Other Parts Discussed in Thread: TUSB2046B, ALLIGATOR
This design has been in production since 2013. Dual hub primary TUSB2077A and secondary TUSB2046B in series to create a 9 port hub product. There have been no documented problems until recent revisions of Windows 10 interaction resulting in random port disable state. In a static connection to 8 devices all under self power and the hub self powered and ports individually monitored for current draw faults, simply disconnecting the USB link to the host Windows 10 and reconnecting results in 0 to 3 devices, mostly 1 or 2, fail to enumerate properly. Missing in the OS record of connected devices is the end point establishment and device ID. Some but not all of the device "hello" data is transferred. Also associated with the failure is pin 42 PORTDIS is low. If all devices enumerate properly then PORTDIS is high. When devices that are connected to the ports are directly connected to host, the device never fails to enumerate. This leads me to the conclusion that there is some sort of race condition causing the hub to detect a port disable and it remains in that state until the next connection cycle. I need advice from an FAE on how to tackle this problem.
The attached schematic shows pull-down resistors on the DPx and DMx pins - these are unpopulated in the production build and reset duration has been extended with larger R values 232k at R49 and 118.
