Hi Biser,
Thanks for the prompt response and your suggestion.
Unfortunately, I cannot change the hardware right now.
Like I said, I do have a workaround - the device enumerates everytime if I revert the patch mentioned above.
So I am trying to understand why the small code change seems to have this effect and why it is not consistent (as in on some reboots the device is enumerated and on some reboots not).
Do you think this inconsistent behaviour is due to the 5V directly tied to USB1_VBUS ?