Measurement Effects
Figure 16
Figure 16 shows the results of the measurement system on the signal that is being measured. The rise time of the signal that is being measured is shown as tdut (or the generator). The system risetime is affected by the RC loading of the probe, tRC, the rise time of the probe, tprobe, and the rise time of the oscilloscope, tscope. If the system is assumed to be Gaussian in nature, the observed rise time, tobserved, will be the square root of the sum of the squares of tdut, tRC, tprobe, and tscope.
Now compare the effect of using an HP 10441A passive probe with that of an HP 1145A active probe. Look at the effects of these probes on measuring a 1-ns rise time. The bottom of figure 16 gives the capacitive and resistive values for the probe and oscilloscope along with the rise times. The observed rise time using the HP 54542A with the HP 10441A is approximately 1.57 ns. Using the HP 54542A with the HP 1145A active probe, the observed rise time is approximately 1.33 ns. There is about a 20% improvement when using the active probe vs. the passive probe. Note that even using the active probe the rise time being measured is actually 33% faster than the rise time that is seen on the oscilloscope. In order to make accurate rise time measurements, the key is to have significantly more measurement bandwidth than the signal that is being looked at.
The waveform that is seen on the screen is a result of a number of factors that have been discussed earlier in the probe loading section regarding the probe and the measurement system. The R-C loading effects of the probe can affect the circuit under test, while the probe/oscilloscope can affect the overall bandwidth of the waveform that is being measured.
Figure 17
Figure 17 shows the impedance vs. frequency of various probes that have been discussed. The loading impedance of the HP 10441A starts to increase at approximately 20-kHz. It is the most versatile of the four probes shown here, thus making it the most general-purpose probe for verification. The HP 10441A is good at measuring both voltage and time where precision is not required. If you need a certain parameter to be measured more accurately, there are other probes to choose from. These other probes are more appropriate for characterization, but there will be tradeoffs. For the HP 54006A (the resistive divider probe), its corner is out well past 1-GHz, but it only has a 500-
impedance and can (in some cases) significantly load the circuit. The probe that has the lowest overall loading effects above 3 MHz is the HP 54701A probe.
Figure 18
Figure 18 shows the measurement error as a function of the ratio of the system rise time to the measurement rise time. If the system-to-measurement rise time ratio is 1:1 (system rise time is equal to measurement rise time), then there is a 42% error in the observed rise time. If the system-to-measurement rise time ratio is 4:1 (the system is 4 times faster than the measurement), then the error in measuring that rise time is about 3%. The 4 to 1 ratio is used as a boundary between making functional measurements (or verification) versus characterization measurements. For a 7:1 ratio, in which the system is 7 times faster than the measurement, the result has an approximate 1% error in the rise time measurement. We recommend that the scope and probe be chosen to match the measurement accuracy needed. If troubleshooting or making functional measurements is the desired application, then it is possible to use something less than the 4:1 ratio between the system performance and the signal being measured. On the other hand, if characterization is desired, then the system performance must be at least four times faster than the circuit being measured.
Page 4 of 7