Before taking a measurement, choose the correct type of oscilloscope probe for the measurement being taken. The true bandwidth of the oscilloscope and the probes must be at least the bandwidth of the signal to be measured to avoid distortions in the measurement. Since most oscilloscopes and probes are not specified as true bandwidth but as -3dB bandwidth, using an oscilloscope and probes with at least 10 times the bandwidth of the signal to be measured makes sure that the signal is not distorted. For example, a signal that rises linearly in 10ns has a bandwidth of 0.35/10ns = 35MHz (Bogatin), and so setting the oscilloscope to 350MHz or greater and using probes with bandwidth 350MHz or greater is sufficient.

When choosing oscilloscope probes, remember that probes are also susceptible to parasitics. Long probe cabling can give an inflated drop is transient tests. Capacitive 10X probes can add additional noise to the LDO's output signal. Additionally, introducing long ground loops while using a 10X probe can add unwanted noise to a measurement. 10X probes also effectively increase the signal to noise ratio because the probes divide the signal down by a factor of 10 and then the oscilloscope digitally gains the signal back up by a factor of 10. For this reason, when measuring small signals, sometimes using a 1X probe is appropriate. When using a 1X probe, remember that the voltage going into the oscilloscope channel is not divided down, so care must be taken to avoid exceeding the voltage rating of the oscilloscope channel.

In this section, two oscilloscope probes, a SMA probe and a probe with a ground clip are evaluated, while taking a load transient measurement.

Figure 3-1 shows a load transient measurement using a probe with a standard Alligator ground clip. Figure 3-2 shows the setup for a standard Alligator ground clip, where there are not convenient test points on which the probe tip and ground clip can be connected. The available banana connectors are used to run out wires to the probe. Figure 3-1 shows that while using a standard Alligator clip, the measurement looks visually very noisy, and users can conclude that the device appears to be much more noisy than the data sheet specifies.

Figure 3-3 shows a load transient while using a ground spring attachment for the oscilloscope probe. Figure 3-4 shows how to use the ground spring connector that comes with every oscilloscope probe. If there are no SMA connectors on the board, using this ground spring connector is the best way to get the shortest ground return path using a standard oscilloscope probe, and results in the least amount of noise coupling into the measurement. Figure 3-3 shows a much less noisy measurement using the ground spring connector.

Figure 3-5 shows a load transient with a SMA connector, and Figure 3-6 shows the setup for using an SMA connector. Figure 3-5 shows the least noisy measurement. SMA connectors are shielded which reduces noise coupled onto the cable, and the ground return loop is also minimized. The result is a very low-noise measurement. Whenever possible, TI recommends using SMA connectors for the lowest noise measurement capabilities. Many of TI's modern LDO EVMs employ SMA connectors for measuring critical nodes, such as the input and output voltages near the device pins.