Performance Evaluation
52
CG635 Synthesized Clock Generator
Table 23: HP 54120A digitizing scope setup for CMOS timing
Parameter Setting
Trigger
0.0V level, positive slope, HF sens. off, HF reject off
Trig. Slope
Positive
Volts/div
50 mV for channel 4
Offset
125 mV for channel 4
Time/div 200
ps
Delay
Reference at center, value adjusted to center transition in the display
Display
Channel 4, averaged with count of 4
V Markers
Marker 1: 50 mV
Marker 2: 200 mV
T Markers
Start Marker: adjusted to time where waveform crosses V Marker 1
Stop Marker: adjusted to time where waveform crosses V Marker 2
The scope delay will have to be adjusted manually until the rising edge of the CMOS
output transition from the 200 MHz output is located. Once the delay is properly adjusted
to center the transition on the display, the falling edge transition can be readily obtained
by adding 2.5 ns to the delay.
Measure the rise-time by adjusting the T Markers so that the Start Marker is located at the
point where the waveform crosses V Marker 1 and the Stop Marker is located at the point
where the waveform crosses V Marker 2. Record the
¨
t of the markers as the rise-time of
the given waveform.
Add 2.5ns to the reference delay to center the falling edge transition in the display.
Measure the fall-time by adjusting the T Markers so that the Start Marker is located at the
point where the waveform crosses V Marker 2 and the Stop Marker is located at the point
where the waveform crosses V Marker 1. Record the
¨
t of the markers as the fall-time of
the given waveform.
The measured transition times should meet the specifications given in Table 24.
Table 24: Maximum allowed transition times for CMOS output
Output Meas. (Rising, Falling) ps
Max. ( Rising, Falling) ps
CMOS
(1000,
1000)
Frequency Synthesis Tests
Basic functionality of the CG635 is verified by testing the generation of several specific
frequencies from DC to 2 GHz. This ensures that all the dividers are working properly.
Overall stability of the CG635 is tested by measuring the phase noise, jitter, and wander
at a couple of specific frequencies. The test points ensure that both crystals used in the
frequency generation are operating at specification. At the factory, we use a proprietary
mixer and filter which allows us to measure the jitter directly with a voltmeter. Here we
provide an alternative method which should provide equivalent results.
