Phase continuous waveform generation for ATSC (N.American Digital TV standard).

I am trying to generate a test waveform which is compatible with the
ATSC standard. I am obtaining the MPEG packets from a TV clip obtained
from the web. I am ensuring that the MPEG packet frame rate of the
transport stream packets is compatible with the ATSC standard. The
waveform is converted from IQ -->RF and fed into a set top box, which
decodes the signal and feeds the DVI output into a monitor.

When I generate the RF signal at a TV frequency (i.e. 584.25 Mhz), the
set-top box indicates receipt of a broadcast signal but does not
display the actual video corresponding to the MPEG packets which
constitute the RF signal. Also, the RSSI (received signal strength
indicator) shows only a weak signal and very occasionally shows receipt
of a strong signal...

The RF waveform is being played out of the arb (arbitrary waveform
generator) in a looping mode i.e. the waveform samples are played out
of a circular buffer. My suspicion is that the set-top box gets
confused when the RF signal loops back and the MPEG packets go out of
sequence...Further, all the downstream processing namely descrambling,
Reed-Solomon decoding, convolutional de-intereleaving and trellis
decoding may get affected due to the loop-back in the arb buffer...

Is anyone aware of any easy way to trick a set-top box to display a TV
signal from MPEG packets using an arbitrary waveform generator??How do
I improve the signal quality going into the box?

Any help would be greatly appreciated.

Regards
Vikram

In alt.tv.tech.hdtv cvikram@mac.com wrote:

| The RF waveform is being played out of the arb (arbitrary waveform
| generator) in a looping mode i.e. the waveform samples are played out
| of a circular buffer. My suspicion is that the set-top box gets
| confused when the RF signal loops back and the MPEG packets go out of
| sequence...Further, all the downstream processing namely descrambling,
| Reed-Solomon decoding, convolutional de-intereleaving and trellis
| decoding may get affected due to the loop-back in the arb buffer...

if it just cuts off and loops back in the middle of the data segments
or data frames, it's probably never going to get back in sync.


| Is anyone aware of any easy way to trick a set-top box to display a TV
| signal from MPEG packets using an arbitrary waveform generator??How do
| I improve the signal quality going into the box?

You need to be sure it fully conforms to ATSC even at the loop back point.
An ATSC generator is likely needed.

-- 
-----------------------------------------------------------------------------
| Phil Howard KA9WGN       | http://linuxhomepage.com/      http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/   http://ka9wgn.ham.org/ |
-----------------------------------------------------------------------------

How long is the loop before it repeats?

You may have other issues, there are a lot of MPEG layer things that
have to be right for the STB to diplay video, like  PSIP info and PID
numbers etc.

Do you have the correct DIGITAL RF center frequency?... the RF freq you
listed sounds like an analog video carrier frequency.

There can be confusion about RF freqs.
Analog TV uses the video carrier freq.
Digital CATV (64 or 256 QAM) uses the center freq of the channel
I'm not sure what over the air digital uses (8VSB) it may use the pilot
carrier frequency or it may use the center frequency or something
else???

If your not getting RSSI, you very well may have the wrong RF frequency
dialed in.

Mark

In alt.tv.tech.hdtv Mark <makolber@yahoo.com> wrote:

| I'm not sure what over the air digital uses (8VSB) it may use the pilot
| carrier frequency or it may use the center frequency or something
| else???

There is a specific frequency offset from the lower end of the channel.
I read it in one of the documents but didn't need to remember it's exact
value right now (but it is based on some multiples to minimize adjacent
channel interference).  But I suspect it should work at any frequency as
long as the relationships are correct.

-- 
-----------------------------------------------------------------------------
| Phil Howard KA9WGN       | http://linuxhomepage.com/      http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/   http://ka9wgn.ham.org/ |
-----------------------------------------------------------------------------

Phil/Mark
Thanks for the reply.

I am pretty sure that I have the RF frequency correctly dialed in. In
fact, I tuned into the ABC/KVUE station here in Austin that transmits
its programs digitally using the ATSC standard at this center frequency
(584.25 Mhz).

The ATSC standard employs the 8-VSB modulation which is pilot assisted;
the pilot is transmitted at the suppressed carrier frequency and
located at 310 Khz to the right of the band-edges [a total of 6 Mhz
bandwidth].

However, I think that the problem is associated with the looping time.
The time for which I am transmitting the MPEG stream is 6.59 seconds
after which the waveform loops back...

Now, the tuner may have its own tuning time followed by which it
attempts to perform MPEG decoding. I see a lot of havoc being caused in
MPEG decoding as a result of the loopback.

Is there any way to set some MPEG header information to inform the MPEG
receiver decoder about the phase discontinuity...

Vikram

cvikram@mac.com wrote:
> Phil/Mark
> Thanks for the reply.
>
> I am pretty sure that I have the RF frequency correctly dialed in. In
> fact, I tuned into the ABC/KVUE station here in Austin that transmits
> its programs digitally using the ATSC standard at this center frequency
> (584.25 Mhz).
>
> The ATSC standard employs the 8-VSB modulation which is pilot assisted;
> the pilot is transmitted at the suppressed carrier frequency and
> located at 310 Khz to the right of the band-edges [a total of 6 Mhz
> bandwidth].
>
> However, I think that the problem is associated with the looping time.
> The time for which I am transmitting the MPEG stream is 6.59 seconds
> after which the waveform loops back...
>
> Now, the tuner may have its own tuning time followed by which it
> attempts to perform MPEG decoding. I see a lot of havoc being caused in
> MPEG decoding as a result of the loopback.
>
> Is there any way to set some MPEG header information to inform the MPEG
> receiver decoder about the phase discontinuity...
>
> Vikram

OK, the STB is Receiving on the correct RF frequency...

what RF frequency is the ARB waveform gen set to transmit on?
Did you set the generator to 584.25 MHz?  That would probably NOT be
correct.


Can you compare the KVUE singal and your singal on a spectrum analyzer?
 
Mark

I'm going to guess that you need to set the AWG to 583.310 MHz which is
the supressed carrier freq.

thats 584.25 - 1.25 = the lower band edge = 583 MHz

583 + 0.310 = 583.310

I'm not sure this is correct.

Mark

<cvikram@mac.com> wrote in message 
news:1123223473.886421.91120@f14g2000cwb.googlegroups.com...
>I am trying to generate a test waveform which is compatible with the
> ATSC standard. I am obtaining the MPEG packets from a TV clip obtained
> from the web. I am ensuring that the MPEG packet frame rate of the
> transport stream packets is compatible with the ATSC standard. The
> waveform is converted from IQ -->RF and fed into a set top box, which
> decodes the signal and feeds the DVI output into a monitor.
>
> When I generate the RF signal at a TV frequency (i.e. 584.25 Mhz), the
> set-top box indicates receipt of a broadcast signal but does not
> display the actual video corresponding to the MPEG packets which
> constitute the RF signal. Also, the RSSI (received signal strength
> indicator) shows only a weak signal and very occasionally shows receipt
> of a strong signal...
>
> The RF waveform is being played out of the arb (arbitrary waveform
> generator) in a looping mode i.e. the waveform samples are played out
> of a circular buffer. My suspicion is that the set-top box gets
> confused when the RF signal loops back and the MPEG packets go out of
> sequence...Further, all the downstream processing namely descrambling,
> Reed-Solomon decoding, convolutional de-intereleaving and trellis
> decoding may get affected due to the loop-back in the arb buffer...

Instead of looping, can you just go up and down through the sequence?  For 
example, instead of 1 2 3 4 1 2 34 how about 1 2 3 4 3 2 1 2 3 4 etc.

>
> Is anyone aware of any easy way to trick a set-top box to display a TV
> signal from MPEG packets using an arbitrary waveform generator??How do
> I improve the signal quality going into the box?
>
> Any help would be greatly appreciated.
>
> Regards
> Vikram
>