Probes, probes, probes,...

For time-varying signals, the oscilloscope is the most important measuring instrument. In addition to the basic capability of reading time and voltage parameters—which, although with greater difficulty, was also possible with analog instruments—digital oscilloscopes perform more advanced data processing as well.
To properly take advantage of an oscilloscope’s capabilities when measuring signals, it is necessary to avoid basic errors caused by incorrect connection of the oscilloscope to the circuit under test. A fatal mistake is connecting measuring equipment with characteristics that completely alter the state of the measured component or circuit. The worst-case scenario—which unfortunately does occur—is that the device under test temporarily or permanently stops functioning.
Oscilloscope features
An oscilloscope is not the best voltmeter. The converter usually has only 8-bit resolution, which in itself represents an error greater than 0.8%. The number of effective bits in today’s oscilloscope A/D converters does not exceed 7, i.e., 1/128. The overall accuracy on the vertical axis is therefore specified in units of percent. In addition, the specified bandwidth must be taken into account. The definition assumes a maximum drop of 3 dB, which in the voltage domain represents almost a 30% error when measuring amplitude. The vertical system has the following frequency characteristic:

Fig. 1 Amplitude attenuation versus frequency of a 100 MHz oscilloscope (idealized system)

where AU is the displayed amplitude
A is the signal amplitude
fBW is the bandwidth
From this relationship it follows that an error of several percent is introduced into the result already at one-fifth of the oscilloscope’s frequency range. An additional error can be introduced into voltage measurements by using AC coupling at low frequencies.
On the horizontal (time) axis, the oscilloscope provides much more accurate information than on the voltage axis. Measurements of period, frequency, and rising or falling edges are easily handled by digital oscilloscopes. It is also possible to measure phase shift between channels or to use cursor-based measurements. The ubiquitous bandwidth parameter also plays a role here, especially in edge measurements. Because oscilloscope inputs, due to their impedance, have their own time constant, the oscilloscope’s intrinsic rise time is also included in the measurement result.
The resulting rise time (or fall time) is expressed by the following relationship:
where ts is the actual rise time
fBW is the bandwidth
The resulting distortion can be considered the maximum possible error, because the bandwidth is usually specified with a margin that reduces the error. In percentage terms, the magnitude of the error is similar to that of voltage measurements. If a probe is included in the measuring setup, the following term must be included in the previous relationship:
.
Here we have shown how large the errors introduced into the measurement are by the input circuit itself. At the moment when the oscilloscope is connected to the circuit via a cable or a probe, additional distortion is introduced into the measurement chain.
Connecting an oscilloscope using a cable
If we do not work with a matched transmission line—where the source, oscilloscope, and cables are matched to 50 or 75 Ω, or another value depending on the application—but instead use, for example, a common banana/BNC cable for the connection, the error caused by such a connection can even be fatal.
The explanation of the first error component is relatively simple. If, for example, we load a 10 V DC source with an output resistance of 200 Ω using a cable and the oscilloscope input impedance of approximately 120 pF and 100 kΩ, the oscilloscope will display 9.98 V due to the voltage drop across the internal resistance of the source. The second error component becomes apparent at higher frequencies as a limitation of the system bandwidth.
Bo = 1 / 2πRC, i.e., approximately 6.6 MHz!
This naturally affects both the voltage and time axes when measuring signals with frequencies from about 1 MHz upward, as described above. Therefore, oscilloscope probes are used for such measurements.
Connecting an oscilloscope using a passive probe
Passive probes are the most commonly used probes and are very often included as part of the standard accessories of a new oscilloscope. Passive probes should have an input resistance at least 10× higher than the output resistance of the circuit under test. Therefore, for conventional oscilloscopes with high-impedance inputs, 10:1 probes with 10 MΩ and 10 pF are used. Such probes are generally suitable for frequency ranges up to 500 MHz; an example is the Tektronix TPP1000 probe, 10:1, 1 GHz, with an input capacitance of only 4 pF and 10 MΩ.
1:1 probes must have lower input resistance and higher capacitance, and are therefore suitable for frequencies up to 20 MHz. 1:1 probes are used only exceptionally, and only when the sensitivity of the oscilloscope’s vertical circuits is insufficient even when using a 10:1 probe.
Probes with ratios of 100:1 up to approximately 5 kV, or 1000:1 up to 20 kV RMS, or up to 60 kV for short-term voltage values are also manufactured. In the case of high-voltage probes, designs and materials corresponding to the magnitude of the measured voltage are required. In addition, such a probe must be designed to dissipate the heat generated at the probe’s input resistance (approximately 4 W).
For high-frequency applications in 50-ohm circuits, passive probes with low input resistance (hundreds to thousands of ohms) and an input capacitance of 1 pF are used. Such probes can be used up to approximately 10 GHz. They are not suitable for low frequencies and operate only at voltages on the order of tens of volts. However, they can be used with spectrum analyzers. Due to the very low input resistance, these probes must be handled with great care and used only where the probe does not excessively load the circuit under test.

Fig. 2 High-voltage probe Tektronix P6015A, 75 MHz, 40 kV pulse, 20 kV RMS

Fig. 3 Low-input-resistance probe 3 GHz, 50 V
Because oscilloscope input circuits are manufactured in specific impedance ranges, it is advisable to use probes supplied with the given instrument. The ideal condition occurs when the product of the probe resistance and capacitance (including the cable) equals the product of the oscilloscope input resistance and capacitance:
.
For this reason, probes provide a compensation adjustment. Basic adjustment can be performed using the oscilloscope’s calibration output or another source of a high-quality square-wave signal. If the displayed rising edge is not terminated with a right angle, the probe must be adjusted using the capacitive trimmer. The probe is then matched to the given input. If the probe is connected to another channel of the instrument, the compensation must be repeated, as the input impedances differ.

