PDF1. Introduction
One of the most important features of high-speed digitizers and digital storage oscilloscopes (DSO) is the available onboard acquisition memory. Given the fast sampling rates of today's high-speed instruments, it is important to store the data in the instrument's onboard memory before transferring the data to the PC for further analysis, data storage, and display or distribution. GPIB-based DSOs outpaced the 1-8 MB/s transfer rate of the IEEE 488.2 bus long ago, and today's high-speed digitizers out perform the 132 MB/s transfer rate of the PCI bus, especially in applications that require multiple channel or high-resolution acquisitions. High-speed digitizers with deep memory enable you to efficiently automate your oscilloscope measurements while allowing you to:
- Maintain high sampling rates
- Increase frequency resolution
- Prevent aliasing
- Integrate with other instrumentation
2. Maintain High Sampling Rate
It is important to determine the maximum sampling rate over a given acquisition time period or time-base setting. Maintaining an optimal sampling rate over the required time period often determines the ability to perform comprehensive analysis on the measured signal. The maximum sample rate that a high-speed digitizer or DSO can sustain over a given acquisition time period is dependent on the amount of onboard acquisition memory available on the instrument.
Maximum Sampling Rate = Onboard Acquisition Memory / Acquisition Time Period
Or in terms of time-base:
Maximum Sampling Rate = Onboard Acquisition Memory / (10 * Time-base setting)
*The manufacturer’s instrument specifications also limits the maximum sampling
Measurements that require the capture of "packetized" signals such as video, disk drive read channel, and serial communication signals are just some examples that require high-speed digitizers with deep acquisition memory. "Packetized" signals are often comprised of high frequency components with relatively long time periods. For example, when acquiring serial communications packets, the individual bits will often have short time periods with fast rise time edges, but the entire message or signal envelope may take several milliseconds to complete.
The same is also true when capturing analog signals. For example, to detect high frequency transients in slow varying signals you must acquire the signal with a high sample rate to capture the high frequency content and long enough to capture the slow varying waveform to perform comprehensive analysis.
Capturing a signal with a wide range of frequency components such as 100 Hz, 5 kHz, and 10 MHz requires a digitizer with deep memory to capture the entire 10 ms period of the 100 Hz content and to maximize the sampling rate for the 10 MHz component.
An input signal with 100 Hz, 5 kHz and 10 MHz frequency components requires a minimum 10 ms acquisition time period. With 1 MB of memory, you can maintain a 100 MS/s sampling rate over the entire 10 ms period to capture the details of the high frequency 10 MHz component for further analysis.
An input signal with 100 Hz, 5 kHz, and 10 MHz frequency components requires a minimum 10 ms acquisition time period. With only 250 kB of memory, you must decrease the sampling rate to 25 MS/s to acquire the entire 10 ms period, which causes the loss in signal information.
3. Increase Frequency Resolution
The most common method for performing frequency domain analysis is to compute the
FFT on the sampled waveform. The frequency span of the resulting FFT is between DC and half the sample rate. The sample rate and the number of points used to calculate the FFT determines the frequency resolution, or the smallest frequency step in an FFT spectrum.
With this equation, you can see the frequency resolution improves by decreasing the sampling rate or increasing the number of points acquired. Decreasing the sampling rate is typically undesirable because it decreases the frequency span and increases the risk of aliasing. One of the benefits of deep memory discussed earlier is the ability to maintain high sampling rates over long time periods, which effectively increases the number of points acquired and improves the frequency resolution over a given frequency span.
The graph above demonstrates the importance of optimized frequency resolution when performing frequency analysis on an acquired waveform. Two signals closely spaced in frequency appear as one peak in an FFT analysis with limited frequency resolution. The acquired signal below contains 12.000000 MHz and 12.000180 MHz frequency components. A digitizer with 100 kB of memory and 100 MS/s maximum sampling rate cannot resolve these two frequencies. Even lowering the sampling rate to exactly twice the signal bandwidth would not enable the instrument to resolve the two signals. You can resolve these two peaks with the use of a digitizer with deep onboard memory by acquiring 4 million samples. Also notice the drop in the asynchronous noise floor, which is another benefit of using deep memory to acquire more data for FFT analysis.
