TUSB546-DCI: DisplayPort SBU signal behavior

Hi,

I'm looking into this now.

Do we see breakdown in AUX/SBU communication when we see this common mode shift?

The common mode in this waveform look ok. Based on this waveform this looks like is a capacitor charging and discharging. Would there be a reason there is a capacitor charging and discharging here? (physical disconnect/ reconnect may cause this)

Hi,
Thank you for your reply.

>Would there be a reason there is a capacitor charging and discharging here? (physical disconnect/ reconnect may cause this)
=> No, we just implement only AC coupling capacitor same value(0.1uF) as EVM.

>Do we see breakdown in AUX/SBU communication when we see this common mode shift?
=> We do not have any issue on our side. However, we got issue on compliance test site.
After getting issue on compliance test site, when we check waveform, we saw such offset drift. Therefore I asked this question to you.

Best Regards,

Hi,

Thanks for clarifying.

Do we see this waveform consistently or is this only at the start or end of communication?

Hi,

>Do we see this waveform consistently or is this only at the start or end of communication?
At least, we saw start and end of communication.

Best Regards,

Hi,

Can we do some measurement to test if this happens in the middle of communication. This is most likely due to discharging/ recharging of the AC caps during plug and un-plug events.

The AUX interface is an AC coupled interface as seen here.

In a USB Type-C DisplayPort alt. mode system, the AUX and SBU is MUXed as follows. 

Before a cable is connected to the USB Type-C receptacle, the AUX and SBU path is open. This means that the SBU1/2 nets will be pulled down to GND through the 2M PDOWN, the AXP signal will be pulled down to GND through the 100k PDOWN, and the AUXp signal will be pulled up to 3.3V through the 100k PUP. 

When a cable is plugged in the AUX DC voltages will level to the appropriate common mod voltages, but the AC caps will discharge/charge which results in this waveform. 

Ryuuichi machida said:

After getting issue on compliance test site, when we check waveform, we saw such offset drift. Therefore I asked this question to you.

What is the issue we see on the compliance test side? Is this failing compliance?

Hi,

Thank you for your reply.
>This is most likely due to discharging/ recharging of the AC caps during plug and un-plug events
If root cause of posted waveform is related above description , I wonder why DC shift does not occur on first packet.
(As I show following figure, DC shift occur for second packet not first packet.)
Can you explain about this ? 



Best Regards,

Hi,

Additional information about below two.

>>Do we see this waveform consistently or is this only at the start or end of communication?
>At least, we saw start and end of communication.
=> We saw this waveform not only start and end of communication but also other points.

>What is the issue we see on the compliance test side? Is this failing compliance?
=> We failed following compliance test contents.
- AUX_CH (Manchester-II) EYE Test (Normative)
According to result, we observed variation of DC offset of differential signal. We understand that this is different from phenomenon which we posted at first, but we found DC offset variation under single end measurement during investigating root cause of compliance test failure.

Best Regards,

Hi,

Ryuuichi machida said:

Can you explain about this ? 

Is this waveform from the start or end of communication. Its not clear. My understanding was that this is the end of the AUX/SBU transfer meaning that the second packet is the final packet sent.

Ryuuichi machida said:

=> We saw this waveform not only start and end of communication but also other points

In this case this may not be the charging or discharging of caps. Are you able to send the waveforms from before (AUX inputs) and after (SBU outputs) of the TUSB546?