A “probe” can be thought of as any device used to transmit a voltage signal from a DUT (device under test) to an oscilloscope; this includes 50 ohm cables, active probes, passive probes, differential probes, wire leads and ground extensions, etc. Probes are used to deliver signals to an oscilloscope from the device and are inherently lossy, plus their characteristics vary from probe to probe. The characteristics can depend on frequency, temperature, manufacturing variations, connection methods, and of course damage. For the purpose of this paper, the probes considered are active and passive probes.
As oscilloscopes continue to achieve higher bandwidths, the cables used are becoming the bandwidth bottleneck of the systems. To keep up with the increasing bandwidth needs of their customers, oscilloscope vendors now invest in high end semiconductor chips on probes as well as their oscilloscopes. However, even with high bandwidth technology, probes remain lossy and probe vendors use digital signal processing techniques to ensure a flat frequency response and less loss. The evolution of various methods and techniques of probe correction have followed scopes’ path toward feature-richness and sophistication. The advantage of digital signal processing is that it provides more flexibility over hardware implementations, and with modern microprocessors in oscilloscopes, present very little impact on update rate performance.
The terms “calibration” and “correction” can be used interchangeably, however normal nomenclature is that calibration tends to represent the process of setting up a correction, which is then applied to a signal once calibration is completed. To understand and characterize scope probes, it’s important to understand that any circuit with a probe attached effectively becomes a new circuit that includes the probe and probe accessories. VSource represents a circuit signal with no probe connected, VIn represents a new signal with the probe’s effect included (the voltage at the probe’s tip), and VOut represents the signal as passed through the probe. An understanding of these signals is essential to any discussion of probe correction.
DC Correction Methods
The most common correction method for probes is the DC adjustment, which entails the adjustment of probe gain and probe offset. Probe gain correction simply adjusts the scope’s scaling factors of the signal displayed on screen to properly match the correct DC values. In other words, the instrument forces VSource and VOut to match (at DC only) by scaling the vertical axis (voltage) accordingly. In Keysight Technologies, Inc. oscilloscopes the DC voltage source (“probe comp” or “cal out”) has a source impedance of 0 ohms, so VSource = VIn (probe loading does not affect this circuit.
It’s important to note that this scaling can be greater or less than one (gain or attenuation); for example, a 10:1 passive probe must be scaled up by 10, and many active probes must be scaled down. Typically a given probe will not have exactly nominal gain characteristics due to manufacturing variability (a 10:1 passive probe might actually be 9.85:1, for instance). Probe offset variation stems from similar manufacturing differences and simply represents the output of the probe when it measures a signal of 0 V. The instrument subtracts the offset from future measurements after probe calibration to ensure this case is true.
These DC corrections represent a very simple straight line, y = mx + b situation; to determine gain and offset coefficients the user connects a given probe to the instrument, with its tip, or input, connected to an instrument output (typically known as “probe comp”). Once the calibration is initiated, the instrument outputs known DC voltages and compares these values to the scope input after the signals pass through the probe. While only two points are technically required to determine the gain and offset coefficients, most instruments collect more than that. These coefficients don’t typically drift much with time; performing probe DC calibrations several times a year is enough.
AC Correction Methods
As oscilloscope performance increases to multi-GHz levels, the use of DC correction methods becomes less reliable, since many probe characteristics are strong functions of frequency, particularly at higher bandwidths. The term “AC correction” refers to correction schemes that vary with frequency and that attempt to adjust a probe’s characteristics to be in line with those of an “ideal probe.” An ideal probe is one that has a flat frequency response up to its bandwidth (-3dB point, or the point at which the signal level is attenuated by the probe to 71% of the original signal), and that minimally loads the circuit to which it is connected. The loading of a probe is a complex impedance, and ideally would be infinite, since in that case it would have no effect on the DUT circuit whatsoever. Unfortunately probe manufacturers are limited by the realities of physics, and a number of other real-world constraints which lead to deviations from the ideal.
AC correction methods require understanding some key terms:
VSrc - The signal at the probe point before the probe is connected which would be the signal at the probe point if an ideal probe with infinite input impedance were connected
VIn - The signal at the probe point while the signal is being loaded by the probe. Probe loading is caused by the input impedance of the probe making a voltage divider with the source impedance of the circuit being measured.
VOut - The signal that is output from the probe
Vout/Vin Correction - The signal at the output of the probe is an accurate representation of the signal that currently exists, as it is being probed
Vout/VSrc Correction – The signal at the output of the probe depicts the signal before it was probed or VSrc
귀하의 양식이 성공적으로 제출되었습니다.