Pulsed Measurement Capability of CMT VNAsApril 20, 2018
Pulsed S-parameter measurements are important when testing a DUT at a higher power than it can handle without damage in the steady state, or when the normal operating mode of the DUT involves RF pulses. Examples include amplifier chips or circuits without their thermal packages and radar components or systems.
This application note demonstrates how the minimum measurement window and trigger timing behavior of CMT VNAs can be experimentally verified. It was developed based on the 8 GHz Planar 804/1 VNA using v3.49 software. To ensure the measurement settings and pulse timing match the descriptions in this application note, please install v3.49 software or later.
Copper Mountain Technologies VNAs are capable of measuring externally generated or modulated pulsed signals. When the pulse signal is generated externally from the VNA, the system is setup as shown below. The 10 MHz reference signal will synchronize the two instruments to eliminate any frequency offset between the pulsed RF signal generators and the VNA receivers. The trigger output from the function generator is connected to the Planar 804/1 trigger input to initiate the VNA’s measurement synchronous with the pulse occurring on the function generator output.
Connected as such, the system is not much useful for measuring S-parameters of a DUT, but rather is configured for experimental measurements of the trigger timing and measurement window behavior.
Figure 1: Block diagram of the test setup.
The signal to be measured is produced by the function generator, and is specified by its pulse duration and pulse rate:
Figure 2: Pulse duration and interval/rate.
Figure 3: Photograph of the test setup.
Since the signal is generated by the function generator instead of the VNA itself, several instrument configuration steps are required in order to obtain a proper measurement:
· Turn off the VNA’s RF output (Stimulus > Power > RF Out > Off)
· Change the measurement type from S-parameter to absolute (Measurement > Absolute > Receiver A (Source port 1) or Receiver B (Source port 2))
· Change the Trigger settings
o Stimulus > Trigger > Continuous
o Stimulus > Trigger > Ext Trigger > Position > Before Sampling
· Change the reference source
o System > Misc Setup > Ref source > External
The behavior of each of these settings, as well as alternative configuration options, are detailed in the VNA’s operating manual.
IF bandwidth setting
Setting the IF bandwidth (IFBW) properly is essential for obtaining reliable measurements. The IFBW setting will determine the length of the VNA’s measurement window. The relationship is approximately ??????????? = 1.2 ???? . e.g 1 kHz IFBW > 1.2 msec; 10 kHz IFBW > 120 usec; 30 kHz IFBW > 40 usec; etc. And some overhead time that relates to the trigger jitter time and uncertainty of measurement time needs to be added as well, which is about 10 usec.
The length of measurement window must be shorter than the pulsed signal itself in order to obtain a stable measurement result. Otherwise, the VNA’s measurement will include the edge of the pulsed signal and some portion of time with no signal present, which contributes only noise to the measurement result.
External Trigger Delay
External trigger delay allows for introducing a delay between the occurrence of trigger event and initiation of the actual measurement. It can be useful when trying to avoid the rising edge, or when measuring a specific portion of a pulsed signal.
The timing is as shown below:
Figure 4: Timing diagram of signals in the system.
External Trigger Delay
Measurement delay is applied between Setup and Sampling. Measurement delay can be setup by Stimulus > Meas Delay. Measurement delay is useful when extra time is need for the DUT to be ready, e.g. when measuring a DUT with a long internal time delay or slow time constant.
Figure 5: Measurement sequence for each point in the sweep.
Note that measurement delay will only shift the measurement window relative to the trigger event if the trigger is set to “Before Setup” (Stimulus > Trigger > Ext Trigger > Position > Before Setup)
In this case, the analyzer will start Setup (retuning and levelling of its generator output) after receiving the trigger signal. The total amount of time (Setup+Delay+Measurement) will vary a little bit, since the time needed for setup varies.
Figure 6: Oscilloscope screen capture of the pulsed 30 MHz signal, 300 usec duration; blue trace, along with the trigger signal (Red trace).
The following example shows measurements of a pulsed 30 MHz sinusoidal signal where the duration of each pulse is 130 usec. Each setting addressed here can be customized for any particular measurement.
In this measurement, we only want to measure the power at 30 MHz, so the stimulus setting is set to Start frequency = Stop frequency = 30 MHz.
If we want to measure 201 points, just set “Points” to 201. Under this configuration, 201 external trigger events will complete a full cycle of measurements (a complete sweep in the plot window).
Suppose we want to measure power at 3 particular frequencies, for example 30 MHz, 70 MHz and 100 MHz. We can setup a segment sweep for those three frequencies (Stimulus > Sweep Type > Segment) and configure the segment table (Stimulus > Segment Table).
For each trigger signal received, we want to measure one point. Therefore, set the trigger event to on point (Stimulus > Trigger > Ext Trigger > Event > On point).
Since the pulse duration is 130 usec, we want to set the measurement window to be shorter than that. So we set the IFBW to be 30 kHz, which corresponds to approximately 40 usec. And to make sure it’s not measuring the rising edge of the pulsed signal, the external trigger delay is set to 50 usec. Now we are ready to begin measurements. The measurement timing diagram can be shown as follows:
Figure 7: Timing diagram with 50 usec trigger delay. The pulse and trigger signals begins simultaneously; Trigger Delay shifts the measurement window to a later portion of the pulse.
VNA measurement results at 30 MHz, 70 MHz and 100 MHz (3 points in total) are shown below. It shows that the first point (30 MHz) measures about +0.2 dBm signal strength and the following two points (70 MHz and 100MHz) do not measure much power. This is what we would expect, since the generated signal only contains energy at 30 MHz.
Figure 8: Measurement of 30 kHz pulse with external trigger on point.
The following screenshot shows the same VNA measurement when the external delay is set to 400 usec. When the external trigger delay is too long, the pulsed signal will be missed by the measurement window, so we will see little or no signal energy measured at 30 MHz.
Figure 9: Same test configuration but with trigger delay too large to catch the pulse.
Here is the approximate shortest pulse duration measurable for several CMT VNA models, corresponding primarily to each instrument’s maximum IF Bandwidth:
Copper Mountain Technologies VNAs are capable of measuring externally-generated or externally modulated pulsed signals, provided the pulse duration is at least as long as the measurement window plus timing jitter/overheads. By adjusting the external trigger delay setting, the position of the measurement window with respect to an externally-generated trigger signal can be optimized. Planar series VNAs from CMT are able to measure pulses as short as 50 usec in duration, while Cobalt series instruments can measure pulses 20 usec in duration or shorter.