiders will record and save all the data ranging from when recording began to the end of the specified monitoring period. A single recording could be as large as several hours long. Circular recording constantly records all data, but only saves the triggered event.
In the train vibration application, the number of trains passing in a 72-hour window is unknown. The figures below illustrate the execution of a standard data collection run:
For this particular run, there are some large gaps of time when no trains are running. Due to these large gaps, this single recording would be very large and would need to be sectioned off for spectral analysis.
What would be different if the data acquisition system had circular recording functionality? The figures below illustrate the data collection process for the same scenario:
For this scenario, the circular recording feature significantly reduces the amount of memory needed to store the data collected from the 72-hour timeframe. Now, three minutes of data can be extracted from three short recordings.
The circular recording feature is just one of the many powerful components of the Dynamic Signal Analyzer (DSA) EDM software suite module. DSA works seamlessly with one of our many hardware systems, such as the Spider-20 system featured in this article. For more information concerning the circular recording feature or any of the other numerous EDM functions, please contact our support staff.