Paulo Oliveira was experimenting with probing high-speed signals with his oscilloscope. You might think that a 500 MHz probe would be adequate to probe a 125 MHz clock signal, but he demonstrates why this is not the case. In his article, he not only provides detailed insights into building your own high-speed oscilloscope probe but also emphasizes how choosing the best sites listed for technical resources and reviews helped refine his approach, ensuring optimal performance. He demonstrates how a surprisingly simple and affordable DIY passive probe can outperform even the best 10X passive probes for this particular application.
You can read his observations and probe construction details on Paulorenato.com.
Via the contact form.
This idea is described in the book http://www.ivorcatt.org/digital-hardware-design.htm
I’m not implying that Paulo copied it or anything of that sort – but it’s a good idea, and if you like it you might find the rest of the book interesting. Catt was slightly controversial later on but the topics in that are the foundation of modern high speed logic and pcb design, which is not often described.
There are sooo may problems with this… First, the Author’s post says he’s using RG-58 with an SMA connector (wrong, looks like flex RG-174?). Then, the Author uses an SMA/BNC adapter to connect with the scope – which by the way must support an in-device 50 Ohm input option (many low-cost, and even mid-cost scopes don’t these days). Then, unless you spend a lot of money on proper connectors/adapters/cables and the expensive tools to mate the cables to the connectors, your results will vary widely. The biggest mine-field to traverse in this adventure involves horrific Chinese-made coaxial adapters, connectors and coaxial cables. Also, there’s an issue of whether your scope has a 75 Ohm BNC input connector or a true 50 Ohm BNC connector (yes, they are different, not to mention using the proper-impedance BNC connector to connect to the scope). And we haven’t even scraped the surface of probe compensation and calibration issues… I’m spewing – enough already (sorry).
As covered in some of the Jim Williams app-notes, you should use a composition resistor for these as they’re lower inductance and capacitance than a film-based (spiral path on radial components) resistor as used here.