TUSB542: USB communication is disconnected

Yamaguchi-san

Then, the jitter changes with the RX signal on the processor side.
What is the factor that changes the jitter due to the difference in the above parameters?

*** Are we referring to the total jitter in this case? 

When (CNFG_A2, CFNG_B2) changes from (Float, High) to (Low, High), the TX output swing (OS) and De-emphasis stay the same but the RX equalizer changes from 6dB to 0dB. the RX equalizer helps to compensate for the loss between the TUSB542 and the connector and in turn addressed inter-symbol-interference (ISI) jitter. The ISI jitter is part of the deterministic jitter.  

Question.2

In the example circuit in the datasheet (Figure 16), there is no bypass capacitor for the power supply. How much capacitance should I connect to the bypass capacitor to avoid worsening the jitter?

***

• I would recommend a 10uF and 0.1uF decoupling caps for the 1.8V. 
• In the layout, I would recommend routing the 1.8V as a power plane, not as a fat trace. You also want to make sure the 1.8V power plane has an adjacent ground plane.
• Per the TUSB542 datasheet, you also want to make sure the TUSB542 has at least 88% solder coverage of its thermal pad. 

All of these will help reducing the power supply noise, which is a source of random jitter.

As mentioned above, it seems that there is sensitivity to the occurrence rate of the disconnection phenomenon by changing the parameter settings on the RX side. By the way, does the behavior on the TX side affect the RX signal?

*** it depends on the system design. One example is that TX routed too close to the RX which could result in crosstalk effect.

Thanks

David