Fig. 4 Examples of possible probe adjustment variants
In addition, any probe should be immune to external interference. A simple test can be performed by connecting both the probe tip and the ground lead to the circuit ground—the displayed trace should be a straight line. If the signal is noisy, it indicates that a different probe position or a different connection method should be used. In general, the longer the non-coaxial section, the lower the probe’s immunity to interference. The non-coaxial section behaves as an inductance that acts as an antenna and also affects the impulse response of the entire system—see Fig. 5.
Special connectors can then be used for connection; see Fig. 3 and Fig. 6.
Probes are very often supplied with a variety of small cables and clips for connection; this should also be taken into account, as it can easily happen that one “runs out of hands.”

Fig. 5 Demonstration of the effect of probe connection on pulse edge display

Fig. 6 Possible probe connection using an adapter
Differential probes
If it is necessary to measure differential voltage at two different points, two probes can be used, with their ground leads connected together, and a mathematical channel subtraction performed on the oscilloscope. Similarly, a purely passive pair can be assembled, but because this approach has its limitations, active differential probes containing a differential amplifier are commonly used.
For this reason, a differential probe—like active or current probes (which will be discussed later)—requires power from an external source. Modern oscilloscopes already feature input connectors designed with surrounding power and identification contacts, allowing probes to be powered directly from the oscilloscope.
For differential probes, the parameter CMRR (Common-Mode Rejection Ratio) is also specified. Probes are manufactured either for high-frequency applications up to tens of GHz at voltage levels typical of standard logic signals, or for floating measurements up to hundreds of volts (and hundreds of MHz). In general, the wider the bandwidth of a probe, the lower the voltage that can be measured with it.
Since oscilloscopes are usually grounded and floating measurements are very common, differential probes are essential for on-board applications. At low frequencies, this includes measurements on switched-mode power supplies; at high frequencies, serial data interfaces such as USB 3.0, HDMI, PCIe, GBE, and others dominate.
A special category is the so-called Tri-Mode probes from Tektronix. These probes allow, using a single probe and without changing the connection, measurement of differential voltage, signal-to-ground voltage, and common-mode voltage with respect to ground.
Probes intended for high-speed signals are very sensitive to static charge and electromagnetic coupling; careless handling can easily damage their input circuits.
Active probes
If the circuit under test is excessively loaded by a passive probe with low resistance, or if the frequency range of high-resistance passive probes is insufficient, an active probe can be used. Active probes are designed using FET transistors, with a division ratio typically of 5:1 or 10:1. The probes are connected to oscilloscopes with a 50 Ω input impedance and very often include memory in which correction and calibration data are stored and can be transferred to the connected oscilloscope.

Fig. 7 Diagram of the active probe P7240
The frequency bandwidth ranges from 500 MHz to 7 GHz. For frequencies below 500 MHz, it is more advantageous to use a passive probe, as active probes do not achieve comparable voltage ratings (maximum 40 V, but typically only a few volts). Despite their cost, active and differential probes have become indispensable for measurements on today’s very fast digital circuits; moreover, due to the ever-decreasing size of electronic components, physically connecting a probe to the circuit is becoming increasingly difficult.

Fig. 8 Examples of probe connections to miniature components

Fig. 9 Miniaturization does not simplify probe connections
Current probes
Let us recall that in the list of SI units, the quantity we have been measuring so far with various probes is not voltage, but electric current—the ampere. Oscilloscope current probes come in two basic designs. Either they are current transformers that operate from a non-zero frequency, or a combination of a transformer with a Hall sensor. Current probes virtually do not load the circuit under test; however, the closer the measurement point is to the grounded end, the better. RF current probes are grounded in order to prevent interference of the probe secondary caused by induction from external voltages.
A typical example of a passive current transformer is the Tektronix CT1 and CT2 probes. These miniature probes with a closed magnetic core operate up to 1 GHz. The disadvantage is precisely the closed core—the conductor on which the current is to be measured must be disconnected, threaded through the probe, and then reconnected. The clamp-type design eliminates this inconvenience; unfortunately, the air gap in the core adversely affects the quality of the results, and ensuring the stability of this gap is technically challenging.
Probes intended to measure DC currents also have a Hall sensor inserted into the magnetic circuit.

Fig. 10 AC current probes and connection diagram
Hall effect – a phenomenon in which a voltage (Hall voltage) arises between opposite edges of a semiconductor plate through which current flows longitudinally while a magnetic field is applied perpendicular to it.
At low frequencies, the AC and DC components are summed; at higher frequencies, only the transformer section operates. Because probes with a Hall sensor require power, the probe either contains a battery or the Hall sensor is powered from an external source or directly from the oscilloscope. Such probes operate up to approximately 120 MHz, while the simplest versions operate up to 100 kHz.
Given the operating principle, it is very important to note that the maximum current a probe can measure is strongly dependent on the signal frequency. The higher the frequency, the lower the measurable current.
Oscilloscope probes serve to physically connect the circuit under test to the oscilloscope. Hopefully, this brief overview will help with better orientation in selecting the appropriate probe for a given measurement so that the obtained data correspond as closely as possible to reality.