4. Prevent Aliasing
The ability to maintain a maximum sampling rate over the acquisition time period helps to prevent effects of aliasing. According to the Nyquist Theorem, the digitizer must sample at a rate greater than twice the highest frequency component of the measured signal to reconstruct the waveform properly. High-speed digitizers with deep memory allow you to maintain high sampling rates over long acquisitions. With deep memory, you do not need to decrease the sampling rate to capture data over the entire period of interest, which helps protect against the possibility of aliasing. The following graph shows the relationship between the acquisition time period with various acquisition memory options and the need to decrease sampling rate as the acquisition time period increases.
The ability to maintain a maximum sampling rate over the acquisition time period helps to prevent effects of aliasing. According to the Nyquist Theorem, the digitizer must sample at a rate greater than twice the highest frequency component of the measured signal to reconstruct the waveform properly. With deep memory, you do not need to decrease the sampling rate to capture data over the entire period of interest, which helps protect against the possibility of aliasing. The following graph shows the relationship between the acquisition time period with various acquisition memory options and the need to decrease sampling rate as the acquisition time period increases.
With 32 MB of acquisition memory, a digitizer with a sampling rate of 100 MS/s can maintain its maximum sampling rate for up to 320 milliseconds. On the other hand, a 100 MS/s digitizer with only 100 kB of memory has to reduce its sampling rate for acquisition periods greater than 1 millisecond. The decrease in sampling rate results in greater risk of signal aliasing, loss in the accuracy for time-domain analysis, and decrease in the frequency span for frequency-domain analysis.
5. Deep Memory Instrument Integration
You can also integrate high-speed digitizers with arbitrary waveform generators to capture waveforms for an extended period of time and regenerate the waveform. Often it is a challenge to create complex test signals for manufacturing test applications. For these applications, you can capture a "gold" standard signal with a high-speed digitizer and output the acquired signal with an arbitrary waveform generator to the device under test. In these applications, it is important to maximize the sampling of the high-speed digitizer to capture the maximum amount of detail in the original "gold" standard signal. Using a digitizer with deep memory enables you to maintain the maximum sample rate over the entire acquisition of the "gold" standard signal. One example would be to capture video signals with the NI 5112 high-speed digitizer and regenerate the same signal on the NI 5411 or NI 5431 to test video components in television, medical imaging, and DVD equipment. In addition, you can synchronize the NI 5112 with the NI 5411 to perform stimulus-response measurements over long periods of time. These applications are possible with the advanced timing and triggering capabilities inherent to National Instruments modular instrumentation architecture. With the NI 5112 and the NI 5411 you can lock to the same reference clock, otherwise known as phase-lock, to ensure precise time base alignment required for these applications. In addition, these devices can share triggers using either the Real Time System Integration (RTSI) bus for PCI modules or the PXI trigger bus for PXI modules.
High-Speed Digitizer and Oscilloscopes with deep onboard memory allow you to capture today's complex signals at high sampling rates ensuring no loss in signal detail. Maintaining high sampling rates provides more accurate measurements in the time domain and better frequency resolution in the frequency domain while helping to prevent aliasing in the acquired waveform. Learn more by reading the Benefits of Deep Memory white paper.
See Also:
Capture of a reference signal on the NI 5112 digitizer and regeneration of the signal on the NI 5411 arbitrary waveform generator
Stimulus/Response with a burst square pulse using the NI PXI-5411 arb and NI PXI-5112 digitizer
Stimulus/Response with a burst sine tone using the NI PXI-5411 arb and NI PXI-5112 digitizer
Swept sine stimulus/response measurements with the NI 5112 digitizer and NI 5411arb: Bode plots
