Understanding Phase Noise Fundamentals: White Paper - Version 01.01 - Paul Denisowski
Understanding Phase Noise Fundamentals: White Paper - Version 01.01 - Paul Denisowski
FUNDAMENTALS
2 About oscillators.............................................................................................................................................3
2.1 Oscillators and phase noise.............................................................................................................................3
2.2 Ideal oscillator..................................................................................................................................................3
2.3 Non-ideal (real) oscillator................................................................................................................................3
6 Summary.......................................................................................................................................................13
2
1 OVERVIEW
A stable frequency source is a common requirement for many electronic and most RF
devices. Phase noise is the term used to describe measurements of the short-term fre-
quency stability of these sources. This white paper provides a brief technical introduction
to phase noise concepts as well as an overview of how phase noise is measured and
reported.
2 ABOUT OSCILLATORS
2.1 Oscillators and phase noise
Oscillators are commonly used as frequency references and therefore phase noise is
commonly used to quantify oscillator stability. All real-world oscillators exhibit some
amount of frequency or phase variation, and although there are ways to minimize this
variation, it can never be completely eliminated. As discussed below, excessive frequency
instability can create serious problems in many applications, so accurately quantifying or
measuring this level of instability is very important.
t ω = 2πƒ f
t ω = 2πƒ f
Mixing is widely used in RF receivers for two main reasons. First, it is generally easier
to work with lower frequency signals than higher frequency signals. Second, mixing
allows the use of fixed frequency filters, amplifiers, etc. In a receiver, signals can simply
be mixed “down” to a convenient frequency for processing. Conversely, a signal may be
mixed "up" in frequency, for example, in the case of a transmitter that converts a lower
frequency or "baseband" signal into a radio frequency signal.
4
Figure 3: Mixing produces the sum and difference frequency of the input signal and local oscillator
Mixer
fin
fin – fLO fin fLO fin + fLO
Local oscillator
fLO
Mixer
fin
fin – fLO fin fLO fin + fLO
Local oscillator
fLO
The consequences of spectral regrowth are easily seen in wider bandwidth signals, such
as those found in LTE, 5G NR, Wi-Fi, etc. If the local oscillator has low phase noise, the
resulting signal will be mostly contained within its assigned channel (Figure 5) with very
little power leaking into the adjacent channels. But as the level of phase noise increases,
the signal will become wider and will spread further into these adjacent channels. At high
levels of phase noise (the red trace), this regrowth or “adjacent channel leakage” can
become quite severe, causing significant interference.
A local oscillator and mixer can be used to move the signals down to an IF (intermediate
frequency) for processing. The IF filter (gray) only selects the desired signal and rejects
the larger unwanted adjacent signal.
Mixer
fin
Local oscillator IF
fLO
However, if the local oscillator has excessive amounts of phase noise, energy will spread
from the adjacent unwanted signal into the IF filter, making it difficult or impossible to
recover the smaller signal (Figure 7). Local oscillator phase noise therefore should be kept
as low as possible, since this phase noise reduces both the sensitivity and selectivity of a
receiver.
Figure 7: LO phase noise reduces sensitivity when interferer power spills into IF
Mixer
Local oscillator
IF
6
4.4 Phase noise and communications systems
Phase noise can also create problems for communications systems that use some form
of phase modulation. Most modern high-data rate wireless technologies use modula-
tion schemes that are based on phase and amplitude modulation, for example, APSK
(amplitude and phase shift keying) or QAM (quadrature amplitude modulation). These
modulation schemes are often represented using polar constellation diagrams, where each
point in the constellation is a “symbol” composed of a unique amplitude and phase
(Figure 8).
Phase noise causes a rotation of the constellation, with higher levels of phase noise
causing greater rotation of the points (Figure 9). If this rotation becomes high enough,
symbols can be interpreted incorrectly, leading to bit errors or a higher bit error rate.
Spectrum analyzers are general-purpose instruments and are the traditional tool used for
measuring phase noise. In almost all cases phase noise measurements using spectrum
analyzers are performed by means of an automated phase noise measurement applica-
tion. The greatest advantage of using a spectrum analyzer for phase noise measurements
is that a spectrum analyzer is a flexible, general-purpose instrument that can be used for
a wide range of other measurements as well.
