Oscilloscope probe loading experiment

Oscilloscope Probe Loading Experiments - Hands-on lab experiments and probing tutorials for electrical engineering students

When you connect an oscilloscope probe to an in-circuit test point, the probe itself becomes part of the circuit being measured and can affect the measurement results. This is often referred to as "probe loading." This experiment, using a simple 2-resistor voltage divider network, will show empirically how the frequency-dependent probe impedance can significantly affect measurement accuracy.

Required equipment and components

– 2-channel oscilloscope (≥50MHz bandwidth)
– Function generator (≥10MHz)
– Two standard 10:1 passive oscilloscope probes
– Breadboard
– Two 10kΩ resistors

Compensation probe

Before you build your circuit and experiment with it, it is very important to properly compensate your oscilloscope probes, or your measurements will be inaccurate. To compensate the probes, connect one probe between the oscilloscope's Channel 1 input and the probe compensation test terminals located on the oscilloscope's front panel. Connect a second probe between the oscilloscope's Channel 2 input and the same probe compensation test terminals. Don't forget to connect the ground leads of both probes to the ground terminal on the oscilloscope's front panel. After that, set the probe attenuation factor for both input channels to 10:1 (ten to one). Note that some high-end oscilloscopes will detect if a 10:1 probe is connected and automatically set the probe attenuation factor for you.

Next, set the V/div setting and the sec/div setting for each channel so that one or two cycles of the probe compensation signal appear on the oscilloscope display. The probe compensation signal is typically a 1 kHz square wave, so an appropriate sec/div setting would be 200 μsec/div.

Use a small flat-blade screwdriver to adjust the adjustable compensation capacitor of each probe so that both waveforms are "flat" responding, as shown in Figure 1. This adjustable capacitor is near the tip or part of the probe, close to the oscilloscope BNC input where it plugs into.

Figure 1: Adjusting probe compensation for each passive probe

Figure 2: Compensating a 10:1 passive probe using an oscilloscope's 1kHz probe compensation signal.

Figure 3: Improperly compensated probe.

Figure 2 shows what the normal Channel 1 and Channel 2 waveforms look like if the probe compensation for each probe is adjusted appropriately. Figure 3 shows an example of the Channel 1 probe (yellow waveform) being overcompensated and the Channel 2 probe (green waveform) being undercompensated.

What does probe compensation consist of? We will find out later.

Create experiments, predict outcomes, and measure results

Figure 4: 2-resistor voltage divider network.

Create a 2-resistor voltage divider network using a breadboard and two 10 kΩ resistors as shown in the schematic in Figure 4. Note: If you do not have a breadboard, solder the two resistors together instead of simply connecting them together via long cables and clips. Long cables will add inductance to this experiment, which we want to avoid. Before firing up the function generator and performing any measurements with the oscilloscope, answer the following questions:

Now let's test this circuit and compare the results with our predictions.

Function Generator Setup and Connections:

1. Set the output load impedance to High Z (instead of 50 Ω)
2. Set the waveform shape to Sine Wave
3. Set the amplitude to 5Vpp
4. Set the offset to 0.0V
5. Set the frequency to 10kHz
6. Connect the output of the generator to R1.
7. Connect the generator ground to the circuit ground.

Oscilloscope setup and connections:

1. Connect the Channel 1 probe between Vin and ground.

2. Connect the Channel 2 probe between Vout and ground.

9. Measure Vin and Vout (peak-to-peak) using hand-placed cursors or automatically performing the measurement, or simply count the number of divisions and multiply by the vertical scale factor (1.0V/div).

4. Ensure that the probe attenuation factor is still set to 10:1 for both channels of the oscilloscope.

.Set the vertical scale of Channel 1 and Channel 2 to 1.0V/div.

6. Use the vertical position/offset controls to center the Channel 1 and Channel 2 waveforms on the screen.

7. Set the horizontal scale (time base) to 20.0 μs/div.

8. Set the rising edge trigger on channel 1 to approximately 0.0Volts (typical default setting).

9. Measure Vin and Vout (peak-to-peak) using hand-placed cursors or automatically performing the measurement, or simply counting the number of divisions and multiplying by the vertical scale factor (1.0V/div).

The oscilloscope will now display something similar to Figure 5.

Figure 5: Using two channels of an oscilloscope to measure Vin and Vout at 10 kHz.

