2
\$\begingroup\$

On USB host port (and usually on device ports, too) the VBUS line is filtered with a ferrite (Intel EMI Design Guidelines recommends a 500mA part, for example), and that is obviously fine for standard USB ports (500mA is the maximum current available on USB 2).

This time I need to make a charging port, which allows a lot more; specifically a charging downstream port which does both high power and data service. They make special power switches for that which make it somewhat easy (like the ST STCC2540, and TI makes many variant of that).

The big question is: evaluation boards simply tie the switch output with the port VBUS (with the mandatory 150µF capacitor) and no ferrite; also evaluation board are notorious to be lacking in the EMC department so I don't really trust that to be the right thing to do.

The USB specification doesn't talk about ferrites (except saying "do not put them on data lines" for obvious reasons) and the BC standard only talks about voltages and currents.

What would be the right thing to do? Direct connection like the Chinese supplies and the eval boards do or a beefy ferrite? About 1.5A is not trivial but should be doable.

\$\endgroup\$

1 Answer 1

1
\$\begingroup\$

Looked around and found the answer at last.

The answer is simply huge ferrites. Würth Elektronik has reference designs for both USB 2 and USB 3.1 with PD; I'll take the data circuit from the USB 2 AN and the power line filtering from the USB 3 note.

Of course the want to sell their product but for reference the VBUS is filtered by a twin-pi filter: 1206 ferrite, inductor and another 1812 ferrite (personally I think it's overdesigned but if you want to be sure…)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.