A phase noise analyzer, as the name implies, is an instrument containing specialized hard-
ware specifically designed for making phase noise measurements. Phase noise analyzers
usually have higher speed and higher sensitivity than traditional spectrum analyzers, the
increased sensitivity primarily being a result of the cross correlation method implemented
in many phase noise analyzers – this will be covered in chapter 5.5. In addition, many
modern phase noise analyzers also have other functionality used in testing oscillators,
such as the ability to measure amplitude noise and spurious emissions or the ability to
characterize voltage controlled oscillators.
Pc
1 Hz
Pn
Offset
8
5.2 Single sideband phase noise
In Figure 10, phase noise was measured at a positive frequency offset from the carrier.
Since the “sidebands” created by phase noise are usually symmetrical around the car-
rier, measured phase noise is normally the same for a given positive or negative offset
from the carrier. In Figure 11, phase noise is –70 dBc/Hz at both +10 kHz and –10 kHz
offsets from the carrier. Therefore, phase noise is normally only measured on one side of
the carrier (Figure 11) and this measurement is called “single sideband phase noise”. By
convention, positive offsets (the upper sideband) are used when measuring and reporting
phase noise.
f
–10 kHz +10 kHz
–70 dBc/Hz
f
+10 kHz
–95
–100
–105
–110
–115
–120
–125
–130
103 104 105 106
Since phase noise is undesirable, lower values in a phase noise plot indicate “better”
phase noise performance. Note that many phase noise plots have distinct “regions” in
which the phase noise graph has different slopes. These are highlighted in Figure 14.
These different regions exist because the causes or sources of phase noise are often dif-
ferent at different offsets from the carrier.
–95
–100
–105
–110
–115
–120
–125
–130
103 104 105 106
10
5.4 Spot noise
In addition to the single sideband phase noise plot, another common way of represent-
ing phase noise measurement results is spot noise. Spot noise is the numerical phase noise
result (in dBc/Hz) at one or more specific frequency offsets. Spot noise is often measured
at decade offsets, that is, offsets which are powers of ten, e.g. 1 kHz, 10 kHz, 100 kHz,
etc., although it is also possible to measure spot noise at arbitrary, user-defined offsets.
Spot noise is commonly reported in table form and is often used to verify that phase
noise at a given offset is below a specified threshold or specification value.
–95
Offset Phase noise
–115
–120
–125
–130
103 104 105 106
All measurement instruments contain local oscillators whose phase noise will be added to
the phase noise of the device under test (DUT). In cross correlation, the incoming signal
from the device under test (DUT) is routed through two “identical” measurement paths in
the instrument. These “identical” paths have independent local oscillators, each of which
has slightly different, that is, uncorrelated phase noise. When these two paths are fed into
a cross-correlation function, the uncorrelated phase noise generated by the instrument
can be minimized or removed, thus allowing a more precise and more sensitive mea-
surement of the phase noise in the signal from the DUT. Increasing the number of cross
correlations further increases the sensitivity, allowing the measurement of extremely low
levels of phase noise.
Ref 1
I/Q
Low pass filter
Low noise amplifier
…
Cross
correlation
DUT I/Q
Low pass filter
Low noise amplifier
…
Ref 2
Phase noise analyzers therefore have the advantage of being much faster, especially
when measuring close-in offsets, as well as having much greater measurement accuracy
and sensitivity when using the cross correlation method.
∫ L(ƒ)dƒ
12
6 SUMMARY
An ideal oscillator produces a signal whose frequency, amplitude, and phase do not vary
over time. Phase noise describes the short-term variations in the frequency or phase of
signals produced by real-world oscillators or other devices. Phase noise is undesirable
for many reasons and phase noise measurements are used to quantify phase noise per-
formance. Phase noise measurement results are typically given in the form of a single
sideband plot, which shows phase noise as a function of carrier offset, and as spot noise,
which is a measurement of phase noise at specific offsets. Two types of instruments are
used when measuring phase noise. Traditional spectrum analyzers often support auto-
mated phase noise measurement personalities in addition to their many other functions.
Dedicated phase noise analyzers contain special hardware designed to make very fast
and very accurate measurements of phase noise and also frequently incorporate a cross
correlation function in order to reduce the influence of instrument phase noise and maxi-
mize measurement sensitivity.
www.rohde-schwarz.com
3683467852