
SR1 Operation Manual
186
© 2014 Stanford Research Systems
be one, two, or three unit intervals long. If the one-two-three pattern is completely unrecognizable the
digitizer will abort processing at this point. If the one-two-three UI pattern is recognized, the digitizer
does a more careful measurement of the zero-crossing positions and attempts to recreate the original
clock signal by minimizing the jitter at one of three user-selectable points throughout the input record.
Selecting "Preambles" in the Jitter Detection control instructs the Digitizer to reconstruct the clock
based on minimizing the jitter at the trailing transition of the 3-UI pulse found in every preamble of the
digital audio signal. Because this transition occurs at the end of the longest stable interval in the signal
(remember that all pulses in the digital audio signal are one, two or three UIs long and the 3-UI pulses
only occur in preambles) pulse pile-up effects due to limited bandwidth are at a minimum at this
transition and the "Preambles" selection yields the jitter that is most representative of the intrinsic jitter
of the transmitting device rather than cabling effects. Since preambles occur only once every 32 bit-cells
in the digital audio signal the "sampling frequency" for the calculated jitter will be 1/32 of the bit-cell rate
and the number of calculated jitter points will be equal to the number of preambles in the input record.
This implies that the bandwidth for jitter measured with this selection is (1/2 of 1/32) or 1/64th of the
original bit cell rate which is equal to the sampling rate of the embedded digital audio signal.
Selecting "All Bits" reconstructs the clock by minimizing the jitter at each bit cell transition. This method
has the maximum measurement bandwidth for the calculated jitter signal. The jitter bandwidth will be 1/2
of the bit cell rate or 32 times the embedded digital audio sampling rate. Because the jitter is being
sampled at each bit cell transition the calculated jitter will include contributions from both the intrinsic
jitter of the transmitter as well as pulse pile-up effects due to bandwidth limiting in the transmission
channel. Since during preambles there are bit cell boundaries where no actual transition occurs the
digitizer interpolates the position of these "virtual" transitions.
The "Stable Bits" selection is a midpoint between the "Preambles" and "All Bits." The jitter for this
setting is calculated at every 4th bit-cell starting from the first transition in the preamble. The jitter
bandwidth is thus 1/8th the bit cell rate or 8 times the embedded audio sampling rate.
For non digital audio signals (square waves) the available selections are "Square Rising", "Square
Falling", or "Square Both". Square Rising or Falling calculate the jitter on the respective edges of the
clock signal. The Jitter bandwidth for these selections is 1/2 the square wave frequency. "Square Both"
calculates the jitter on both clock edges leading to a jitter bandwidth equal to the square wave
frequency.
After reconstructing the clock, the digitizer calculates the jitter at the user specified points in the input
record by comparing the zero-crossing of the reconstructed clock with the actual measured zero
crossings in the input record. The total rms jitter and the effective embedded audio sampling rate are
calculated and displayed in the "Measure" tab of the Analysis box. The jitter sampling frequency and the
number of points at which the jitter is calculated are shown in the "Advanced" tab.
Asymmetry
Several physical processes can lead to a situation where negative going pulses in the carrier signal have
systematically different widths from the positive going pulses. For pulses that are roughly trapezoidal in
shape an offset in the signal will cause this effect as shown below:
Содержание SR1
Страница 5: ...Part I Getting Started Audio...
Страница 7: ...Getting Started 7 2014 Stanford Research Systems...
Страница 12: ...SR1 Operation Manual 12 2014 Stanford Research Systems...
Страница 27: ...Part II SR1 Operation Audio...
Страница 156: ...SR1 Operation Manual 156 2014 Stanford Research Systems Passband Group Delay of Elliptical Filter...
Страница 258: ...SR1 Operation Manual 258 2014 Stanford Research Systems...
Страница 272: ...SR1 Operation Manual 272 2014 Stanford Research Systems on the amplitude sweep...
Страница 289: ...SR1 Operation 289 2014 Stanford Research Systems...
Страница 290: ...Part III SR1 Reference Audio...