RE: Clause 48: enable_cdet
I believe the enable_cdet functionality is the same as the required EN_CDET
signal in the pma service interface of 802.3z. When true, the PMA will
force any serial +comma to align to it's 10 bit output. I believe that is
the implied functionality of the enable_cdet signal in 802.3ae but it's not
In the LOSS_OF_SYNC state, the enable_cdet signal is forcing the PMA to
align to a comma. Once a valid comma is detected on the 802.3ae service
interface , enable_cdet can be deasserted. In 802.3z, there is another
required signal called COM_DET that indicates this condition. There is no
such required interface signal in 802.3ae so it up to the user to figure out
how and where he/she wants to detect this condition and implement the state
The COMMA_DETECT1-3 states are looking for 3 properly aligned commas in the
absence of an asserted enable_cdet signal. This would indicate a high
likelihood that the alignment that was forced in the LOSS_OF_SYNC state
(with an asserted enable_cdet) was a good one.
> -----Original Message-----
> From: Rogers, Shawn [SMTP:s-rogers@xxxxxx]
> Sent: Monday, March 26, 2001 2:10 PM
> To: 'rich.taborek@xxxxxxxxx'
> Cc: '802.3ae'
> Subject: Clause 48: enable_cdet
> Rich, I'd like some clarification on this before deciding on whether to
> enter a comment on D2.3.
> In Figure 48.8,
> In the LOSS_OF_SYNC state, enable_cdet = true.
> Exiting LOSS_OF_SYNC state is dependent on valid signal detect and
> detecting a comma.
> In the COMMA_DETECT_1 state, enable_cdet = false.
> Exiting COMMA_DETECT_1 to go to COMMA_DETECT_2 is dependent on a valid
> How do you get to COMMA_DETECT_2 when comma detect is turned off?
> Also, on a more generic note, why turn off comma detect so early in the
> sync state machine?
> Shawn Rogers, PMP
> Serial Gigabit Products Dev. Mgr.
> Texas Instruments/ 12500 TI Boulevard
> M/S 8732/ Dallas, Texas 75243
> Tel: 214-480-2678, Fax: 214 761-6954 Pager: 972-597-1803