Minutes of Weekly Meeting, 2012-02-06

Meeting called to order: 11:05 AM EST

1. Roll Call

Adam Ley
Patrick Au
Tim Pender
Carl Walker
Heiko Ehrenberg
Ian McIntosh
Brad Van Treuren
Peter Horwood
Mark Ellis (Tellabs Inc.)
Brian Erickson (joined 11:11)
Harrison Miles (joined 11:18)

Excused:
Eric Cormack

Adam introduced Mark Ellis from Tellabs, who was visiting Asset and had joined this week's call.

2. Review and approve previous minutes:

01/30/2012 minutes:

  • Draft circulated on 01/30/2012.
  • No corrections noted.
  • Tim moved to approve, seconded by Brad. No objections or abstentions.

3. Review old action items

  • Adam proposed we cover the following at the next meeting:
    • Establish consensus on goals and constraints
    • What are we trying to achieve?
    • What restrictions are we faced with?
  • All: do we feel SJTAG is requiring a new test language to obtain the information needed for diagnostics or is STAPL/SVF sufficient? see also Gunnar's presentation, in particular the new information he'd be looking for in a test language
    (http://files.sjtag.org/Ericsson-Nov2006/STAPL-Ideas.pdf)

4. Discussion Topics

  1. Tutorial on IEEE Std. 1149.7 from Adam Ley
    • [Ian] Having taken a quick look at the slides Adam has for us, I think it's likely that we'll need more than one session. Are you OK with carrying on into another week, Adam?
    • [Adam] I am happy to do so if it is of interest to the group and they want me to do that.
    • [Ian] I'm sure they will but let's see how we get on this week. I'll pass control of the meeting to Adam now.
    • {Adam shared his presentation, previously circulated by email}
    • {Due to the impracticality of representing the entire transcript, the following notes only record some key data points and exchanges}
    • [Adam, at slide 9] In TMS(C) and TCK(C), (C) is shorthand for 'compact'. Where that is shown the signal has extended or enhanced functions over the 1149.1 signal.
    • [Brad] I see that nTRST still looks like a discrete signal for the asynchronous reset?
    • [Adam] It is a discrete. It is drawn with a dashed line to show that it is optional. Where it is not present, the chip is assumed to have it's own power on reset mechanism.
    • [Slide 13] IDCODE is mandatory in 1149.7, while it is optional in 1149.1.
    • [Slides 14/15] Zero Bit Scans (passing through DR scan without entering Shift-DR state) in Bypass are effectively NOP instructions. The number of ZBS cycles determines the current control level, which is then locked in by entering Shift-DR.
    • [Slide 16] Commands can be supplied in Class T1 at command level 2. This is done without the use of TDI/TDO, by counting the cycles that you stay in DR Scan. 5-bit opcode may be followed by 5-bit operand in a second DR scan. A third DR scan may provide an argument for selecting a path in the Extended Protocol Unit.
    • [Slide 17] In T2 scan formats, chip level bypass can bypass both IR and DR.
    • [Slide 19] Adam noted that the bit fields on this slide need to be verified. NODE_ID allows up to 256 instances of the same part on the network.
    • [Tim, at slides 21/22] Is there a maximum TCKC rate for 1149.7? Does it tend to be higher, or make use of system clocks?
    • [Adam] The intention is that 1149.7 should support 100MHz clock, and there is no other clock source.
    • [Tim] So you could use something more reliable like LVDS to route TCKC and TMSC?
    • [Adam] There is nothing specific in the standard, just some timing. Levels are not defined, just as they're not defined in 1149.1, but differential TCKC and TMSC would be nonstandard.
    • [Tim] I meant through the backplane and then convert back to single ended on the board.
    • [Adam] Remember that TMSC is now bidirectional, so translator would need to be protocol aware.
    • [Tim] OK.
    • {Stopped after presenting slide 23 as time had run out}
    • [Ian] Can we continue this next week?
    • [Adam] I should be OK for that. If any one has any questions or anything they'd like me go into more detail on, can they let me know for next week so I can have something ready?
    • [Ian] If emails are sent to the group then that might prompt others to think of their own questions. {ACTION}

5. Key Takeaway for today's meeting

None.

6. Schedule next meeting

Next Meetings:
13th February - Eric will be absent, Heiko may be absent
20th February
27th February

7. Any other business

None.

8. Review new action items

  • All: Any questions regarding Adam's 1149.7 presentation or suggested topics for more detailed explanation should be sent to the group - group(at)sjtag.org.

9. Adjourn

Carl moved to adjourn at 12:09 PM EST, seconded by Patrick.

Respectfully submitted,
Ian McIntosh