6. Unit Calibration (Converting Sensor Signals to Physical Quantities)
The next approach was to "create a system where Repolyzer doesn't make users aware of the measurement process." This involved a method where hardware settings such as sensor sensitivity and voltage range, as well as software settings such as analysis conditions, could be configured without the user having to think about them.
To address the often-considered difficulty of setting up measuring instruments, such as sensor sensitivity (calibration), we have incorporated a sensor database as a standard feature.
Once sensor information is registered, subsequent measurements can be performed by selecting the sensor model, serial number, etc., simplifying the calibration process.
More recently, products with built-in sensitivity data, known as TEDS sensors, have started to be sold, and using these allows you to read sensitivity data directly from the sensor.
FFT analyzers have a variety of settings and analysis conditions, and the settings and operating procedures were not exactly easy to understand. To address this, we have centralized all measurement and analysis conditions so that they can be set from the channel palette and configuration board, and by using a graphical interface, we have ensured that the operation is the same whether it is used for dedicated FFT analysis or multi-application analysis, prioritizing ease of use and clarity.
In the initial concept (imagination) phase of development, we envisioned automatically setting the optimal values for the hardware and software based on graph data. For example, if the user wanted to perform octave analysis on sounds up to 20kHz and simultaneously perform FFT analysis on vibrations up to 2kHz, the idea was that the user would place a graph with the horizontal axis representing 20kHz octaves and another graph with the horizontal axis representing 2kHz FFT on a piece of paper, connect the sensor to the hardware (front-end), connect the hardware to a computer, and then simply press the start button to complete the analysis.
We envisioned the vertical axis of the graph automatically scaling and displaying optimally to match the measured data. However, because it's such an intelligent system, we determined that it's difficult to implement at this stage. Ideally, we'd also like a function that compares the current data with past data and analyzes notable features after the analysis is complete, but I think this falls under the realm of a "measurement assistant robot," so I look forward to its realization in the near future.
7. Signal preprocessing (bandpass filter, time-domain double integration)
In sound and vibration measurements, signal processing, such as bandpass filtering, is often performed before analysis. Repolyzer achieves this signal processing through digital preprocessing, making preprocessing possible using only a PC.
Even today, signal preprocessing is implemented using analog circuitry, such as filter units and calculus units, but Repolyzer makes digital preprocessing easily accessible.
As an extreme example, data recorded on DAT (Digital Audio Tape) at the site is passed through an analog unit after returning to the office, and the processed signal output is input to analysis devices such as FFT analyzers and octave analyzers for analysis. However, with Repolyzer, it is possible to record raw data while simultaneously performing FFT analysis on the pre-processed signal at the site.
Once back in the office, the recorded raw data can be digitally preprocessed and analyzed using only a computer.
Furthermore, because it's a digital process, even extremely small numerical values obtained through double integration have the advantage of being valid without being buried in noise.
8. Future scalability
Finally, there's the extensibility of Repolyzer as a platform. Here again, the fact that everything is processed on a PC makes a big difference. Until now, we've focused on hardware and developed software for each application, such as FFT analysis software and octave analysis software. While the ease of use of dedicated software is an important point, the operation differs for each software.
We believe that the fact that Repolyzer allows users to perform analysis and report generation with the same user interface is a major selling point, in a different sense.
For example, by adding a new "sound quality" analysis function, it becomes possible to perform not only sound quality parameter analysis but also sound quality tracking analysis using the same operating procedure.
Furthermore, a new "torque" analysis function has been added, allowing for easy analysis of the correlation between torque fluctuations and sound/vibration using the same procedure. This means that even if new analysis functions are added to Repolyzer in the future, users will not have to learn new operations, just like driving a car.
From a hardware perspective, this is addressed by multi-front-end support. This allows for measurement and report generation with the same operation regardless of the hardware, even when dealing with signals outside the scope of sound and vibration measurement, such as data loggers requiring low-speed sampling or oscilloscopes requiring high-speed sampling. This is the philosophy behind RepolyzerWorld.
Finally, Repolyzer anticipates future increases in PC processing power and performs all processing solely on the PC.
Recently, personal computers with multi-core CPUs that perform parallel processing have also been released.
Repolyzer supports this. I'm already looking forward to using an even faster Repolyzer in the future.
Repolyzer is a product that incorporates "new ideas and concepts" like this, but the problem is that it's so unique that it doesn't get recognized. I'm happy that it's starting to gain recognition little by little recently.
-Complete-
(Excerpt from the email newsletter issued on September 21, 2006)
