RE: SJTP: Presentation
Ben,
This is a good summary.
The "Objectives" show that we still haven't articulated what this is for, in
the context of the standard.
Slide 2 has (my numbers added):
1. Guard against sub-standard PLL designs
2. Shorten measurement times
3. Make measurements more consistent
4. Perform Diagnostics
5. Fill a hole in the standard !
and Tom adds
6. guard against inappropriate frequency response somewhere.
If this is the list I jotted down, I listed a few weak reasons to see if
someone else could articulate some strong ones.
Going through the list:
1. I think this is the main motivation. What we don't know is whether
any special test pattern is needed, or whether any "long" scrambled data or
pattern will be fine.
2. Is nice, but not essential for interoperability
3. Has a significant lack of consistency been identified?
4. Is not needed for interoperability
5. To fill it, it would help to identify where it is.
6. As 1. We don't specify frequency response in detail. Bad
implementations will be caught by the regular stressed sensitivity test. We
don't know whether any special test pattern is needed.
Slide 3 lists several tests. These have associated pattern criteria
described in the draft standard. Many of these criteria have been crafted
quite carefully. They aren't holes.
So, an interesting study, worth our time, but nothing here justifies
imposing a mandatory way of doing things.
Piers