![]() |
||||
|
Commutation A complete scan by the multiplexer (one revolution of the commutator) produces a frame of the stream of words containing the value of each measurand. Every scan produces the same sequence of words. Only the value of a measurand is captured, not its address (name). If only the measurand’s data is captured, there is no way to distinguish the owner of one value from the next. Thus, a unique word called the frame sync is added at the end of each frame to serve as a reference for the process of decommutating the stream’s data (i.e., extracting it into individual measurand values). The example above shows a frame sync word for only 10 data words. In practice, the frame may incorporate hundreds or thousands of measurands. While this reduces overhead, it increases the amount of data lost if the frame sync is corrupted or bits of the frame are lost and the location of measurands cannot be guaranteed. In a simple commutator, each data word is sampled once per revolution at a rate compatible with the measurand with the fastest changing data. Since the rate of change of a measurand's value varies tremendously, the sampling frequency rate must accommodate it. As an example, to characterize vibration requires many more samples per second (thousands) than temperature (fractions). According to the Nyquist Theorem, you must sample data at twice the maximum frequency component for the signal to be acquired. Sampling rates of 5 times the maximum frequency component are typical. A low pass filter is used to eliminate any frequencies that you cannot accurately digitize to prevent aliasing. If we were to take a worst-case approach to sampling all measurands at the highest rate, we could expect much waste in carrier frequency spectrum and power. Sampling rates should therefore vary with respect to frequency content and be somewhat independent of other measurands with different periodic acquisition rates. Highly sampled measurands are super-commutated with multiple occurrences of the measurand in each frame. The opposite scheme occurs in subcommutation and embedded asynchronous data streams, where one position over time has multiple meanings. More on these commutation schemes appears later. Representing the telemetry stream as a continuous string of values in a diagram, while possible, is very cumbersome as shown below. In addition to the data words WD2 through WD10, you will notice the FS for frame synchronization. Frame syncs mark the end of a frame so that the original data can be reconstituted in the ground station. As you can see, it can be cumbersome to visualize the simulated serial output data in this format. An easier way to visualize data is presented in the table below and is defined in Chapter 4 of the IRIG-106 Standard. The standard includes both naming and numbering conventions of words and frames as seen below.
|
|||