Raytheon SNV-12, Инструкция по установке и эксплуатации

Страница: 146 / 206

SNV-12 Operations Manual
5-4
Three different RX Audio throughput configurations exist:
5.7.1 A- Standard Voting With No Audio Delay.
The voter processes all signals as quickly as possible. Throughput delay is minimal, on the
order of 6 milliseconds. This is simply the duration of the process of digitizing the input,
analyzing it in the digital domain (includes measuring signal quality, notching pilot tone, etc.)
and returning to analog format.
A longer period of time expires before the system votes and starts passing the voted audio
signal. The CPM module must poll all SVM modules, and sufficient noise detector settling
time is needed to ensure an accurate initial measurement. Note that a similar detector settling
delay would be required if an analog detector, rather than DSP one, were used.
The end of the voting sequence does not have a similar processing delay; this reason is that
while the SVM must be told by the CPM to place its audio on the voted audio bus, but it does
not require CPM instruction to cut the audio off when its associated radio signals that it has
become squelched.
Note that some of the initial RX audio that enters the voter before the voting sequence starts is
clipped off. This is normally not noticed as there is typically no audio present this early in the
field transmission. If there are other delays in the system (outside of the voter), this initial
clipped audio can be problematic. This can be rectified by moving to configuration B.
5.7.2 B- RX Audio Delay; No Delay Added To The Onset Of Unsquelch Declaration
If audio delay is added, but no delay is added to the unsquelch declaration (factory default
setting with SVM SW5-4 left Off), the audio is held in a buffer for the set delay time. The
unsquelch declaration is not delayed, but the re-squelch declaration is held off for the set delay
time so that the end of the RX Audio message is not clipped.
This delay can be set to equal the time it takes to vote plus other delays later in the signal path,
thereby ensuring that all audio that reaches the voter while a site is unsquelched will reach the
voted out port. When used for this purpose, the same delay should be set to all SVMs.
Note that the vote decision is being made on the undelayed audio.
5.7.3 C- RX Audio Delay; Unsquelch Declaration Also Delayed
If SVM dipswitch SW5-4 is set to On, the unsquelch declaration is delayed along with the RX
audio. This is useful if the RX Audio Delay function is used to compensate for a slow to arrive
signal by delaying all of the “quicker” SVMs to match up with the “slow” SVM (for example,
if an SVM is getting its audio transferred via a network while the others are not). This lines up
the entire sequence of a late-to-arrive signal with any SVMs that get an undelayed input. Note
that, in this case, the signal quality reporting is delayed as much as the initial unsquelch
declaration and RX audio are delayed. This can be thought to as what would happen if an
external delay function slowed down the radio signals (audio and COR) coming into an SVM.
In Figure 5-1, the Audio Input signal at the top of the page is the input to an SVM module. For
clarity, this is shown using a hardwired Unsquelch input; though the basic functions are the
same if pilot tone is used. The UNSQ IN signal denotes the state of the COR input (or pilot
tone detect function). The Vote Out control signal shows the time when audio is available at
the SNV-12 Voted Audio Output port. Voted Out represents this audio signal.