Design Tip
RF wireless burst analysis
using oscilloscopes
By Mike Hertz
D
evelopers of wireless technology have
historically relied on frequency domain
equipment as their only source of feedback
for high-speed wireless designs. With realtime oscilloscope bandwidths now topping
18 GHz, capturing and analyzing the full band
of ultrahigh-frequency RF signals is now possible due to rapid performance advancements
in real-time acquisition technology. Real-time
oscilloscopes offer unique advanced analysis
capabilities not available using traditional
frequency domain equipment for RF wireless
technologies. Advanced analysis capability
through statistical characterization techniques
and customized processing can provide greater
insight and increase design efficiency.
To enable capture of wireless RF signals,
an amplified antenna configuration can be
used. Using standard components from the
TV industry, an RG6 video coaxial cable
with F connector can be stripped to use as
an antenna, and powered inline amplifiers
connected in series to the BNC adapter on the
oscilloscope will form an amplified antenna
array capable of detecting the tiny wireless
RF transmissions nearby. The connectivity
of hardware configuration is as follows. The
receiving antenna is connected to the inline
amplifier and power supply, and this structure
is then connected via a BNC adapter to the
oscilloscope input channel to form the amplified antenna array.
Figure 1 shows something you may not
have seen before—an octal display depicting
eight views of the same real-time wireless
RF burst with varied zoom ratios. To see
both the burst characteristic and waveshape
requires a large amount of waveform acquisition memory. In the capture of the wireless RF
signal example 10 million acquisition sample
points were used. In the final zoom ratio
250,000:1, actual data sample points can be
seen along the rising and falling edges of the
RF signal. As shown, the high sample rate of
the single-shot capture is more than adequate
to accurately represent the true signal shape for
this signal. The high-level view of the entire
burst can be viewed along with the smallest
details at the same time using this multizoomed, multigrid display technique
In addition to time-domain views of an
RF burst, the frequency domain view (FFT)
can provide more insight into the fundamental frequency, harmonics and modulation
characteristics of the RF signal. Figure 2
shows several types of analysis useful for RF
60
702RFD33.indd 60
performed simultaneously on the
acquired traces. The FFTs shown
in Figure 2 are applied to specific
zoom areas.
When panned, the FFT result
can change depending on the frequency content encompassed by
the function. As the zoomed areas
are panned throughout the original
waveform region, the FFTs show
the spectral content of a user-defined
portion of the waveform, allowing
isolation of frequency content of
a portion of the burst separate
from surrounding waveform data.
Using a zoom trace as the source Figure 1. Eight simultaneous views show both the big picfor an FFT, then positioning the ture and the detailed view of this real-time RF burst.
zoom to isolate specific areas of the
waveform, will allow the frequency
domain view to isolate the portion
of the waveform of interest.
The histogram in function F8
shows the statistical distribution of
instantaneous pulse widths of each
cycle captured in the RF burst. The
track view in function F4 is plotting
the instantaneous pulse width values
for every cycle and is time-correlated
with the 5:1 zoomed view. This
track waveform has measurement
results for its Y-axis and time for
its X-axis. In this case, the track
is plotting duty cycle as a function Figure 2. Advanced frequency content analysis of wireless
RF burst.
of position, and the X-axis scaling
of the RF burst and the track are identical. math-on-math capability allows the scope
As shown by comparing the track view and to rapidly square, sum and rescale the wavezoom view, pulse width is inversely correlated forms to produce an instantaneous power
with amplitude in the RF signal. This unique output waveform. Automatic measurement
view provides insight into the modulation parameters derive the maximum and mean
characteristics. Any anomalies of the tracked values of the power waveform and can be
parameter can be easily identified in the track divided to rapidly determine the ratio of
view and isolated in the acquired waveform peak to average power of the RF signal directly
by multizooming the track and acquisition at on the scope display.
the point of interest.
Function F5 shows a custom MATLAB
ABOUT THE AUTHOR
script calculating a simple wavelet transform,
running within the scope application and
Mike Hertz is a field applications engineer
outputting its result directly into the scope’s
with LeCroy Corp. in Michigan. Before
graticule. This type of customization can be
joining LeCroy, he was an applications
used to apply digital filtering and other signal
engineer with Agilent Technologies. He
conditioning directly to the data both before
has three U.S. patents pending in the
and after the other analysis has taken place. Any
area of oscilloscope measurement deprocess that can be described algorithmically
sign. Hertz received a BSEE from Iowa
can be applied to the RF analysis process.
State University and an MSEE from the
When acquiring RF signals that contain an
University of Arizona.
I (in phase) and Q (quadrature) component,
www.rfdesign.com
February 2007
2/8/2007 1:10:55 PM