RF Column 43 - June 1995
Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED

TOPICS:   Editorial on ADTV - Keep it flexible
          Comments and summary of ADTV sessions at NAB 95
          Transmission equipment at NAB 95
              Comark's Innovative PS-Squared power supply kills arcs
              Acrodyne & Thomson show Diacrode - 60 KW tube the size of
                   a lawn mower tire!
              Higher power from solid state transmitters
          Update on Internet providers
          Interesting Web Sites for RF Column readers

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

As you might have suspected from reading my April column, I've been real
busy at my real job and didn't have time to do a May column. After a
month's delay and what I believe was the most significant NAB of the 20
I've attended, I have enough material to fill several columns. This
month I'll start my NAB coverage off with some personal comments on ADTV
developments at NAB and follow with reports from some of the technical
sessions and observations on UHF transmitter developments at NAB. I also
have some updates on Internet services and resources.

I've been saying since last year that using the digital bandwidth of the
broadcasters' second channel for big screen pictures only didn't make
sense. Apparently other people felt the same way and at Sunday's NAB
Technical Sessions focusing on ADTV the Grand Alliance revealed their
proposal for "flexible use". My personal opinion is that if the F.C.C.
and Congress allow it, this will dramatically shorten the time it takes
for digital television to become established. The reason I say this is
that for many broadcasters, the idea of investing a million dollars or
more in a new transmission plant for ADTV didn't make sense if all it
would do is provide perfect big screen pictures to a limited number of
sets. If they couldn't sell it, it would be tough to sit on that kind of
an investment for the five years or more it would take for the consumer
to catch up.

"Flexible Use" changes all that. First, as demonstrated by the G.A.,
additional multimedia material can be imbedded in the picture stream
with commercials. This could be a catalog, an interactive computer
program which would help guide the customer through a purchasing
decision or some useful reference material that carried the advertiser's
logo. This extra value is something stations should be able to sell.
Look at how quickly companies are jumping on the Internet with World
Wide Web home pages. Second, it opens up additional "broadcasting"
opportunities. When I started in radio a couple decades ago the
main-stay of Sunday morning programming was local church services. Now
that camcorders are cheap, it should be fairly easy for a church to tape
a service, supply it to the TV station, and have it transmitted at low
cost using fairly heavy compression along with other programming. Third,
because "Flexible Use" is not limited to video, it would be excellent
public relations to allow community organizations and governments to put
up Web like hyper-text, multimedia documents during times when
opportunistic data was available. Of course, paying advertisers would
have the same options.

I am concerned that Congress, after seeing the flood of money created by
the PCS auctions, will try to charge broadcasters for that second
channel. Flexible Use opens the door to such an effort. My feeling is
that as long as the material transmitted can be received by the public
at no charge without restriction, just as broadcast TV programs are
today, there should be no additional charge or tax on the broadcaster.
We are only moving our existing broadcasting structure into the
information age, one where several sources, all of which I've seen want
to charge you for reception, are competing. The digital second channel
with the flexibility to use it to deliver free programming to the
public, be it HDTV, SDTV (Standard Definition TV), digital multimedia or
computer data is essential for broadcasters survival over the next ten
years.

My recommendation -- charge the broadcaster the "market rate" (whatever
that is) for any bytes transmitted that are not receivable by the
general public (with appropriate equipment) without additional charge.
The rate should match what other spectrum users pay. That would be fair
to broadcasters. A smaller station without the capital to build a
digital transmission site could work a deal with a data seller to supply
the capital. If his taxes were comparable, he should be able to compete
with other vendors for the bandwidth. Add a condition that limited how
much of the data bandwidth could be used for these for-fee services and
the public gains as well by getting a new, free, digital TV service.

If the regulators can stay out of the way, we could see regular ADTV
transmissions start by 1997. At Sunday morning's opening session,
Richard Wiley, Chairperson for the FCC's ACATS, said that the over the
air field tests in Charlotte NC were on schedule for this summer and
that he expected an FCC proceeding to establish a new video standard by
the end of the year.

Wayne Luplow, from Zenith (a Grand Alliance member), remarked that the
field tests done last year showed that a noise limited 8-VSB signal
would give equivalent coverage to a conventional Grade B contour, even
after allowing a 12 dB peak NTSC to average ADTV ratio. I won't go into
details here because most of the information is the same as I reported
in my column on the World Media Expo / SBE / SMPTE convention in Los
Angeles last October. Some of the improvements that have been made since
then include building in greater interference rejection and speeding the
response to multipath by use of data-directed equalization. Speaking of
multipath, Mr. Luplow said they did try using an indoor bow-tie antenna
during the tests. As anyone who has tried to optimize an indoor UHF
antenna for good reception knows, it is very difficult to get a
multipath free signal when people are moving around in the room. The
Grand Alliance 8-VSB, according to Mr. Luplow, performed well with the
indoor antenna and experience no data errors from the time varying
multipath. In addition, he noted the 8-VSB system outperformed NTSC in
the presence of burst and impulse noise.

