3
\$\begingroup\$

What is the role of the outer metal connector? (the one with two square holes) Does it have a potential? Is it insulated from the 4 pins? Are the 4 pins insulated from each other?

enter image description here

\$\endgroup\$

marked as duplicate by Maple, winny, Axeman, Lior Bilia, Voltage Spike Aug 30 '18 at 20:25

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

8
\$\begingroup\$

The metal shroud around USB connectors is called "shield".

The shield serves two purposes, (1) To protect from over-the-air ESD events, and (2) to shield internal high-frequency noise from being emitted out and meet emission regulations for EMI levels. These two processes have different electrical characteristics, so the treatment of shield connection must have somewhat more sophisticated handling than just grounding. It also depends whether this is a portable device, or a stationary equipment.

(1) ESD event is a high-voltage one-time discharge pulse about 50ns long. The purpose of shieled is to intercept the plasma filament, and route the discharge current (10 A-50 A) aside from the signal ground at pins of ICs. Therefore the best way is to ground the shield solidly to system ground plane and chassis.

(2) EMI: USB uses high-speed signaling, which employs internal switching frequencies of 480MHz and higher harmonics of it. Even if the external signaling is differential over a shielded bi-axial cables (which is supposed to cancel direct emissions), inner workings of digital electronics and unbalanced return currents create so-called "ground bouncing". In essence, the digital signal ground in the device is noisy, bouncing. If the shield is directly connected to this ground, the conductive braid along the entire USB cable will emit as a good antenna. Therefore, the shield should be disconnected from digital ground.

As one can see, these requirements are contradictory. The industry solution is to use a de-coupling circuit between the shield shroud and signal ground. Different manufacturers recommend slightly different solutions for this filter. Googling for something like [EMI shielding and ESD protection of computer interfaces] will give a lot of recommendations, LC, RC, etc.

My preference is to use a 0.1uF ceramic cap 0603 size, with a resistor of about 330 Ohms to provide galvanic path. The filter works on specifics of this particular size of ceramic caps. These caps have an impedance of capacitive type at frequencies up to 10-20MHz (so a 50ns pulse gets well-coupled with ground plane and dissipates in power supplies), but at frequencies above 50MHz it becomes an inductor, so if provides a good de-coupling of cable shield from the noisy digital ground.

\$\endgroup\$
3
\$\begingroup\$

The outer metal shield of the USB connector is part of the cable screening or shielding. The cable will have either a woven wire sheath that is connected to the metal connector parts OR the cable will have a conductive foil wrap that is intermingled with a single wire bare wire that connects to the connector shield part.

When the cable is plugged into the mating equipment the cable connector metal shield contacts with the like shield of the connector on the equipment. The shields of the equipment connectors are connected to the chassis or internal ground system of the product. One of the purposes of this shielding connection system it to keep the two attached systems at equal potential.

Most often the connection of the equipment connector shield is to the same internal circuit node as the USB GND wire. Careful design is often used to arrange things so that the shield shunts external electrical disturbances, such as electrostatic discharges (ESD), directly to the chassis in the shortest path possible to a possible earth wire connection. This shunting attempts to keep as much of the high voltage ESD out of the internal circuitry as possible where it could cause damage to the electronics.

Another purpose of the shielding is to help prevent internal RF signals in the products from being emitted out on the cable. It is often necessary to arrange the product USB connector shields a part of an EMI containment system to prevent RF leaks from going over legal limits.

\$\endgroup\$
  • \$\begingroup\$ does the metal connector have a voltage? \$\endgroup\$ – minusatwelfth Nov 14 '16 at 11:01
  • \$\begingroup\$ One would hope that it would not have a "voltage" during normal operation. Like I said it is normally connected to the equipment GND. \$\endgroup\$ – Michael Karas Nov 14 '16 at 11:10
  • \$\begingroup\$ Hmm, is the connector which is touching the sheath, touching the live wire? A lot of devices don't have a ground prong at the mains side, so would the outer connector have some voltage, in that case? (it would be connected to a ground that doesn't exist) \$\endgroup\$ – minusatwelfth Nov 14 '16 at 11:54
  • \$\begingroup\$ For devices that run on battery or AC/without Earth Wire the USB connector shield is still simply connected to the internal DC GND. The internal DC GND in this situation is most likely the equivalent of an internal "chassis" or EMI shield case. Some cases a PC (or other type of equipment) at the other end of the USB cable will have the USB shield connected to the DC GND and to the chassis + earth GND. Like I said before one intent of the USB cable shield is to help keep the two equipment shield connections at equal potential and applies whether there is a earth connection at (continued) \$\endgroup\$ – Michael Karas Nov 14 '16 at 12:13
  • \$\begingroup\$ (continued from above) either end or not. The design should not however attempt to use the shield connection for transferring current for the USB +5V circuit. That is what the USB_GND wire is for. \$\endgroup\$ – Michael Karas Nov 14 '16 at 12:14
0
\$\begingroup\$

The outer metal connector is the USB connector shield. It will shield the wires from EMI. Note that the shield should not serve as connecting wire, i.e. should just be grounded by one end.

\$\endgroup\$
  • \$\begingroup\$ if it was NOT grounded (no third prong at mains), would the emf induce a potential in the connector? \$\endgroup\$ – minusatwelfth Nov 14 '16 at 12:00
  • \$\begingroup\$ Yes, if it is not grounded it will not have the desired effect. \$\endgroup\$ – staringlizard Nov 14 '16 at 15:29

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