I designed a Keyboard PCB using an ATmega32U4 with the data lines wired up to a Micro-USB B Receptacle. Brand: Molex; part number: 1050170001. The USB ground is directly connected to the Micro-USB B receptacle shielding as shown on the double layer PCB design below. I have a ground pour on the F.Cu side of the PCB represented in red, no ground pour on the B.Cu.

PCB Design

So, the issue here is that my keyboard only operates when using USB cables with dedicated shield wire (basically any cable that passes the continuity test on both ends of the USB metal housing). If I use a USB cable that doesn't pass the continuity test on the USB metal housing, the keyboard doesn't work due to insufficient voltage.

I did some probing with a multi-meter on the cable that doesn't work and here are my results:

PCB's VCC - GND: 1.4V

Probing VCC - GND VCC - GND Value

USB Cable voltage across USB cable metal housing: 3.6V

Probing Micro USB Receptacle Probing USB A Cable metal housing Shield voltage value

What I want to find out is:

  • What is causing this issue? Why is the 5V on the USB cable without a dedicated shield wire splitting across my PCB's Vcc and the shield?
  • Why does it only work when the USB cable has a dedicated shield wire? (If I short the metal housings of the cable, the keyboard gets detected)
  • How will this issue affect the board in the long run? Will it fail one day?

Notes: I do have an ESD protection chip (USBLC6-2SC6) running across the data lines.

Thank you for your time.

  • \$\begingroup\$ What is the power consumption of this device? \$\endgroup\$ Mar 7, 2019 at 5:48
  • \$\begingroup\$ @Ale..chenski Measuring the power consumption would be difficult, but I can safely say that it requires below 2.5W as the standard USB port give 5V and 500mA. \$\endgroup\$ Mar 7, 2019 at 6:04
  • \$\begingroup\$ A standard USB port can "give" much more than 500 mA, 1 - 2 A is not unusual. What is so difficult to put 0.1 Ohm resistor in series with VBUS and measure the voltage drop? \$\endgroup\$ Mar 7, 2019 at 6:33
  • \$\begingroup\$ @Ale..chenski Sorry, can you elaborate on what you want me to do and the reason as well? It's difficult because I don't have a 0.1 Ohm resistor. Can you explain how does adding a 0.1 Ohm resistor in series with VBUS helps me measure the voltage drop? \$\endgroup\$ Mar 7, 2019 at 6:57
  • 1
    \$\begingroup\$ @Ale..chenski OMG, you were right, my ground usb ground pin was not soldered properly. After fixing it, everything works. Thank you so much! \$\endgroup\$ Mar 11, 2019 at 13:50

1 Answer 1


If this is a LS device, you can't use LS-grade cables, the cable must be "captive". And there should be no LS detachable cables without shield.

If this is a FS device, then any certified cable should provide enough conductivity to keep voltage drop within 0.25V typical, shield or no shield (125mV over ground wire, and 125 mV over VBUS) with 500 mA current. If your device consumes more than that or you are using bootleg uncertified cables, it is not a USB problem.

Alternatively, there could be bad ground pin, broken, folded inside the cable itself, of cold solder on PCB.


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.