DP83867CR: RGMII TX and RX jitter specification

Hi Tarik,

Unfortunately, we don't have the specific jitter specification for TX and RX ports on RGMII communication. May I ask what is the jitter budget that you are looking for?

--

Regards,

Hillman Lin

Thanks Hillman for your answer. It is however unfortunate that you don't have such numbers. I am not looking for specific jitter budgets. I intend to connect two of your Eth phys through a wireless connection and I need to know the jitter budget that I have in hand (based on the phy's RX and TX jitter parameters) to estimate the RF link budget that my connection can handle.

I could find such numbers for Eth phys from your competitors and wanted to make a similar evaluation for TI's phys. In the absence of such numbers, I could make dedicated measurements on your EVKs but I would only get "typical" numbers and this wouldn't give me the max values (accounting for process, temperature impact).

Hi Tarik,

Hillman is out of office today, and will respond tomorrow.

However, in response to your statement, can you please provide a block diagram for how the PHYs will be connected via wireless?

Sincerely,

Gerome

Hi Gerome,

I cannot disclose all the details here. Let's simply say I take the RX and TX RGMII diff signals of one phy and connect them to their counterparts on the second phy with the signals going through a transmission medium. My objective is to know how much jitter budget I can reserve for this medium, regardless of how it is made.

Thanks again

Tarik 

Hi Tarik,

Unfortunately, we don't have the exact number in the RX and TX jitter. If possible, could you share the jitter budgets range that you are looking for? We can reach out to the team internally based on your requirement.

--

Thank you,

Hillman Lin

Hi Hillman,

Something in the 30%.UI would be good.

Tarik

Hi Tarik,

30% UI should be good for TX and RX jitter budgets.

Just want to make sure my understand is correct. May I ask what is the value for 30%UI in the units seconds you are looking for? What speed are you operating at for your PHY?

--

Thank you,

Hillman Lin

Hi Hillman,

Sorry for the late answer, I was OoO.

I plan to use both 1GbE and 2.5GbE meaning that actual data rate would be 1.25 Gbps and 3.125 Gbps. 30%UI would then be 240 ps and 96 ps respectively and this would be the max amount of jitter that my medium would add between your PHT TX and your PHY RX.

Tarik

Hi Tarik,

Hillman is currently OoO, please expect a follow-up before Thursday 2/29.

Thank you,

Evan

Hi Tarik,

Thank you for the calculation. DP83867PHY only able to support up to 1GBE so I will reference the max jitter value as 240ps. DP83867PHY have the jitter of RX and TX within 30%UI. 

--

Regards,

Hillman Lin

Hi Hillman,

I am sorry but this is not clear to me. Is the above a confirmation that, in 1GbE mode, the DP8386PHY RGMII RX is able to sustain data jitter as high as the sum of worst jitter output of the DP6886PHY RGMII TX + 240 ps?

Tarik

Hi Tarik,

I might miss-understand your requirement. We only have the specification that RGMII RX (output signal) will have jitter less than 30%UI.  RGMII TX (input signal) can tolerate 30% UI jitter but make sure setup and hold time is greater than the jitter tolerance.

--

Regards,

Hillman Lin

Hi Hillman,

Ok but then the input can probably tolerate more than 30%UI because of 30% is the maximum (and the output jitter could be as high as 30%UI) then there wouldn't be any jitter budget left for the medium.

If output max jitter is x and input tolerated max jitter is y then the max medium jitter would be z = y - x and it is this z value that I wanted to get.

I hope my ask is clear at least so that you can see if you can answer it or not. 

Thanks again for your help

Tarik

HI Tarik,

Sorry for miss understand your requirement Tarik, I don't think I have the information that you are looking for.

--

Sincerely,

Hillman Lin