RF Column 23 September 93
Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED

TOPICS:  Predict whether your signal meets new cable TV requirements
         Calculate signal levels with typical antennas at VHF & UHF
         Tips and techniques for improving reception
         I visit G.I. and report on their Digicipher compression system

--------------------------------------------------------------

This month I'm taking a suggestion from Bill Hayes at WSAZ TV. A while
back he sent me a message on CompuServe asking about methods for
predicting whether or not a TV station's signal would meet the
requirements of the Report and Order to FCC Docket 92- 259 on Mandatory
Carriage/Retransmission Consent". It is the implementation of the Cable
Television Consumer Protection and Competition Act of 1992. He thought
it would make a good column, so I dug into the books to come up with
some details on exactly what the Report and Order requires and what a TV
station can do to meet the signal level requirements. On a different
subject, I'll give you a quick update on G.I.'s Digicipher video
compression system. I had a chance to play with it some more and did
some multi-generation (compress/decompress) tests. I was surprised at
the results.

I'm not going to cover all the legal issues of the Cable Act here, just
the ones that involve technical issues relating to a station's right to
must carry. The Report and Order excludes TV stations that do "...not
deliver to the principal headend of a cable system either a signal level
of -45 dBm for UHF signals or -49 dBm for VHF signals at the input
terminals of the signal processing equipment, unless such station agrees
to be responsible for the costs of delivering to the cable system a
signal of good quality or a baseband video signal."

This rule leaves a lot of room for interpretation. Cable companies have
to designate a headend and put that information in their public
inspection files. The signal levels are measured at that headend. An
unscrupulous cable operator could put his headend under ground or behind
a mountain range or hill to avoid incurring any must carry
responsibilities. They could also do something like use RG-174 miniature
coax for the antenna cable or insert a couple attenuator pads in the
line. Such moves are probably in violation of the FCC's "Cable Technical
Report and Order" which requires cable companies to use good engineering
practices and make a reasonable effort to avoid unnecessary degradation
of a TV station's signal. Besides, the cable operator has to tell the TV
station how they got the measurement.

If your station received a notice from a cable company that it didn't
meet the required signal level, the FCC's Report and Order on the Cable
Act requires them to provide you information on the measurements. The
Report and Order states that in addition to an equipment list, "...the
cable operator must include a detailed description of the reception and
over-the-air signal processing equipment used, including sketches and a
description of the methodology used by the cable operator for processing
the signal at issue." It also states: "This information must include the
specific make and model numbers and age of all equipment." One thing
that has surprised me is the number of cable companies using Radio Shack
U-75 antennas for UHF reception. I suppose for a wide band antenna it
isn't a bad choice, since the same antenna often shows up in over the
air HDTV tests.

The Report and Order requires cable companies to give broadcasters the
opportunity to improve the signal at the headend. They can do this via
microwave, translators (if they would normally be entitled to carriage)
or by improving the cable companies equipment. I heard that Scala, one
of main manufacturers of high quality receiving antennas, has been
deluged with orders from broadcasters ordering antennas for cable
companies. If you don't have their catalog, give them a call at
503-779-6500. There a number of other companies manufacturing high
quality receiving antennas, SITCO (Simplicity Tool Company) is another
one. Their number is 503-253-2000. They have a UHF helix antenna that
may of interest to broadcasters using circular polarization. Early
versions were not that rugged, so you might want to check on the current
model if you intend to use it in a rough environment.

A simple way to increase the signal level at the processing equipment is
to add a preamplifier to the antenna system. Several cable equipment
companies make preamps for VHF and UHF reception. A less expensive
option, with a very low noise figure, are the GAsFET preamps
manufactured by Advanced Receiver Research in Burlington Connecticut
(Phone 203-582-9409). If you're a ham, you've probably seen their ads in
QST magazine. They will custom built preamps for any VHF or UHF TV
channel. These preamps are not waterproof, so you will have to put them
in a housing like a "Tupperware" container with a drain hole for
condensation.