There was quite a bit of interesting discussion about COFDM (Coded
Orthogonal Frequency Division Modulation), which uses multiple carriers
to allow each carrier to transmit the data at a slower rate, resulting
in a more rugged system in the presence of on-channel interference and
multipath. I also covered this in previous columns, but some practical
test information emerged at NAB. Vidar Ringset from SinTef Delab in
Norway Yiyan Wu from the Canadian Communications Research Center
discussed their testing of COFDM for terrestrial ADTV transmission.
SinTef Delab built the modems used in the tests. I won't go into the
technical specs, but these were very flexible units, capable of
generating up to 13,000 carriers using inverse Fast Fourier transforms.
In addition, because the carriers are computed digitally, it is easy to
design an almost rectangular spectrum shape and build in guard intervals
to allow for local "gap filler" transmitters.

Yiyan Wu pointed out some of the disadvantages of COFDM versus 8-VSB.
First, higher peak powers are required, even though the average power
remains the same. He estimated an additional 2 dB would be required,
though more studies need to be performed. The system is also more
sensitive to frequency errors. COFDM can have a long guard interval to
allow on channel repeaters, but only if the bandwidth is 8 MHz., a bit
large for the United States and Canada. However, because there is no
demand in the U.S. for a nationwide single channel network, a smaller
guard band may work. Some of the hurdles to implementing "gap filler"
transmitters are the requirement they be locked to less than one Hertz
of each other and that they be adjustable in delay. Phase noise could be
a problem with COFDM if the phase noise on each carrier is high, but the
total phase noise can be the same as for a single carrier system like
8-VSB. I was disappointed to hear that the modems were not designed to
work in a mobile environment. Some of the proponents of COFDM have been
touting this as one of its advantages. In a "flexible use" environment
it could be essential in some applications.

Its not too late to switch from 8-VSB to COFDM, but at the rate the
Grand Alliance and the A.T.S.C. are moving to get proposals to the
F.C.C. I think it will be an up hill battle to prove the system worthy
before 8-VSB becomes the standard. I hope we pick the right system!

The last two years have had very little ADTV gear on the floor. This
year all the major transmitter manufacturers demonstrating how the Grand
Alliance 8-VSB signal made it through their exciter and or transmitter.
Not surprisingly, the only two transmitters I saw operating through the
final amplifier stage were solid state low power TV transmitters from
Acrodyne and I.T.S. I'm still waiting for the NAB when Nat Ostroff fires
up a 60 KW Comark transmitter on the floor of the convention center to
demonstrate a working FULL power ADTV transmission system!

That may not be as impossible as it seems. In one of the most
significant advancements in I.O.T. (or Klystrode) transmitter design at
this year's NAB Comark showed an alternative to the big, heavy,
expensive oil filled high voltage supply. It also solved a major
weakness in I.O.T. systems. One of the big problems with I.O.T.'s (note
that when I say I.O.T. I am not excluding Klystrodes) is been their
tendency, like any tube operating at high voltage and high power, to arc
now and then. The small vacuum space inside the I.O.T. makes it
vulnerable to the gas produced from arcs and the grid structure is prone
to damage from arcs. To protect against this, tube manufacturers require
the power supply shut down immediately to protect the tube. This has
been discussed in previous NAB papers and in tech bulletins from tube
and transmitter manufacturers. Usually the protection comes from a
crowbar circuit which shorts the HV supply to ground, tripping circuit
breakers and causing all kinds of transients in ground and power
circuits.

Comark's PS-Squared power supply borrows from the Thomcast high power AM
transmitter division's experience with high voltage AM modulators. The
supply consists of eight 4 KV supplies in series, each controlled by a
fiber link. The supply requires no oil filled transformers and consists
of rack mounted units that can be placed next to the transmitter if
desired. Because the supplies are regulated, no external AC regulator is
needed at most sites. Because the supplies, unlike a transformer supply
with a huge capacitor, stores very little energy and can supply very
little energy in excess of the rated output, crow bar circuits are not
needed. And finally, because the supplies can ramp the voltage up and
down and shut off quickly, they shouldn't trip the AC circuit breaker,
so an automatically resetting breaker isn't required.

