V/m for USB2.0 HS & FS

[bobb-uk]

Hi all,

I'm a new user of openEMS and am slowly getting to grips with it.
I plan to eventually use it to examine the fields produced from a USB2.0 differential pair (480Mbps HS and 12Mbps FS).
Does anyone know what the excitation amplitude should be (roughly) for those two ?

Thanks,
Bob.

[thliebig]

I would not worry about amplitudes used for excitation. In openEMS everything is linear. Thus you can always re-normalize all results to whatever voltage/current/power/whatever you desire...

[biergaizi]

I plan to eventually use it to examine the fields produced from a USB2.0 differential pair (480Mbps HS and 12Mbps FS). Does anyone know what the excitation amplitude should be (roughly) for those two ?

For reference, USB 2.0 FS is 0 - 3.3 V, and USB 3.0 HS is 400 mV, LVDS-like. The transmitter and receiver has a differential impedance of 45 Ω to ground at each pin (90 Ω differential). See USB 2.0 Specification, Chapter 5.

But you should be clear about what you're trying to do.

If you only want to check signal integrity, you don't need to care about voltages or E&M fields, as long as the S-parameters are extracted, You can check the raw S-parameters by hand by comparing S11/S21 against your link budget. The S-parameters can be used by circuit simulators, which simulates whatever you need - step response, transient response, eye diagrams, etc. These tasks are outside the scope of openEMS, and many circuit simulators are proprietary, but there are a few FOSS tools like scikit-rf (which can calculate step responses from S-parameters), SignalIntegrity by Peter Pupalaikis (which can simulate eye diagrams), Qucs (general-purpose RF simulation).

If you want to use it for electromagnetic compatibility to check radiated interference, then things start to get complicated. In a EMC test lab, you run a stress test of USB traffic for minutes, and receive the radiation meters apart using a reference antenna. But this is obviously impractical to replicate by simulation, since each FDTD simulation runs at nanosecond time scale... But perhaps you can calculate the transfer function (S-parameters) from the data line to a nearby antenna, which allows you to estimate the radiation? But there's the problem of near-field to far-field transformation... Alternatively, perhaps the data lines can be modeled as an antenna to get its radiation efficiency? No idea, but it's an interesting open problem to do it in openEMS. It would be the best to discuss it in a separate thread.

If you want to use it for visualizing real USB electromagnetic fields for educational purposes, only then you would care about the actual signal amplitudes. But it would be an even more complicated project...