TUSB212: Output sign is probability reduction

Minghao,

Thank you for your question.

What is the trace length and EQ setting?

Best,

Stanton

EQ pin is floating， and it seems the trace length is not vert long in PCB.

Hi Minghao,

The change in the signal amplitude near the middle of the packet is likely due to reflections which may be caused by too high AC and DC boost.  This can be captured by using single-ended probes to capture the signals instead of a differential probe.

We would recommend using AC boost level 0 and DC boost level Medium and retake the waveforms.

Best,

Stanton

Customer reply that tusb212 is 950mil to device and about 1.5m to host.

But customer don't know why there is some normal waveform before the problem?

They test that normal waveform is about 300mV (differential voltage), while the small waveform is only 180mV. 

This is schematic diagram.

Customer will make test in few days. I will ask for feedback.

If the problem is solved, I will close this question.