• ???? USB 3.0 ?????? ???? !! ???????????? - ???? HKEPC Hardware

    From Mr. Chang Man-wai@3:633/10 to All on Sun Aug 30 18:36:24 2026

    ???? USB 3.0 ?????? ???? !! ???????????? - ????
    HKEPC Hardware
    <https://www.hkepc.com/26591/>

    ???????? on X: "USB 3.0????????????????????????????
    ???????2.0???????3.0?? ?????????????" / X <https://x.com/ap_kokura/status/2093534111885664549>

    --

    @~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
    / v \ May the Force and farces be with you! Live long and prosper!!
    /( _ )\ https://sites.google.com/site/changmw/
    ^ ^ https://github.com/changmw/changmw
    The game is afoot... Meow...


    --- PyGate Linux v1.5.19
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Paul@3:633/10 to All on Tue Sep 1 06:28:53 2026
    On Sun, 8/30/2026 6:36 AM, Mr. Chang Man-wai wrote:

    ???? USB 3.0 ?????? ???? !! ???????????? - ???? HKEPC Hardware <https://www.hkepc.com/26591/>

    ???????? on X: "USB 3.0???????????????????????????? ???????2.0???????3.0?? ?????????????" / X
    <https://x.com/ap_kokura/status/2093534111885664549>


    First article [Translated via Firefox]
    -------------

    USB 3.0 will slow down! Recently, the well-known Japanese PC chain Applied (?????)
    shared a little knowledge: if the USB 3.0 device is slowly when it is inserted into
    the port, the transmission speed will be reduced from USB 3.0 to USB 2.0, which has
    aroused the attention of the majority of netizens and computer enthusiasts.

    This is true, not any metaphysics or urban legend, but the physical edge state (Edge Case) generated by the USB 3.0 Standard-A interface for ?perfect downward compatibility.? In order to improve the transmission speed on the original USB Type-A physical structure, USB-IF adopted a double-layer metal contact design that is backward compatible. The port opening is shallowly configured with a traditional USB 2.0 signal pin, while the depth is equipped with a USB 3.0-specific
    high-speed data transmission pin.

    The base of USB 3.0 Type-A is located at different depths within the male head, with 9 contacts:

    Shallow (front 4 pins): Traditional USB 2.0 contacts (Power, GND, D+, D-).

    Deep (5-pins behind): USB 3.0 SuperSpeed Dedicated Connector (RX+/RX-/TX+/TX-/GND).

    If you slowly plug the USB 3.0 device into the Type-A port, the USB 2.0 contact in front will
    be touched first, and the host chip will launch a handshake (Handshake). At this time,
    the deep contact has not yet been connected, and the system will directly determine
    as "USB 2.0 device".

    For the device to be precisely identified as USB 3.0, it must be inserted into the
    end at a faster and smooth speed, so that the USB 2.0 and USB 3.0 pins can be exposed
    at the same time in a very short time difference, so that the host can successfully
    enable high-frequency high-speed transmission channels.

    How slow is it to slow down? When the plug touches the USB 2.0 contact in front, the
    power supply (VBUS) is immediately turned on, the control chip of the device (Controller)
    is started and enumeration, and the USB 3.0 controller will launch the SuperSpeed
    link training (Link Training).

    This handshake Timeout is about 100 to 200 milliseconds (0.1~0.2 seconds), so as
    long as you hesitate a little, staton, or push to the end of the moment, the deep
    RX/TX signal foot is not connected or the connection quality is unstable, the system
    will directly determine "SuperSpeed failure, downgrade (downgrade) to USB 2.0 mode."

    Unexpectedly, even USB has the knowledge of pumping speed, depth and rhythm, and gently
    "fast to the deepest" is right.

    ******************* End: Translated via Firefox **************

    The config space has a field that reports the PHY capability of the peripheral.

    Vendor ID : 0x0781 (Western Digital, Sandisk)
    Product ID : 0x5581
    USB Version : 3.2 Gen 1 <=== This USB Root Hub has six ports total...
    Port maximum Speed : SuperSpeed
    Device maximum Speed : SuperSpeed
    Device Connection Speed : SuperSpeed

    Vendor ID : 0x0325 (Unknown Vendor) [Actually OCZ Rally2] Product ID : 0xAC02
    USB Version : 2.0
    Port maximum Speed : High-Speed (Companion Port 1-5 supports SuperSpeed) <=== Same port as before
    Device maximum Speed : High-Speed
    Device Connection Speed : High-Speed

    This allows the system to notice any mis-match, where the negotiation
    did not uncover the full potential of the device.

    Port maximum Speed : SuperSpeed <----+ \
    Device maximum Speed : SuperSpeed \___ "SuperSpeed failure...
    Device Connection Speed : High-Speed <----+ /

    The most likely device type, where this would be a concern, is a plastic-barreled USB stick
    into a metal computer port. Those tend to jam. If the plastic barrel enters on an angle,
    the metal can bite into it a bit (the capture process of plastic on metal was likely
    never considered when the scheme was designed). When metal USB3 keys plug into metal USB3 A type,
    the insertion process is normally a consistent level of friction, mainly because
    the metal-on-metal case cannot happen on an angle like the plastic-on-metal can.
    After "mechanical capture" occurs, metal-on-metal slides straight.

    The feature of the pins is called "Advance Contact". You want VCC and GND to make
    electrical contact first (device powers up). And, you only want that to happen, when the metal barrels have mated sufficiently to define a "straight path for insertion",
    not the "crooked path of plastic on metal" [which can be bad enough to actually break off a USB3 pin in the set of five pins]. Next the D+ and D- connect, and any strapping resistors are detected. Finally, the diff interfaces of the
    high speed RX pair and TX pair touch. The fifth pin is a ground for
    crosstalk "isolation".

    Do NOT buy devices with plastic barrels. Your USB sticks should have only metal barrels.
    I lost good usage of a 128GB USB3 stick, just because the plastic barrel caused one of the USB3 pins to snap off. The connector is so tiny and flimsy, no combination of solder and epoxy glue, can make a sufficient electrical connection again.
    The stick continues to run at USB2 rates, but who wants to fill a 128GB device at USB2 rates.

    Paul




    --- PyGate Linux v1.5.19
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Mr. Chang Man-wai@3:633/10 to All on Sat Sep 5 01:09:22 2026

    [HK01]USB 3.0?????2.0????????????????????
    ???:
    <https://www.hk01.com/%E6%95%B8%E7%A2%BC%E7%94%9F%E6%B4%BB/60385817/>

    --

    @~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
    / v \ May the Force and farces be with you! Live long and prosper!!
    /( _ )\ https://sites.google.com/site/changmw/
    ^ ^ https://github.com/changmw/changmw
    The game is afoot... Meow...

    --- PyGate Linux v1.5.19
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)