[opendtv] Re: DVB Receiver Standard

  • From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
  • To: <opendtv@xxxxxxxxxxxxx>
  • Date: Fri, 9 Dec 2005 17:45:37 -0500

Dale Kelly wrote:

>> It reads more like a combination of A/53 and A/74.
>> It describes  the standard as well as give performance goals.
>> Interestingly, the terrestrial receiver requirements are not
>> all that stringent.
> I obviously failed to make my point; which is that the DVB
> standard protects those who purchase a receiver labeled as
> such, verses the ATSC non standard, that basically says: "buyer
> beware". However, the new information from Terry Smith may
> indicate that the DVB possibly does note enforce its labeling
> policy other than by listing which product on the market are
> unregistered.

And that was interesting info from Terry. But I didn't miss
your point, I simply moved beyond it, to see whether Nordig in
fact did anything to "guarantee" good reception any more than
A/74 does.

And what I discovered is that it does not. At least, not that I
could see, with respect to terrestrial reception at 3.3 b/s/Hz.

The RF receiver voluntary reception parameters stipulated in
Nordig are in many ways not as stringent as A/74. Which means,
just because you comply with Nordig does not mean your receiver
will work any better than one which complies with A/74. The
single exception to this is echo tolerance in terms of max/min
usec, where the usec requirement for echo is stated as being
from 1.95 usec (note: not closer to 0) and 95 percent of GI.
Note how this does not encompass a really hard requirement,
which is Brazil E.

For example, where are the N +/- 2 and N +/- 3, etc. specs in
Nordig? Maybe they are covered by the single 28 dB requirement
for "image channels," I don't know for sure. If so, A/74 is
quite a bit stricter.

For example, Nordig stipulates a max signal strength of -35 dBm
at 3.3 b/s/Hz, and -77 or -69.6 dBm sensitivity. A/74 doesn't
specify the sensitivity with interference, but without
interference, it says -83 dBm sensitivity, and max signal
strength is -8 dBm. So both are quite a bit more stringent.

Nordig says the C/N margin at 3.3 b/s/Hz with no echo should be
20.2 dB, yet all the receivers tested by the CRC measure right
at 15 dB. I don't think anyone would be impressed with a
receiver that can just manage 20 dB C/N requirement in gaussian

In terms of C/N margin in echo environments, Nordig says that
with a single 0 dB echo, at 3.3 b/s/Hz, a C/N margin of 27.6
dB is required. And yet in the more difficult Brazil E tests,
the CRC has measured all the newer ATSC receivers at 25 or
better (the Samsung came in at 19.9 dB). So again, complying
with Nordig would not be an automatic formula for success, as
far as I can tell.

I understand that A/74 is not the min standard document that
Sinclair wanted from the FCC (and, FWIW, I agreed with
Sinclair). However, it HAS BECOME the document that Nordig is,
at least as far as testing receivers goes. Check out the CRC
tests of the Samsung receiver. They look at just about all the
A/74 recommendations, and test against them. *Including* the
echo profiles in Wash and NYC, which the CRC emulates.

> The DVB standards requirements  the are not "goals". The
> document states that receivers "shall" meet certain
> requirements to qualify as DVB-T receivers. One very
> important performance specification unaddressed by the ATSC
> document is that of receiver noise figure. Industry sources
> have stated that the typical ATSC receiver, using the cable
> ready configuration (the vast majority of receivers sold), has
> a noise figure of 12 to 14 DB. On the other hand, DVB allows
> only a maximum of 8 db and recommends 7db or less.

Okay, so let's add noise figure to A/74. I'm not sure why you
need to specify noise figure if you specify the sensitivity, the
C/N margin, and the bandwidth. I'm thinking it is C/N margins
and sensitivity that we're really after, right? But sure, add
more parameters to A/74, if that helps get a better prediction
of performance.

Noise figure equals dB sensitivity - thermal noise floor - C/N
margin for TOV - 10*logbase10(bandwidth). So, all else equal, the
noise floor of an 8 MHz bandwidth receiver has to be better than
a 6 MHz bandwidth receiver, agreed?

> That is a non-sequester.  I believe that we should establish a
> new rule for this forum; one can't use" vapor ware" product
> performance to support any argument,

Are CRC test results vaporware? Nordig and A/74 fit exactly the
same purpose, as long as the test measurements done on receivers
must be compared with these documents. Obviously, if some
manufacturer merely makes a claim, that doesn't hold much water.
But the CRC results, I suggest, are not just baseless claims.

Not a non-sequester at all ! ;)

You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at 

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

Other related posts: