USB on fire


We have a problem with a destroyed notebook USB port when it was connected to BeagleBone Black.
Our system: we designed an extension board connected to the BBB pin-headers P8 and P9. Let’s call this the ‘ERTEC board’. The ERTEC board supplies power to BBB via the connector P9, pins 5, 6, which is VDD_5V. The VDD_5V rail in BBB is basically the barrel jack input. The picture below shows the setup:


Normally the system works fine.

But now one colleague reported this problem:

  1. He had the Mini-USB of BBB connected to a PC (which is grounded…), and the whole system was powered from a +24V isolated supply and running.
  2. While the system was running, he pulled the USB cable from the PC and connected it to a USB3.0 port of a notebook.
  3. The USB connector of laptop started smoking. The laptop and BBB didn’t work after that.

Any idea why this could happen? So far it seems like an one-off event, but one could never know :-/


The mini-USB power is connected to the BBB power. I can run my BBB just from the USB cable.

So you probably just paralleled 2 power supplies. The list of outcomes includes the escape

of the black magic smoke. Some regulators do not like getting started when there is already

a backward voltage from the load. The LM317 comes to mind. Its data sheet includes info

on how to make it survive. (Just as an example. it’s unlikely that your laptop contains one.)

Maybe it is a good idea to use a USB port replicator during experiments with high-priced laptops.



“smells” like the laptop was grounded…

Either thru it’s powersupply, “ethernet” cable, or something else…

Or even thru the user “holding” it…




we checked that the power source from the other board and from the PC USB are not connected in parallel.
See the picture below: we power BBB via the VDD_5V which is identical as using an external 5V supply with the barrel jack. The PC USB puts power to the USB_DC rail. The U2 TPS65217C is supposed to draw power preferentially and automatically from VDD_5V and not USB_DC. I wish there is a 0R jumper between USB_DC and pin 12 of U2.



Dne středa 17. října 2018 13:52:27 UTC+2 Gerhard Hoffmann napsal(a):