This video documents an attempt to measure and compare the rise times of various oscilloscope probes. The creator's initial effort with a 74AC logic-based pulser proved too slow, leading him to construct a faster, sub-nanosecond avalanche pulse generator based on a classic Jim Williams (App Note 47) design using a 2N3904 transistor.
Using a low-inductance test fixture, he measures several probes from manufacturers like Rigol, HP, and Tektronix, as well as an unbranded generic probe. The results are inconclusive and counter-intuitive, with measured rise times clustering between 1.2 ns and 1.5 ns, regardless of the probes' specified bandwidths. A cheap generic probe and a 100 MHz probe performed unexpectedly well. The creator concludes that the measurements are fundamentally limited by the 350 MHz bandwidth of his oscilloscope, which itself has a rise time of approximately 1 ns. The experiment highlights that the entire measurement system (pulser, probe, and scope) contributes to the final reading, and a faster signal source and oscilloscope are needed for accurate probe characterization.
Measuring Oscilloscope Probe Rise Times with a Jim Williams Avalanche Pulser
(0:05) The Challenge: Finding a Fast Signal: An initial attempt to test probes with a 74AC logic pulser failed because its 2.4 ns rise time was slower than the probes being tested, making it impossible to differentiate their performance.
(1:58) Building an Avalanche Pulser: To get a sub-nanosecond signal, the creator builds a pulse generator based on a Jim Williams design. It uses a 2N3904 transistor forced into avalanche breakdown with ~120V, creating a very fast pulse from the discharge of a capacitor.
(5:05) Low-Inductance Test Fixture: A "dead bug" style circuit is constructed with a special ground coil. This allows the probe tip to connect with minimal ground inductance, which is critical for accurate high-speed measurements.
(6:06) Probe Test Results: A series of oscilloscope probes were tested, yielding confusing results that were all very close to each other:
Rigol PVP2350 (350 MHz): 1.25 ns
HP Probe (vintage 275 MHz): 1.24 ns
Tektronix P6106 (250 MHz): 1.49 ns
Tektronix P2200 (200 MHz): 1.4 ns
Tektronix Probe (100 MHz): 1.3 ns (Performed better than higher-rated Tek probes)
Generic "Cheap Chinese" Probe: 1.22 ns (Fastest rise time, but showed ringing and incorrect voltage division)
(7:59) Key Takeaway: The Oscilloscope is the Bottleneck: The creator realizes that all measurements are clustered around 1.2-1.5 ns because he is primarily measuring the limitation of his 350 MHz oscilloscope, which has a theoretical rise time of around 1 ns. The probe's actual speed cannot be accurately determined when it is faster than the scope.
(16:14) Conclusions and Future Plans: The stock Rigol probe performed surprisingly well, matching or exceeding more expensive probes in this test. For more accurate results, a significantly faster pulse generator (ideally < 500 ps) is needed. This project serves as a practical demonstration of how the entire measurement chain—not just the device under test—affects the final result.
Summary of Viewer Comments
Viewer comments provided technical corrections, suggestions for improvement, and alternative approaches.
Primary Limitation Identified: The most common comment was that the 350 MHz bandwidth of the oscilloscope is the main limiting factor. Viewers explained that the scope's own rise time (calculated as 0.35 / bandwidth) is approximately 1 ns, which dominates the measurement.
Circuit Construction Advice: Several viewers suggested improvements to the avalanche pulser for better performance, including using surface-mount (SMD) components, ensuring very short lead lengths on the transistor, and building the circuit directly on a BNC connector to measure the scope's baseline rise time.
Alternative High-Speed Signal Sources: Commenters proposed several alternatives for generating fast pulses, such as using a commercial pulse generator (Leo Bodnar), a mercury-wetted relay, tunnel diodes, or even the fast signal edges from an HDMI source.
Alternative Measurement Methods: It was suggested that a more accurate way to characterize probe bandwidth is to use a spectrum analyzer with a tracking generator or a network analyzer to measure the frequency response (-3dB point) directly.
Critique of Methodology: One commenter noted that using a narrow pulse instead of a true step function can skew rise time results, potentially making slower probes appear faster than they are. Others mentioned the importance of checking probe compensation before testing.