A change is coming in the tools for measurements in both pulsed RF aerospace/defense and I/Q vector-modulated communications application. Whether for multi-channel analysis or for wider analysis bandwidth, high-bandwidth oscilloscopes are taking the place of traditional spectrum and signal analyzers. That’s because they can handle signals with spectral content beyond 1 or 2 GHz. These signals are being created to support the higher resolution requirements in radar systems and move the vast amounts of information in new communications systems.
So how do you create a powerful, wideband RF measurement suite? By coupling a high-bandwidth real-time oscilloscope with RF analysis software. Once you’ve married the two, you achieve a number of enhancements:
- Noise reduction through digital down-conversion
- A wide range of vertical scaling options, including linear and log magnitude
- Key RF measurements including occupied bandwidth (OBW) and power spectral density (PSD)
- Vector demodulation options for communications formats like QAM16
- Analog demodulation options including AM, FM and PM
- Set-up of segmented memory capture
- Statistical pulse analysis
Pulse amplitude, frequency, phase, and FFT measurements
For radar and electronic warfare applications, it’s helpful to perform a variety of measurements on many pulses. This includes things like amplitude variation, frequency, and phase shift across pulses, and a view of the spectrum of signals. For applications such as aircraft warning receivers, you also want the capability to measure time difference and phase difference between pulses associated with the capture of a wave front by multiple antennas on an aircraft. Let’s consider some of these measurements.
In the simplest case, you can measure the basic pulse amplitude, frequency shift, and phase shift across the measured RF pulse. The RF pulse train is sampled by the oscilloscope and then digitally down-converted to reduce noise and allow further signal processing.
For example, in Figure 1, a 15-GHz carrier, 2-GHz-wide linear FM chirped RF pulse signal is shown after vector signal analysis (VSA) processing. Here’s what the image shows:
- 2 GHz-wide spectral content of the signal (upper left);
- Real part of the down-converted I/Q data (lower left);
- 2-GHz-wide linear FM frequency chirp seen across the RF pulse (upper right);
- parabolic phase shift seen across the RF pulse (lower right).
These measurements are taken in the “Vector” measurement mode.
Basic vector mode analysis of FFT, real part of I/Q, FM chirp, and phase shift across pulse seen
Single channel, segmented memory capture, statistical RF pulse analysis
The next level of analysis requires a shift into “Pulse Analysis” mode. Here we use multiple oscilloscope channels to capture RF pulse signals into segments of oscilloscope memory. These are digitally down-converted into baseband I/Q signals, and then evaluated for single and multiple channel pulse analysis. For single-channel measurement, you can make three comparisons:
- the linear FM frequency chirp to an ideal, best-fit linear FM chirp signal;
- the phase shift across a pulse to a best-fit parabolic phase shift profile;
- the amplitude of the pulse envelope to a best fit ideal straight-line best fit reference.
In Figure 2, you’ll see these comparisons being made between measured to reference, and then the “error” between the measured and reference is expanded in vertical scale for a close view.
A pulse table also displays RF pulse parameters, including an RMS error calculation between the measured frequency or phase across the pulse, compared to a best-fit reference signal. It’s also possible to show statistics for the measurements over all the pulses.
Single-channel spectrum, amplitude, phase, and frequency measurements vs. best-fit reference signals
Dual-channel delta pulse amplitude, frequency, and time-delay measurements
You can also make “two-channel delta” measurements, as shown in Figure 3. These measurements are becoming increasingly important in applications such as aircraft warning receiver testing, where multiple signals are being captured from multiple antennas. The time delay and frequency difference of arrival between wave fronts must be measured for angle-of-arrival calculations.
Notice in this example a 1 nsec time delay being measured between two RF pulses. You’ll also see a 0.2-dB difference in amplitude and a 16-kHz difference in frequency, on average.
Pulse analysis is also performed on three of ten captured pulses that are being placed into oscilloscope memory segments. The parabolic phase shift across pulses (lower left), the linear FM chirp frequency shift across pulses (middle right), and the pulse envelope of pulses (upper left) are superimposed for signals coming into two oscilloscope channels. As in the previous example, each scope-channel measured signal can have the measured, reference, and error signal calculations made. Finally, the FFT spectral content for both scope-channel captures of the two pulse trains (center left) is also shown.
Two-channel measurements of RF pulse characteristics including time, amplitude, and frequency difference between two channels
Cross-correlation between pulses for precise time-delay measurements between RF pulses
In the aircraft warning receiver example mentioned previously, you can determine very precise measurements of time delay between RF pulses captured on different antennas on an aircraft by using a cross-correlation measurement between pulses. In Figure 4, a 50-psec time difference of arrival (TDOA) is being measured between two RF pulses captured on two scope input channels. Here pulses have a 10-GHz carrier, 100-MHz-wide linear FM chirp modulation, and a 1-usec width. In this measurement, you can first remove the channel-to-channel skew between oscilloscope channels, including cable delays at the temperature measurements will be taken, through de-embedding. Then a measurement can be made to see the actual time shift between the captured signals. Measurements show a mean delay of 50 psec, with a peak-to-peak variation in delay between 47 psec and 53 psec.
Math function used to measure phase shift between two RF pulses
The difference in phase between two RF pulses is also critical in a variety of radar/EW/warning receiver-oriented applications. Through the use of math functions, the measured phase across one pulse can be subtracted from the measured phase across a second pulse, measured on two oscilloscope input channels. We can measure the same two linear FM chirp signals from the last example to view the phase shift between the two pulse trains by comparing related pulses. Again this might be seen from two antennas on an aircraft. The time shift has now been set to zero on an arbitrary waveform generator, but a 25-degree phase shift is being introduced between the two signals. A capture shown in Figure 5, top center trace C, and related blue marker 1, show this 25-degree phase shift in a mean measurement in lower right Trace D, as well as only a 0.8-degree standard deviation and a 0.7 variance. These are average values over the width of the pulses.
Two-channel phase difference measurement between two RF pulses
More radar/EW/warning receiver applications are driving toward wider modulation bandwidths to increase range and angle-of-arrival precision capability in related systems. At times, this extends beyond 1-GHz modulation bandwidths. Designers increasingly use wideband oscilloscopes as RF receivers to evaluate related wideband signals when validating their hardware prototypes. Although scope measurements directly are of interest, it’s often advantageous to use analysis software to digitally down-convert captured wideband signals to reduce noise and allow more in-depth analysis of baseband I/Q signals. By combining a wideband scope and VSA software with appropriate techniques, you can readily make angle-of-arrival calculations for a variety of systems.