TUSB212: How to judge signal mode

Hi Minghao,

Could you have your customer verify whether the CD Pin is pulled high or low and if the ENA_HS pin is pulled high after reset? If both are pulled high, then the TUSB212 should detect that the signal is a high-speed signal and the signal should be being boosted. If either pin is pulled low, then the device either does not detect a signal or does not detect a high speed signal.

The Eye diagrams on page 14 and 15 of the datasheet can be used to determine what the eye should look like with and without boost.

High Speed data is usually transmitted along links that have 45 Ohm terminations on the DP and DM lanes, and a driving current of 17.78 amps.

For board/PCB designs, we recommend customers follow our Layout Guidelines app note which highlights how traces should be pathed, spacing, and Protocol-specific layout guidance.

Please double check what is being output by the CD and ENA_HS pins, and whether the DP/DM traces follow our layout guidelines.

Thanks,

Ryan

Those pin are high in the test. Is there any reason cause that?

Hi Minghao,

If both the CD and ENA_HS pins are high, then that indicates the TUSB212 sees a connected high speed signal, and should be boosting the signal that comes through the device.

Would it be possible to send eyes of the signal before and after going through the redriver? I would like to verify whether the eyes are being boosted. Additionally, for any eye testing, please make sure to follow the test procedure listed in section 8.2.2.1 of the datasheet.

Thanks,

Ryan

Above is customer waveform, which is USB device receive. This is without redriver.

And following is waveform after TUSB212 and the device receive.

So that is why customer thought the wave is bypass the device. How do you think about it?

Hi Minghao,

Going by the waveforms you sent, it looks like the signal is being boosted by our device. This second picture looks like both AC and DC boost are being applied correctly to the signal. Why does the customer believe the signal is bypassing our device? Are they not seeing an improvement in connection?

Thanks,

Ryan

I have attend their test in customer office. In the second picture, there is some signal is smaller desicion threshold, because of reflection. And this is the signal after the redriver. So I don't know why it seems correctly? Could you tell me more detial?

Hi Minghao,

What are the settings your customer is using? Previously their settings were set for max AC and DC boost, but we have advised them in the past that leaving these settings at max when it is not needed can cause reflections and disconnects. Lower DC and AC Boost settings should reduce the amount of reflections seen while reducing the amount of undershoot we see at the end of that waveform.

Thanks,

Ryan

For DC offset is 60mV, they change their set for DC boost and AC boost. While this topic we discuss is a new problem. Not because of reflection.

Last problem has been solved.

Hi Minghao,

Thanks for the update, confirming that lowering the DC boost to 60mV fixes the issue.  Regarding this post, the redriver is active and is boosting, this is observed in the comparison waveforms with redriver and without redriver.  Is the customer questioning if the redriver is active or not?  As Ryan mentioned, the ENA_HS and CD signals are high indicating the redriver is active.  Is there a different issue, as the redriver is clearly active from the waveforms provided?

Regards,

Undrea