Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_ISAAC] [EXTERNAL] Re: [802.3_ISAAC] Question on Baseline Text Proposal for TDD Based 802.3dm PHY

Hi Wei,

 

In 802.3ch EEE, the receiver has 2560ns (8 frames) to acquire full symbol and frame synchronization BEFORE it needs to receive actual data. In the New-TDD proposal the receiver must acquire full resynchronization with 213ns refresh header (10 times shorter than 802.3ch EEE). The upstream receiver then has another 347ns of data before the receive signal goes silent again for 9040ns. Maintaining frame and symbol synchronization is therefore much harder in the New-TDD than in 802.3ch EEE.

 

To put this into perspective, if the clock offset in 50ppm for 9040ns, the Master and Slave clock will have drifted 452ps, which is 1.37 symbols at 3GHz symbol rate. If the frequency offset is only 5ppm, the drift becomes 45ns or 0.137 symbols, which is still significant drift from an equalization perspective.

 

Another way to look at this is to evaluate crystal frequency drift during temperature change. In this case, the drift can easily be more than 1 symbol per 9600ns period.

 

Based on the above, it is not clear to me that we can assume that the Master and Slave are perfectly timing locked. At least not if we assume that the upstream transmitter (low data rate) is the Master.

 

Ragnar

 

From: Wei Lou <000047a3c8c56bbe-dmarc-request@xxxxxxxxxxxxxxxxx>
Sent: Sunday, July 20, 2025 9:47 PM
To: STDS-802-3-ISAAC@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_ISAAC] [EXTERNAL] Re: [802.3_ISAAC] Question on Baseline Text Proposal for TDD Based 802.3dm PHY

 

Hi, Ragnar, In DATA_MODE, the MASTER and SLAVE are fully timing-locked—both symbol and RS-FEC frame timing stay aligned. This allows both sides to know exactly where each received frame starts and ends, with no cycle slipping between bursts. 

ZjQcmQRYFpfptBannerStart

Prioritize security for external emails:

Confirm sender and content safety before clicking links or opening attachments

    Report Suspicious    ‌

ZjQcmQRYFpfptBannerEnd

Hi, Ragnar,
In DATA_MODE, the MASTER and SLAVE are fully timing-locked—both symbol and RS-FEC frame timing stay aligned. This allows both sides to know exactly where each received frame starts and ends, with no cycle slipping between bursts. It works similarly to EEE mode in 802.3bp/ch, which many are familiar with. 
 
As for delay_count, since the Task Force hasn’t finalized the channel delay, we’ve allowed a range of 0 to 31 for now. This range can be narrowed once the actual delay values are determined.
 
 
Wei
 
________________________________________________________________________
To unsubscribe from the STDS-802-3-ISAAC list, click the following link: https://urldefense.proofpoint.com/v2/url?u=https-3A__listserv.ieee.org_cgi-2Dbin_wa-3FSUBED1-3DSTDS-2D802-2D3-2DISAAC-26A-3D1&d=DwIFaQ&c=nKjWec2b6R0mOyPaz7xtfQ&r=hiHgBSUj2X0k3TORVxe0NCZAlJs6SEDHhwLDz5m9MbY&m=xvKvi1ZkG26FstRLjmdPA59nFQmGjOukBOq4e_a-3YniW8LBJ0e8i4vz52em6MFo&s=ABHsHr5-wCPYUKljJfjdEZmDh5ULsu65gCvkljdYoiw&e=

To unsubscribe from the STDS-802-3-ISAAC list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-ISAAC&A=1