TUSB544: Another verification test failed

Hi,

Majority failures are DP compliance failures with failures associated with the delta between different SWING or Pre-emphasis level.

1. What is the channel length between the source and the TUSB544?

2. Can you please set register 0x0C to 0x60 or 0x68 and re-test the failure item?

3. We may need the source to change its SWING and Pre-emphasis level in order to pass the test.

Thanks

David

Hi David,

As your suggested, they tried 0x68 and 0x60, for 0x68 it is better, but only Lane3 still failed, the test report is under here:

https://tidrive.itg.ti.com/a/SNgH0ExYB89a2AfL/df3fb994-7a9d-40a8-96ba-95576360cf4a?l

I asked them to share the schematics and the PCB layout, they said it has been reviewed by BU before, Quanta. As Realtek and us are p2p, so the external components are following their spec, 0.33uF cap on Lane3, which we don't have it in our reference, will this cause the DC shift of the continued code? The schematics are under the link.

https://tidrive.itg.ti.com/a/cShWX3WoVkFjKFZE/d4a436a8-0341-4346-a694-865519a67415?l

As you said, they may tune the pre-emphasis of the intel platform to fix the problem, can you suggest how to tune it can help on the DC shift?

Best Regards,

Peter Wei

Peter

This design looks like there are 4 TUSB544, which one reported DP compliance failure?

Can you keep 0x0C at 0x68 and reduce 0x10, and 0x11 from 0xAA to 0x88?

Thanks
David

Hi David,

Sorry for the wrong information, after changed the 0x0c to 0x68, all the channels are failed, only the items are less. So the test engineer only provide the lane3's test result to keep the report smaller.

After change the 0x11 to 0x88, it is still failed, only the rising edge changed, but the main problem is the low frequency components. Can we use -1dB gain for trying, since the datasheet is hiding this setting, any problem on it?

https://tidrive.ext.ti.com/u/yeNyCN_jcgPb43IL/2f340e8d-dda8-45bf-927b-9d29e4fe447d?l