The week just after Labor Day always seems to me to feel more like the beginning of the year than the week after New Year’s Day. I guess that’s because vacations are winding down and students are heading back to school. If your education needs a boost, a good way to get started is with a series of YouTube videos produced by Keysight’s EEsof EDA Division.

In an interview at Keysight’s Santa Rosa headquarters earlier this summer, Kaelly Farnham, marketing programs manager for EEsof EDA, said the video  “How to Design an RF Power Amplifier” is the most popular, having been viewed over 16,000 times.

In that video, Matt Ozalas, an RF engineer at Keysight, teaches you the basics of PA design in about 12 minutes. He describes his project workspace, which includes an interactive waveform generator, which supports design, generation, and analysis of RF waveforms.

Ozalas notes that RF engineers tend to understand power from a frequency-domain perspective, but he adds that the successful PA designer will also understand power from a time-domain perspective—which allows you to easily separate RF and DC. He goes on to present a practical power-amplifier topology and an analysis of the waveforms involved. He also makes available his workspace files for download so you can explore further on your own.

In another video, Heidi Barnes, an applications specialist for high-speed digital signal integrity, presents “How to Use Fixture De-embedding to Match Signal Integrity Simulations to Measurements.” She describes a PCB with connectorized structures for measurements. She presents as an example a PCB stripline of width W (50 Ω) that increases to 3W (25 Ω) for 2.54 cm and then reverts to W, resulting in reflections.

Measurement of this structure results in an eye diagram that doesn’t match the simulation. That, she explains, is because the measurement involves more than just the copper trace—it includes the fixture. “Differences in the fixture between simulation and measurement,” she says, “are the leading cause of poor simulation to measurement correlation.” The solution is to use S-parameters and matrix math to remove or de-embed the fixture from the measurement or to embed it in the simulation.

She notes that simulation tools enable a four-step process:

• removing the fixture from the measured S-parameter data,
• verifying and calibrating the simulation with the DUT measurement,
• embedding the fixture back with the DUT model, and
• demonstrating how a digital waveform simulation can match a measurement.

As does Ozalas, she makes her project files available for download.

As of today, there are 21 videos posted.