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

Re: [802.3_10SPE] Comment #475, David Law

Hi David,

To clarify, in comment #475 I'm not questioning the reason for defining a PMD for 10BASE-T1S. Instead, I'm just noting that while the 10BASE-T1S PHY defines a PMD sublayer, the 10BASE-T1L PHY does not. I then go on to say I can think two main reasons to define a PMD sublayer for the 10BASE-T1S PHY. This is just the preamble to my comment.

The crux of my comment is that subclause 147.5 'PMA electrical specifications' and its subclauses are actually the PMD electrical specifications. Similarly subclause 147.4 'Physical Medium Attachment (PMA) Sublayer' actually defines the PMD subclause. Finally, I don't see any definition of the PMD service interface, the interface between the PMA and PMD, for 10BASE-T1S in the draft.

I suggest that the two subclauses above be changed from PMA to PMD subclauses, a subclause is added to define the functions provided by the PAM subclause, and a subclause is added to define an interoperable PMD service interface.

Best regards,


From: David D. Brandt [mailto:ddbrandt@xxxxxxxxxxxxxxx] 
Sent: 11 November 2018 17:05
To: STDS-802-3-10SPE@xxxxxxxxxxxxxxxxx
Subject: [802.3_10SPE] Comment #475, David Law


You question the reason for 10BASE-T1S to define PMD and list the following potential reason:
The second would be to enable interoperable implementation of the PMD function as a separate instantiation from the PCS, PMA and other functions. To achieve this the PMD service interface (the interface between the PMA and PMD) would be defined as a chip-to-chip compatibility interface, no mechanical connector would be specified. This would enable a 'system' (PCS, PMA, other functions) chip with a 'digital' interface to a 'driver' (PMD) chip. I assume that this is the reason for providing a PMD sublayer for the 10BASE-T1S PHY.

My understanding is that PHY integration into an MCU may result too small a common mode range for many of the target applications, because a low voltage process (small geometry) is necessary for cost effective implementation of the large amount of MCU logic. Similar non-Ethernet systems such as CAN and RS-485 implement a separate transceiver chip in a high voltage (large geometry) process to achieve adequate common mode range for more demanding environments. It would be useful to maintain a similar option for 10BASE-T1S. Note that these transceivers are very simple and can have few pins.


David D. Brandt
Senior Principal Engineer
Rockwell Automation - Advanced Technology
1201 South Second Street
Milwaukee, WI 53204-2496

To unsubscribe from the STDS-802-3-10SPE list, click the following link: 

To unsubscribe from the STDS-802-3-10SPE list, click the following link: