[opendtv] Re: Zoran, Thomson team on digital TV reference design

Richard Hollandsworth wrote:

>Have you read A/74???  It's a bunch of Guidelines....with
>few real "spec" requirements.  It is more of a test description,
>rather than PASS/FAIL test criteria.

Of course I've read it. And it is exactly what it claims to be, which
is guidelines that are based on field measurements of echo
ensembles.

>The text only "suggests" a "typical" performance range of  -10
>to + 40 usec, which is shown in Fig 4.3 for D/E > 6 dB.

It simply claims that this is the typical bad multipath environment
encountered in real life, and recommends this range as minimum
acceptable. It also recommends a dynamic range of -83 dBm to
-8 dBm. Figure 4.3 is a summary of the echo tolerance
recommendations, and only the Samsung meets it, as I recall
(I don't have the CRC results handy here).

However, Brazil E consists of 3 0 dB echoes that are only 1 usec
apart. So this very difficult ensemble, even for COFDM receivers,
is within the Figure 4.3 description.

>Larger values for pre-echo and post-echo are also shown on
>the chart, with increasing ratios for tolerated D/E.  But it
>appears that these are outside the "suggested" range.  So
>when a manufacturer claims that they "meet" A/74
>Guidelines, which of these test criteria were met???  My
>guess would be only -10 to +40 usec range....

So? The other guidelines are relatively easier to meet. Even
2nd gen receivers were capable of large pre-echo tolerance
when that pre-echo was 15 dB lower than the strong path.
The dark trace in Figure 4.3 is what they think is the most
important guideline to meet, and I find that reasonable, and
certainly better than what pre-Linx receivers could do.

I agree that there are a lot of unspecified parameters here,
in part due to lack of data. But what's discussed is sensible.

>There are a set of Laboratory Test Ensembles, some of
>which require varying the SNR.  However, the PASS/FAIL
>level is not stipulated in A/74

Well, I don't know how they could have been clearer on
that score. They took actual measurements in Washington
and NYC, and said, in essence, rather than give you exact
specs, here are some difficult environments that you can
be expected to see. Presumably, manufacturers would
test their receivers in tests designed to simulate those
environments. Of course, they can pretend.

>Note that the prototype Zenith and Linx STB's had a
>variety of problems and very clearly did NOT meet
>those A/74 Guidelines that are somewhat clear, such
>as sensitivity and single signal overload

But wasn't that EXACTLY my point? That even though
the LG prototype performed quite acceptably in Mark's
apartment, it didn't measure, in terms of echo tolerance
(including dynamic echo) or sensitivity or selectivity, as
well as the Samsung which did match A/74 in these key
areas. THEREFORE, at least in Mark's apartment, it
appears that those parameters discussed in A/74 must
be at least good enough. As specified, they are
obviously not the weak link at that site.

And receivers that perform well there have tested
well elsewhere too.

So again, to A/74 ought to be added more detailed
discussion of other topics, evidently, whatever it was
that made the LG prototype so good *in spite* of its
echo tolerance and sensitivity shortcomings. As far as
it goes, though, A/74 is not a bad start.

Bert

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

 
 
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at 
FreeLists.org 

- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: