[opendtv] Eye witness account of 1999 Sinclair Baltimore tests
- From: "John Shutt" <shuttj@xxxxxxxxx>
- To: "OpenDTV" <opendtv@xxxxxxxxxxxxx>
- Date: Wed, 21 Apr 2004 21:30:17 -0400
http://web-star.com/hdtv/pbsbalit.html
COFDM AT A DISTANCE
by
Jan Pazral,
Chief Engineer
WXXI Public Broadcasting Council
August 11, 1999
Jan Pazral is the chief engineer at a public television station, and a
member of the engineering advisory group to APTS (America's Public
Television Stations). Since those PBS engineers who first observed the
Baltimore tests were unable to observe distant reception tests, we
encouraged Jan to try to see this in his subsequent trip. His report is
notable in his observations of two remote sites. His report is posted with
permission.
Regards,
Bruce Jacobs
KTCA, St. Paul
----------------------------------------------------------------------------
----
As most of my colleagues in the broadcast industry, I took 8-VSB as given.
This issue has been decided long time ago and explained to us on various DTV
seminars, workshops and conferences. The term Longley-Rice has become as
familiar to us as NTSC or Subcarrier. We were told that DTV, while not as
easily receivable as NTSC, will replicate the analog coverage by simply
using adequate outdoor antennas. Then recently, it came to my attention that
field tests in urban areas, using current generation of consumer DTV
receivers and consumer antennas, revealed serious problems with 8-VSB
reception and, moreover, receivers using COFDM, a modulation method the only
thing I knew about was that it existed, provided solid reception most of the
time. This issue became a topic of discussion at the conference call of the
APTS Engineering Advisory Group in mid-July where testimonials from
engineers who eyewitnessed the tests confirmed that the 8-VSB transmission
could not be received in densely populated areas just a few miles from the
transmitter. I realized that in order to participate in the Advisory Group
discussions on this issue in a meaningful way, I needed to see the tests for
myself. I took the opportunity to travel to Baltimore and observed the field
testing on August 3. The purpose of my trip was not gathering a collection
of scientific data and compiling a list of decibels and bit-rate errors. I
wanted to find out what kind of reception an owner of a consumer DTV
receiver can expect in urban, suburban and rural settings.
The following equipment was used for reception and testing:
ATSC Receivers (8VSB):
Panasonic TU-DST50
Pioneer Elite SH-D500
DVB-T Receivers (COFDM):
Nokia View Master 9600
NDS System 3000
HP Spectrum Analyzer
Antennas:
Bow-tie dipole
Double bow-tie with reflector
13-element Yagi antenna
Winegard pre-amplifier
RF switches and RF attenuators
6-foot antenna mast
(On that particular day, Geocast Corp. was testing an 8-VSB prototype
receiver board made by Oren Semiconductor. This receiver did not have a
video output; it was interfaced to a laptop computer instead.)
The following equipment was used for transmission:
Comark IOT transmitter
Two-panel antenna, 1200 ft HAAT, 50 kW ERP
Zenith 8-VSB modulator (19.39 Mbps)
Rohde & Schwartz SFQ COFDM modulator (3/4 FEC, 1/8 Guard, 1705 total
carriers, 18.662 Mbps)
The testing started at 9:45 EDT at the DTV Channel 40 / NTSC Channel 45
transmitter site near 2000 West 41st Street in Baltimore. The performance of
all receivers was verified by a direct connection to a padded tap on the
transmission line. I will not go into a great detail about the data
collected here. I just mention that the RF threshold for the Pioneer
receiver was -82dBm, for all others about -78 dBm. The reference power for
the spectrum analyzer was -22 dBm for both 8VSB and COFDM. Subjectively, all
receivers worked well.
The location of the first test site was in Downtown Baltimore, Lombard at
Greene Street. While the spectrum analyzer showed strong and reasonably flat
RF signal using any of the three antennas (we tried aiming them in different
directions), none of the two 8-VSB receivers produced any video. The OREN
receiver did not establish a lock either. Having the bow-tie antenna on a
stick in my hand, I was unable to find a spot within a reach of the antenna
cable where 8-VSB would work. The COFDM reception was a different story.
While the RF signal on the spectrum analyzer looked virtually identical,
receivers locked instantly. Again, walking around with a bow-tie dipole
antenna in my hand, it was hard to find a spot where the reception would not
work. Even then, the disturbance was limited to an occasional glitch, rather
than a complete loss of signal.
The second site was on the top level of the parking garage of BWI airport.
The transmitting antenna was clearly visible. All receivers worked
flawlessly using a bow-tie dipole.
The third site, an open parking space randomly found at the bottom level of
the same parking garage, was a tough spot to receive UHF. Analog reception
of channel 45 was ghosty and noisy. There was no 8-VSB reception using any
of the three test antennas. There was, however, solid, glitch-free COFDM
reception using the double bow-tie antenna. The difference in quality
between the COFDM and analog reception was so dramatic that if I was a
viewer having the same conditions, I would gladly pay for a COFDM receiver
to achieve that level of picture quality.
The last two sites were distant sites, picked on-the-fly. In Waldorf, MD,
48.8 miles from the transmitter site (distance measured be a GPS
instrument), we were able to receive both COFDM and 8-VSB using 13-element
YAGI. There were several interesting things about this site: Adjacent DTV
channel 39 out of Washington D.C. was about 15 dB stronger than channel 40.
The site was next to a busy road and antennas were aimed across the road.
There were trees in the direction of the transmitter about 200 to 300 ft
away and there were airplanes crossing the RF signal path. My observation
was that both 8-VSB and COFDM signals were receivable at this location. The
aim and location of the antennas for 8-VSB reception was much more critical
than that for COFDM. The road traffic did not seem to have any effect on
either 8-VSB of COFDM, but the airplanes flying overhead produced
considerably more disturbance to 8-VSB than to COFDM.
The fifth and last test site was Mechanicsville, MD, 62.2 miles from the
transmitter, outside of the grade B contour. I had the opportunity to select
the test point myself. This test site was located about 200 ft from a busy
divided highway, about 150 ft from a parking area with two idling diesel
trucks and about 250 ft from a gas station. Using the single bow-tie antenna
with a pre-amp did not provide a satisfactory reception on any available
receiver. Subjectively, there were slightly more glitches using COFDM than
using 8-VSB. The 13-element YAGI and the double bow-tie with reflector,
however, produced solid, glitch-free reception for both COFDM and 8-VSB.
(The aiming of the Yagi was not critical reception was maintained within +/-
40 degrees for both systems.)
Based upon my observations I came to the following conclusion: The Baltimore
tests are believable. I was favorably impressed with the methodology and the
documentation of those tests. In the fringe areas, 8-VSB and COFDM
performance was virtually equal. In the majority of urban test sites, COFDM
worked well, while 8-VSB did not work at all.
We can expect spotty 8-VSB reception in other urban areas elsewhere in the
country.
While there may be advancements in the receiver technology, 8-VSB (based on
single carrier and high data rates) in multipath environment will pose
greater challenge to receiver manufacturers than COFDM (based on multiple
carriers and low data rate).
Jan Pazral,
Chief Engineer
WXXI Public Broadcasting Council
----------------------------------------------------------------------
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:
