Probe measurement results analysis

We can predict that if we use a probe with a ground lead to measure a signal from a low source terminal resistance source, we will observe artificial ringing and overshoot.

Figures 3.6 and 3.7 allow us to compare our judgment with the actual measurement results. These experiments used a very low capacitance FET type probe, rated for 1.7PF shunt capacitance, with a 3DB bandwidth of 1GHZ, connected to a digital oscilloscope TEKTRONIX11403. The source resistance of the signal source in Figure 3.6 is 25 ohms, and the ground lead is 3IN long. The waveform in the middle uses a bare probe directly in contact with the measurement point, and the ground lead is also 3IN long. The waveform in the middle uses a bare probe directly in contact with the measurement point, and the ground lead is also 3IN long. Obviously, removing the plastic clip of the probe has little effect. These scanned waveforms show that the overshoot is about 15% in the case of 25 ohm source resistance, while the overshoot is as high as 29% in the case of 5 ohm source resistance.

The ringing period shown in the figure is between 2 and 6NS. We can quickly know the time constant of the circuit:

The ringing period obtained from the LC circuit time constant of 0.63NS is:

So far, the measured results are pretty much in agreement with the theory. So what is the waveform at the bottom of the two graphs? Why is it better?

The waveforms at the bottom of the two figures provide us with a good idea to solve the overshoot problem. In the waveform measurement below, we removed the outer plastic shell of the probe and the ground lead, so that the metal shielding layer outside the probe and the probe tip are completely exposed, and then a blade is used to directly connect the outer shielding layer of the probe to the ground of the circuit under test, as close as possible to the signal measurement point. This makes the actual ground wire self-inductance very small. Using this method of directly connecting to the ground wire, the scanning waveforms of the 25 ohm source terminal resistance and the 10 ohm source terminal resistance have been significantly improved in terms of overshoot.

[page]

We can predict that if we use a probe with a ground lead to measure a signal from a low source terminal resistance source, we will observe artificial ringing and overshoot.

Figures 3.6 and 3.7 allow us to compare our judgment with the actual measurement results. These experiments used a very low capacitance FET type probe, rated for 1.7PF shunt capacitance, with a 3DB bandwidth of 1GHZ, connected to a digital oscilloscope TEKTRONIX11403. The source resistance of the signal source in Figure 3.6 is 25 ohms, and the ground lead is 3IN long. The waveform in the middle uses a bare probe directly in contact with the measurement point, and the ground lead is also 3IN long. The waveform in the middle uses a bare probe directly in contact with the measurement point, and the ground lead is also 3IN long. Obviously, removing the plastic clip of the probe has little effect. These scanned waveforms show that the overshoot is about 15% in the case of 25 ohm source resistance, while the overshoot is as high as 29% in the case of 5 ohm source resistance.

The ringing period shown in the figure is between 2 and 6NS. We can quickly know the time constant of the circuit:

The ringing period obtained from the LC circuit time constant of 0.63NS is:

So far, the measured results are pretty much in agreement with the theory. So what is the waveform at the bottom of the two graphs? Why is it better?

The waveforms at the bottom of the two figures provide us with a good idea to solve the overshoot problem. In the waveform measurement below, we removed the outer plastic shell of the probe and the ground lead, so that the metal shielding layer outside the probe and the probe tip are completely exposed, and then a blade is used to directly connect the outer shielding layer of the probe to the ground of the circuit under test, as close as possible to the signal measurement point. This makes the actual ground wire self-inductance very small. Using this method of directly connecting to the ground wire, the scanning waveforms of the 25 ohm source terminal resistance and the 10 ohm source terminal resistance have been significantly improved in terms of overshoot.