SuperSpeed, host and device clock.

[Barry Twycross]

That's the idea.

I see that the 3.1 spec has a whole extra bit about timestamps. It looks like its a lot more thorough on the sort of thing you're wanting to do. I don't know how much of this will apply to 3.0 hosts, which are the majority of hosts you're likely to find.

[ChongHan]

Hi

Me again. Do you all have some screenshot for ITP timestamp packet? A continuous ITP packet?

Is the Delta value in the ITP is fix for each link? Or it will change time by time or interval by interval?

Thanks in advance.

[Barry Twycross]

Do you all have some screenshot for ITP timestamp packet? A continuous ITP packet?

A screen shot of what? Its fig 8-31 in my copy of the 3.1 spec. I don't have any super speed analyzer traces available. I don't know what you mean by "continuous ITP packet".

Is the Delta value in the ITP is fix for each link? Or it will change time by time or interval by interval?

Of what you mean by the Delta value.

[ChongHan]

A screen shot of what? Its fig 8-31 in my copy of the 3.1 spec. I don't have any super speed analyzer traces available. I don't know what you mean by "continuous ITP packet".

Yes. I have the Spec too. However in spec it just shows the theory only. As now i dint have any Hardware SuperSpeed analyzer, I can't really see the exact practical value in the ITP packet. How it's look like in real transfer.

Of what you mean by the Delta value.

According to the spec, the delta value is the value between the bus interval boundary and start of ITP transmission by root hub port. I check it myself by programming the Cypress DVK by sending the ITP back to host by Bulk transfer. Then at the host side, i use the SuperSpeed Software Analyzer to see the data received by host. For this, i see that the Bus Interval Counter is counting 1 by 1, but the ITP delta is always zero. The problem nw is there will a random bus interval has a non-zero delta value. From spec, the ITP is transmitting periodically, so from my understanding, the delta should be same for all bus interval, if delta unit is 1, then all remaining delta should be 1 too. Now the delta is randomly have a non-zero delta value in random bus interval, delta is other Bus interval is zero at all.

So I am seeking for prove is that my understanding is right or not. Creating Bus interval and ITP is periodically in 125us(8kHz).