We've routinely posted on how you can characterize your total measurement system to gain important "situational awareness" when using an oscilloscope to make sensitive measurements. The knowledge gained from these tests helps you properly interpret your measurement results so that you can deduce what is actually going on with your circuit, versus what is an artifact of the measurement system. Listed here are nine important things you should know before making sensitive measurements with your oscilloscope, with links to blog posts that instruct you how to test them.
1. Oscilloscope Bandwidth
Most oscilloscope users are familiar with the Nyquist principle and that the frequency of the signals you need to measure determines the oscilloscope bandwidth required. But to help distinguish measurement system artifacts from actual signal components, it's important to be able to ascertain your oscilloscope's effective bandwidth when operating at or near it's specified limits, and to understand how to use memory properly to optimize the oscilloscope's effective digital bandwidth.
2. Channel Bandwidth
Modifying the input channel bandwidth through the use of filters will also change the shape of the signals you see.
3. Probe Bandwidth
You most certainly know probe bandwidth must be able to accommodate the speed of the signals you need to measure, but be aware of how ground leads and tip inductance impacts probing system bandwidth.
The type of interconnect used also can impact the shape of the signals seen.
4. Vertical Scale and Signal Range
The type of probe (single-ended or differential) and the type of signal (floating or ground-referenced, high-frequency or high-bandwidth) will determine the signal range and input offset range that can be safely and accurately measured and displayed.
The oscilloscope also has its own limits in this regard, and you can take measures to improve the oscilloscope's dynamic range.
Differential measurements require good Common-mode Rejection Ratio (CMRR), and the higher the signal frequency, the better that CMRR needs to be. As with oscilloscope effective bandwidth, you should test your probe's effective CMRR.
5. Oscilloscope Sample Rate
How fast your oscilloscope can sample, or its time resolution, will determine what speed of signals you can hope measure with reasonable accuracy. The sample rate you absolutely need will partly depend on the frequency and rise time of the signals you are measuring, but also on the degree of signal fidelity you need (are you measuring jitter, or decoding serial data?).
If you are measuring in the frequency domain, you want to watch your oscilloscope sample rate, because the highest frequency you can measure is one-half the sample rate.
6. Timebase and Acquisition Buffer Size
Also when measuring in the frequency domain, you will want to watch your oscilloscope timebase (Time/div) and acquisition buffer size (total number of samples), because they will determine your total acquisition time. The lowest possible frequency you can measure (the fundamental) is equal to 1/acquisition time. For example, if your acquisition time is 20 µs, 1/20 µs equals 50 kHz. Every other harmonic will be a multiple of this fundamental frequency.
Another buffer is the total number of acquisitions and measurements that can be held in oscilloscope history before older acquisitions "drop out" of the buffer. Of course, all persistence displays (like eye diagrams) and statistical measurements are impacted by this second buffer. Histograms, in particular, are directly affected by the number of measurements plotted, which can greatly affect their shape and, in turn, any measurements made of the histogram. Be sure you are collecting a large enough sample size when making statistical measurements.
7. Input Impedance of the Channel
Different oscilloscope input impedances relative to the impedance of the transmission line used to input the signal will have different effects on the signal you see. You should understand how to choose between the oscilloscope's 50 Ω input and 1 MΩ input and be aware of where your oscilloscope is currently set.
8. Source Output Impedance and Resistance
You need to be aware of changes in impedance throughout the measurement system that could cause reflections. These reflections may appear to be ringing in the signal, but are actually artifacts of impedance mismatches in the measurement system. Besides knowing the characteristic impedance of the transmission line and the input impedance of the oscilloscope channel, you want to know the source Thevenin output impedance and resistance.
9. Environmental Noise
With their many power supplies and devices operating at different frequencies, laboratories and workshops can be very noisy places, and some RF pickup noise can masquerade as signal noise. Knowing your environmental "background" noise will help identify those bands that commonly insert themselves into your measurements.
No comments:
Post a Comment