USB Protection Circuit

I have this protection circuit on the USB power to my circuit. It has a ferrite bead (L2) and a set of TVS diodes (D3), but it's still susceptible to static electricity resetting the circuit.

The static hits the ground shield on the USB connector - should that shield be left floating? Have it's own protection circuit to discharge static to ground? Or should the ferrite bead and TVS diodes be able to handle more current?

Thanks very much!

hi,

I always connect the USB cable screen to 0V , the same 0V as the USB +5v/0v on the remote unit, no problems
How long is your USB cable.?

E

Sorry, I left out a detail - that's our USB power circuit, but it's susceptible to static when it's unplugged and powered only from the battery. But that ground is shared for the whole circuit.

Is your 'ground' a true low-impedance connection to terra firma, or merely a circuit common?

It's a circuit common - when it's battery powered, the battery provides the ground for circuit.

Can you connect your 'ground' to your mains earth to see if that cures the problem?

I can test that, but what would this test prove? The reset isn't a problem when it's USB powered, only when it's battery powered.

Edit: Tested, confirmed, static will not reset the device while connected to mains earth.

Can you connect the shield as Eric mentioned in post #2?

Then you might have to work out other means such as EMI or metalic coatings on the "assumed plastic box" . I used the terms EMI not because of the radiation, but the metalicness of it. One would hope that the person holding the device is at the same potential of what's being plugged din. You may be addressing the wrong dot.

In general, only the sourcing device gets the shield, but it better be connected to Earth. USB does have bidirectional data paths. Using the reason is to remove ground loops, but if you can make the devices the same potential your good,

For fun, wrap the device in aluminum foil to the USB shield and see what happens. An ersatz metalic box.