Just what does that -45 dBm or -49 dBm mean in the Report and Order
mean? It shouldn't be that difficult, after all, most well designed
antenna systems deliver 0 dBmV! There's the rub - the FCC specifies a
_power_ level, not a voltage level! Will a 0 dBmV signal meet the FCC's
requirements? Lets see. "dBmV" means decibels above one millivolt. Zero
dBmV will be one millivolt. Using Ohm's law, dividing the square of the
voltage (0.001 volts) by the standard cable system impedance of 75 ohms,
the power is 0.00000001333 watts. Pretty small, eh? To express that in
dBm, multiply by 1000 to convert to milliwatts then take the log of the
result and multiply by ten. You should get - 48.75 dBm. Our 0 dBmV
voltage will not meet the Report and Order's signal requirement for UHF
and barely makes it at VHF! If the signal is otherwise clean and free of
interference, adding of one of the preamp's I mentioned earlier should
satisfy the cable company and the FCC. There's no point in both of us
doing the math, so here's the minimum voltage levels in dBmV and mV
required by the new cable rules. Zero dBmV or one millivolt is a good
level for VHF stations (the FCC probably had that in mind, rounding the
-48.75 down to -49 dBm). UHF has to be 4 dB stronger, on a power basis.
This will require a +4 dBmV signal (rounding again) or 1.54 millivolts.

The next question is, if a cable company is within my A grade contour
(74 dBu for UHF, 71 dBu for high band VHF and 68 dBu for low band VHF)
will cable companies be able to get an adequate signal? There's no
simple answer for this one. As I pointed out in my series on
antenna/signal performance testing last year, the FCC contours give
field strength in microvolts _per_meter_. That "per meter" makes the
conversion of field strength to power delivered to an antenna dependent
on the wavelength, hence frequency of the signal. Here is how to convert
the two, taken from my earlier column on signal strength measurements:


I found the following formula in the NAB "Engineering Handbook" for
calculating the voltage induced in a matched load by a resonant dipole.
I've modified it slightly for easier use:

                   Field x (Square Root of Resistance)
   Voltage =       ___________________________________
                       Frequency (MHz.) x 0.179

Field must be in defined in meters. If the field is specified in
microvolts/meter the result will be in microvolts, likewise, if the
field is in millivolts/meter, the result will be in millivolts. The
resistance is the resistance of the matched load connected to the
receive antenna, usually 75 ohms. For quick reference, the square root
of 75 is 8.66 and the square root of 50 is 7.07. From looking at this
formula you will see that for the same field, a higher frequency will
give a lower induced voltage. This is because the area of the antenna is
smaller. Use algebra to rewrite the formula to find the field. Don't
forget to convert relative decibel readings to absolute voltage and
field numbers.


Getting back to the current problem, to make it easier to see if our
grade A contours will provide an adequate signal at cable headend lets
first convert the decibels above a microvolt per meter into millivolts
per meter. 74 dBu/m becomes 5.011 mV/meter, 71 dBu/m becomes 3.548 mV/m
and 68 dBu/m is 2.512 mV/m. Using the equation above, which is for a
dipole antenna, at its A grade contour a channel 3 station (61.25 MHz.
at visual carrier) will induce the following voltage in a dipole antenna
with a 75 ohm impedance:

                 (2.512 x 8.66) / (61.25 x 0.179) = 1.984 mV

This is above the 1 millivolt requirement we calculated earlier, so even
with a dipole antenna a cable company within the Channel 3 station's A
grade contour should get an adequate signal.

What about at UHF? Let's use channel 48 as an example since it is in the
middle of the band with a visual carrier at 674.25 MHz. Will a dipole
work here too? The A grade contour for UHF occurs at a field strength of
5.011 mV/meter. Substituting the frequency and field strength:

                (5.011 x 8.66) / (674.25 x 0.179) = 0.360 mV

This doesn't make it! To get to the required 1.54 mV level, an
additional voltage gain of 4.283 is required, or in dB, 20 x log(4.283)
= 12.6 dB. I don't have gain figures on the Radio Shack U-75 UHF
antennas, but since they are primarily a corner reflector (the untuned
little directors in the front don't add much gain), I'd estimate the
gain to be around 10 dB in the middle of the band. Adding a dB or two of
cable loss, the channel 48 station will not meet the minimum signal
level requirements even with a standard antenna that would be consider
fine for home use. Recommendations? Use a Scala Paraflector, which gives
up to 20 dB of gain at UHF and add a good low noise preamplifier before
the coax run with an additional 10 dB of gain.

Bill Hayes suggested it would be a good idea to have a program that
would calculate the signal level at a headend at any distance and permit
entry of the numbers for antenna gain and cable loss. Unfortunately, I
don't know of any free programs for calculating field strength using the
FCC 50/50 curves. I've seen several programs, but none permitted free
distribution. If anyone has such a program, let me know. If you have
access to such a program or want to do it the expensive way using an
on-line service like Dataworld or Broadcast Data Services you can plug
the numbers into the equations above. I tried some test calculation
using the formulas for free space attenuation and found significant
discrepancies between those calculations and the FCC numbers. If you
want to try some calculations, look up my column from 1991 on aiming
microwave antennas - it gives all the formulas. (You do save these
columns, don't you?). If I can find the time I'll try to put together a
little program that will calculate the received power for a given field
strength at a specific channel. If someone has an easy routine I can add
in to do the FCC field strength calculations, I'll include that in the
package as well.

Video Compression...

I've gotten a lot of response via phone and CompuServe on my remarks on
video compression in the last column. I mentioned the CableLab's tape as
one of the best for checking compression quality. I do not have a source
for the tape and I'm not sure if anyone is distributing it. I did some
compression tests at General Instruments (G.I.) in June using a tape our
network (Telemundo) had produced as a sales promo. It included snippets
of much of our programming and had a number of segments I would expect
to challenge a video compression system. One segment was a noisy,
standards converted soccer game with a lot of motion, another included
scenes with explosions from a promo for the "Back to the Future II"
movie. Music videos and news clips had lots of fast cuts and there were
high resolution shaded computer graphics with motion in front of them.
Before anyone asks, I cannot make copies of this tape, as it contains
copyrighted material.

At General Instruments I put the tape (D-2 format) through their
Digicipher compression equipment. We operated it in the four channel
mode, with a mixture of programming on the other channels. The noisy
scenes in the original tape (the soccer game and some foreign news
footage) made it through with no breakup or noticeable deterioration. No
problems were noted on the other segments either. To see how well the
system would work in a situation where material might be compressed,
decompressed then sent through the cycle again, the video was looped
back through the compression gear again for a second pass. Again, no
noticeable problems. Finally, when G.I.'s engineer was viewing the two
pass recording, we looped it back through the system a third time. I
thought I saw a slight bit of deterioration, perhaps in shading and
noise, but I was never able to pin point it. I wish I had one of the new
Hewlett Packard compression analyzers!

What surprised me was that the noisy video segments did not fall apart
under repeated compression/decompression cycles, nor was their any
evidence of picture blocking or breakup. I didn't have time to try the
test with the slower 6 channel data rate. The results might have been
different. As it was, the G.I. engineer worked through his lunch hour to
do the tests I wanted. I have not had a chance to try the tape on an
MPEG system. I was impressed with the improvements G.I. had made in
their system (software) since my last visit. The slight amount of chroma
bleeding I had noticed before was gone. I was also impressed with the
job G.I. had done with actual program video. One of the engineers
commented that was one of the reasons they invited people to the plant
to test the system - when they found a problem, they could work on it.
MPEG-2 has a lot of support and a lot of companies working on it. Many
are simply designing hardware around the C- Cube chip set. It will be
interesting to see how long it takes MPEG-2 to reach the same level of
maturity that Digicipher has. I expect it will happen, if only because
of the engineering talent being focused on MPEG. The question is when.

I had some tips on satellite dish alignment and an update on my cheap
remote control, but I see I'm out of space! Look for them next month.
Meanwhile, I welcome your comments, ideas and tips. Send them to me via
CompuServe at my PPN of 70255,460 or give me a call at 305-884-9664 or
818-502------(NO LONGER VALID). My fax number is 305-884-9661 if you
have a drawing, tip or product you think I might be interested in
reporting on. I do have a regular job, so don't be disappointed if it
takes a couple days to get back to you. You can also write my at 2265
Westwood Blvd., Suite 553, Los Angeles, CA 90064. Mail is forwarded to
me from that address when I'm traveling.
((8/95 > UPDATE! - Use dlung@gate.net for e-mail!))

Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED