Jackson Labs PHASE STATION 53100A, User Manual

Page: 141 / 225

PhaseStation 53100A User’s Manual V1.04 25-Apr-2022 141
Two-Port Residual Measurements
Residual noise and stability measurements with the 53100A can be conveniently performed in the
Measure single DUT with single external reference
or
Measure single DUT with cross ADEV statistics
role. Typically a two-way 0° RF splitter is used to feed a signal from a common reference source to both
the REFERENCE jack and the input of the device under test. The device’s output, in turn, is connected to
the 53100A’s DUT input jack. To the extent the device contributes its own phase drift, phase noise, or
AM noise to the stimulus signal passing through it, the impact will be evident on the respective
measurement traces.
Residual measurements are an integral part of the 53100A production and maintenance processes.
Each 53100A ships with a test report showing the result of a two-hour residual measurement with 5
MHz and +15 dBm applied to the DUT and REFERENCE input ports via a Mini-Circuits ZFSC-2-1 splitter.
The following notes will be helpful when conducting your own proof-of-performance measurements.
- At 50 Hz ENBW, residual ADEV at 5 MHz should be in the high E-14s at t=1s. Thermal effects
may push the slope out somewhat, especially in an unstable environment or when warmup time
has been insufficient. A look at the
p)hase difference graph will often show a consistent phase
slope for the first 30-60 minutes of a measurement made after a cold start.
- Typical results near 5E-15 at t=1s are achievable using the cross ADEV role at 5 Hz ENBW. This
technique can improve the short-term measurement floor by up to 10x in cases where the
53100A’s ADC noise is either limiting the available performance or is being measured itself.
- Residual PN at 5 MHz should be below -145 dBc/Hz at 10 Hz after a few minutes.
- With a strong, clean input signal at 5 to 10 MHz, the residual PN floor in the 10 kHz – 1 MHz
decades will usually end up near -180 dBc/Hz after less than two hours. Expect lower
performance near either end of the supported frequency range. When possible, both proof-of-
performance tests and actual two-port device measurements should be made in the 5-10 MHz
range where the 53100A’s own residual performance is optimal.
- If the PC can support overlapped FFT processing, consider enabling this option by checking the
Additional options → Overlapped acquisition box in the 53100A acquisition dialog. Very low
close-in noise levels will require much less measurement time when overlapping is enabled.
- Phase hits and large outliers in the frequency-difference trace are never normal in a residual test
with a clean source signal. They should be investigated if they recur with no obvious
explanation. Spurs in phase noise plots should not exceed -100 dBc in a 5 MHz residual
measurement, and will normally be much lower.