Record the measurement results:

Vin @10kHz= ____________

Vout @10kHz = ___________

Vout/Vin @10kHz= _______

Is it very close to what you originally predicted? _____________

Now, change the frequency setting of the function generator to 10 MHz. Also change the horizontal time base setting of the oscilloscope to 20.0 ns/div so that you can see this faster input signal. Measure Vin and Vout again. The oscilloscope should now display something like Figure 6.

Figure 6: Using two channels of an oscilloscope to measure Vin and Vout at 10MHz.

Record the measurement results:

Vin @10kHz= ____________

Vout @10kHz = ___________

Vout/Vin @10kHz= _______

Is it close to the predicted result? __________________

If no, what is the reason? ________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________

Understanding Probe Loading

At 10 MHz, the signal amplitude is reduced as it passes through R2 due to capacitive probe and oscilloscope loading. Ideally, the probe would have infinite impedance and would not affect the measurement. However, whether you are using a spectrum analyzer, power meter, DMM, network analyzer, or oscilloscope, as soon as you connect a probe to the device under test, the probe and the instrument become part of the circuit being measured and affect the accuracy of the measurement. This is especially true when testing high-frequency signals.

Now, let's take a closer look at the probe we just used in this experiment - near the BNC connector at the oscilloscope input. You can see the manufacturer's name and model number associated with this probe, as well as the input impedance specifications/characteristics. As shown in Figure 7, it may say "10MΩ/15pF", etc.

Figure 7: Oscilloscope probe models and input impedance characteristics.

This means that when you connect the probe to the oscilloscope, the equivalent input impedance of the probe is 10MΩ, 15pF. Figure 8 shows the equivalent probe/oscilloscope loading model. It is in parallel with R2 (see Figure 4). You can assume that the 10MΩ resistor is so large compared to the 10kΩ resistor (R2) that R2 can be ignored. You can also assume that at low frequencies, the 15pF capacitor will not affect the circuit. But at 10MHz, what is the reactance of this capacitor?

Xc = 1/(2πfC) = ____________

Figure 8: 10:1 passive probe loading model.

Now, calculate the load impedance including R2 (in parallel with Xc). Remember, you can ignore the 10MΩ resistor.

Determine the approximate output voltage with the input frequency set to 10 MHz—The voltage divider-based network now includes R1 in series with ZLoad.

Vout = ______________

Vout/Vin = ___________

When the input signal is set to 10 MHz, are the calculated results close to the measured results? _________________

So, we now seem to be stuck between a rock and a hard place. We need to measure the output voltage of the circuit, but as soon as we connect the oscilloscope probe to the circuit, it changes the output characteristics. How do we solve this problem?

First, let's use the 10 kΩ resistor in this experiment to illustrate a point. That is, at high frequencies, the capacitive reactance of the probe can "swamp" the impedance of the load resistor (R2). But in reality, most high-frequency designs include low-impedance devices/components. Even in low-impedance designs, the probe will still affect the circuit under test when the frequency reaches high enough (such as hundreds of megahertz or gigahertz signals). Moreover, most PCs today operate in the multi-gigahertz range.

Such applications usually require specialized high-frequency "active" probes. Passive probes, such as those used in this lab, consist only of "passive" components, resistors and capacitors. High-frequency probes usually include "active" components, such as transistors and amplifiers, and these probes require a power source to operate. The input capacitance of active probes is in the sub-microfarad range. This means that at high frequencies their effect on the circuit will be small, but in theory not zero. However, the price of these probes is also far beyond the standard passive probes that are usually provided with the oscilloscope. Active probes are almost always an "expensive" option.

If you need more information about oscilloscope probes, download the Keysight application note "8 Tips for Better Oscilloscope Probing" listed at the end of this article.

Understanding Probe Loading

Figure 9 shows a more detailed (but still simplified) electrical model of a 10:1 passive probe when connected to an oscilloscope using the oscilloscope’s default 1 MΩ input option. Although the electrical model of a passive probe and oscilloscope includes inherent/parasitic capacitance (not included in the design) as well as an intentional compensation capacitor network, we will ignore these capacitive components for now and focus on analyzing the ideal signal characteristics of this probe and oscilloscope system at low frequencies.

After removing all capacitive components from the electrical model, only a 9 MΩ probe resistor remains in series with the 1 MΩ input impedance of the oscilloscope. The net input resistance of the probe is 10 MΩ, consistent with the probe loading model shown previously (Figure 8). Using Ohm's law, you can find that the voltage received at the oscilloscope's BNC input is 1/10 of the voltage at the probe: