[sci.electronics] More need advice on descrambler

rom@xor.Sun.COM (Achyutram Bhamidipaty) (06/23/89)

Follow up to request for help with building a descrambler.

Just to clarify I do have access to an oscilloscope, but its just an old
tektronix that can withstand a direct ICBM hit.

Some notes on what I have observed about the scrambling method.

The sound is fine, when I tune to a scrambled station I can hear all the audio.

If the cable company does anything to the vertical sync it does not affect my
TV, because the picture does not roll.  My assumption that the vertical sync
is normal my not but right but it makes some sense.

The horizontal sync it messed up. I have noticed some interesting things 
though, one is that when the picture is totally out of sync (horizontally)
the color seems normal. Any bits of the picture I can see look ok. However
once in a while the TV will actually lock in the entire picture, when this
happens the picture has really strange colors, it looks like the negative of
the picture, but the only thing I can say for sure is that when the picture is
locked in the color is wrong.

I have hooked my scope up to the monitor output of my TV and this is what the
signal looks like:

###########                  ##########
###########    |--|  -X-|    ##########
###########    |  |  |  |    ##########
###########    |  |  |  |    ##########
###########    |  ---|  |    ##########
###########____|        |____##########____ etc...

The '#' sections are where the picture information is stored. The X represents
a small section where a low amplitude sine wave has been put in, I think. I dont
get a steady trace for the X or the # areas, so the X may have something to
do with the picture. For the sake of having a reference I will show what I get
when I look at the signal for an unscrambled signal.

###########                  ##########
###########                  ##########
###########                  ##########
###########                  ##########
###########                  ##########
###########_____  ___X_______##########____ etc...
               |  |
	       |  |
	       ----

I am doing this from memory I cant remember exactly where the X goes in the
unscrambled version. But I do remember that it was placed in a location 
similar to the scrambled version.

Some other notes on the situation. I plan on using the monitor output of my TV,
this way I dont have to deal with demodulating the signal from the cable 
company. So the final set up, if it works will look like this:

--------|       |-----------|       |--------|
  TV1	|-------|descrambler|-------| TV2    |
--------|       |-----------|       |--------|

-Rom
rom@xor.sun.com

wolfgang@mgm.mit.edu (Wolfgang Rupprecht) (06/23/89)

In article <111777@sun.Eng.Sun.COM> rom@xor.Sun.COM (Achyutram Bhamidipaty) writes:
>The sound is fine, when I tune to a scrambled station I can hear all the audio.
>I have hooked my scope up to the monitor output of my TV and this is what the
>signal looks like:
[picture of the h-sync pulse moved up some, by a dc offset.]

It looks like all they are doing is hiding the sync pulse by adding in
a dc offset during the H-sync interval.  This makes it hard for the TV
to lock onto H-sync (since it is expecting a more negative [blacker]
signal than the normal black).  This will also make it hard for the TV
to lock onto the color-burst which sets the phase for the color
demodulation.

What you have to do is subtract this unwanted DC back out.  On the
system I've seen, the stable reference that you needed to lock into
was actually carried on the audio!  Just get a simple phase-lock-loop
(PLL) such as the kind for FM stereo decoding and tap it into the
demodulated audio before any <15khz filter.  Set the PLL up to have a
center frequency of 15.75 Khz.  Now take two one-shots (eg. 555's).
Hook them up so that the first one is a variable delay for 0-64us
triggered from the PLL.  Set the second one up to trigger off of the
first one-shot and have a pulse width of slightly less than the
"scrambling pedestal".  Now using the scope in dual-trace mode, set the
delay so that your pulse is contained within the scrambling pedestal.
Resistively mix the two signals such that the H-sync pulse now goes
negative relative the the rest of the signal again. You only want the
*real* H-sync to drive the signal negative.

If you can't find a reference sync signal on the audio, you can always
*try* to sync onto the video signal directly....

-wolfgang

Wolfgang Rupprecht	ARPA:  wolfgang@mgm.mit.edu (IP 18.82.0.114)
TEL: (703) 768-2640	UUCP:  mit-eddie!mgm.mit.edu!wolfgang

fiesta@cbnewsi.ATT.COM (eric.c.beck) (06/24/89)

From the "oscillographs" it appears that the scrambling involves adding a
positive offset to the composite video signal during the horizontal
retrace time.  However, depending where the DC reference is, it could also
be interpreted as a large attenuation during the horizontal retrace time.
[This would be the case if the white level in your "oscillographs" (most
positive excursions) is at zero volts.]

The latter scheme is one that was once used by a TV station in
the NY metro area.  This scheme involved attenuating the composite signal
during the retrace time so that the horizontal sync pulse amplitude was
well below the peak of the video information (e.g. the black level).  A 15750 Hz
tone was sent in the audio channel as the new horizontal timing reference.
They also placed the program audio on a 31500 Hz subcarrier (DSSC) so that
it couldn't heard without the decoder.

Schemes for decoding the signal are straightforward.  The 15750 Hz tone is
recovered from the audio output -  a PLL is handy.  For the above scrambling
scheme, a multiplying PLL (VCO running at 31500 and then divided by two) is
employed:  this provides a 31500 Hz tone with which the program audio can
recovered, and a 15750 Hz square wave to control an attenuator/gain block
of some sort.  Since your audio program is already available, you could run
the PLL VCO at 15750 (this assumes that the 15750 is available in your
system).  The recovered 15750 Hz square wave would be used to trigger a
one shot, which in turn would trigger a second one shot.  The two one
shots are adjusted to generate a pulse (pulse position and pulse width) which
overlaps the attenuation interval. The output of the second
one shot would then be used to drive a variable attenuator/modulator through
which the scrambled video is sent.  Viola!  The origonal composite signal
is recovered.

I've seen descramber schematics using PIN diodes as variable attenuators.
I believe you can also use a CMOS bilateral switch (aka 4066) to switch
a simple resistive attenuator.   A balanced mixer might also work.

First things first though - is there a 15750 Hz tone on your audio out?
You may have to look right after the FM demodulator in case the output
is low-pass filtered.  If you have an FM multiplex output on your
receiver the 15750 Hz tone would be available there.

Of course, this is for your *own* viewing pleasure, right?(!)


Eric Beck                        |                #include <std_disclaimer>
fiesta@homxb.att.com             |

torkil@psivax.UUCP (Torkil Hammer) (06/24/89)

In article <111777@sun.Eng.Sun.COM> rom@xor.Sun.COM (Achyutram Bhamidipaty) writes:
#
#Follow up to request for help with building a descrambler.
#
..
#
#Some other notes on the situation. I plan on using the monitor output of my TV,
#this way I dont have to deal with demodulating the signal from the cable 
#company. So the final set up, if it works will look like this:
#
#--------|       |-----------|       |--------|
#  TV1	|-------|descrambler|-------| TV2    |
#--------|       |-----------|       |--------|
#
#-Rom
#rom@xor.sun.com

Tell that to the judge, as the saying goes.

Just because it is there and you are technically capable of stealing it
does not mean that you should try to.

Neither does it mean that you can get away with it.  Cable companies
are aware of the situation and have been known to collect against
entrepreneurs.

If you really want the 'engineering challenge' get a job
with a cable company.

I don't think that it is a good idea to use the Net to proliferate
information on how to transgress the law.  Comments?

torkil

irwin@m.cs.uiuc.edu (06/27/89)

The point in your illustration where you show the X and say there is
a sine wave there, is known as the color burst. It is a few cycles
of 3.58 Mhz to sync the color. If this is missing from the signal,
color sets think it is a black and white picture and will not trigger
on the color circuits. The color "killer" circuit is triggered by
the presence of this burst.

I have seen black and white pictures, where some video information
in the picture will "fake this" causing the color to trigger on. The
stripes on a referee's shirt at a football game taped in black and
white and then shown on a color set, has been known to trigger the
color on.