I was impressed with the PS-Squared technology. It's simple, which means
it should work as advertised. Early adopters of Klystrode / I.O.T.
technology often told me they didn't want to have to deal with the
pulsers required to obtain reasonable efficiency from klystron tubes. I
countered that the I.O.T. itself, with the requirement for a crow bar,
auto-resetting breakers and such, required technology as difficult if
not more so than pulsing. In addition, failure of the crow bar circuit
would almost certainly destroy the tube. Failure of the pulser generally
didn't. Comark finally met these concerns with the PS-Squared. Harris
uses similar technology in its line of high power AM transmitters, so
I'd expect something similar from them, perhaps by next year's NAB.
Kudos to Nat Ostroff and Comark for coming up with a design that at
first glance not only improves reliability, but should cost less, be
easier to install and easier to maintain. This year the I.O.T. faced its
first serious competitor for high powered UHF transmitters since the
MSDC klystron. That competition comes from the Thomson Diacrode.

I wrote about the Diacrode in my report from World Media Expo in Los
Angeles last year. At NAB Acrodyne was showing a compact transmitter
capable of producing 60 KW peak visual power with a combined 6 KW aural
carrier using one tube, the Diacrode. I'll be visiting Acrodyne to see
how their first production transmitter performs. If TV Technology bends
a little in my deadlines and Acrodyne meets theirs I may able to include
the results in next month's column. Note that the Diacrode's
specifications say it should produce more power -- 60 KW combined or 80
KW visual only peak sync power, than a single I.O.T. can produce.

My biggest concern is how well the Diacrode's cavity system will be able
to carry away the heat produced by a device the size of a small lawn
mower tire operating at these power levels. The engineers I spoke to at
the Thomson booth assured me that the cavity could handle the power.
When I asked about the cavity temperature, I was told is should not run
above 40 degrees C in normal service. The cavity is water cooled, so a
pure, distilled water heat exchanger is required. I'm concerned about
high cavity temperatures because most of the failures I see in cavities
are somehow related to heat. Heat can speed corrosion and cause
mechanical stress on components when the tube is turned on and off. Both
effects tend to cause losses to increase, which causes more heat, which
causes more problems. Because these effects accumulate over time, one
thing I'll be looking for when I inspect the operating transmitter at
Acrodyne will be how hot the cavity is running. If it's relatively cool,
reliability should be good. If it runs hotter than other proven cavity
designs running at less power, I'll be concerned.

Solid state UHF transmitters seem to be making some head-way. I saw
transmitters from Acrodyne, Larcan-TTC and Harris capable of operating
in the over 10 KW power range. I heard the folks from Westinghouse, who
are developing high power amplifiers using silicon carbide transistor
technology, were looking for me at the show. Unfortunately, in spite of
trying to track them down Thursday at the Hilton, I never got to meet
with them. What I heard was they were showing an amplifier brick the
size of a bread box which could produce 1,500 watts of power. That is
almost an order of magnitude above what single conventionally designed
modules can produce. There is a lot of work going on with solid state
high power UHF amplifiers. I installed a new Acrodyne TLU-1/KE
transmitter using its new amplifiers. Two amplifier drawers produced the
same power (1 KW combined) as four of their older modules produced.
Performance was excellent. More about that next month. Other
manufacturers are looking at new devices for their amplifiers. I talked
to Acrodyne, I.T.S. and T.T.C. about their solid state amplifiers and
all are using different devices! Indeed, it was interesting see how
excited transmitter manufacturers became when talking about their new
devices and how they along were using them.

My Web site picks for this month focus on ADTV. The first URL is
"http://www.atsc.org". (Use only the letters within the "`s and do not
include the "`s). This site has copies of the Grand Alliance's HDTV
proposal available as zipped Word 6.0 files and has links to homepages
of those Grand Alliance members with homepages. It's interesting to see
which companies have participated in the Web and which haven't. I was
surprised to see Zenith didn't have a link. If you are using a Mac and
can't read Word 6.0, check out Charles Poynton's homepage. He has
converted the G.A. proposal into Adobe Acrobat format. His URL is more
complicated. Use
"http://www.inforamp.net/~poynton/Poynton-video-eng.html". Be sure to
type the capitalization as listed, even though in one case Poynton is
capitalized and in another is not. Poynton also has links to World Wide
Web servers with copies of CCIR standards and much other useful
information. NAB is present on the web, use "http://www.nab.org" to
check it out.

And finally, a few notes on Internet providers. Shortly after I
submitted my last column it appears PING (Prometheus Information Group)
went out of business. I had recommended them for low cost, high speed
800 number Internet access without a monthly fee. Apparently they were
too low cost to stay in business. Netcom's "Netcruiser" account has
improved significantly since the last column. Their new software
supports third party WINSOCK compliant software. I've used it with
Netscape, Free Agent and several Netmanage Chameleon software products
with no problems. Netcom has also improved the speed of transfers across
the network, including 28.8 KB access in many cities. It is, however,
still prone to down time and overloads, particularly at busy times like
Saturday night. And, for what ever reason, I have not been able to
connect through the Dallas number (it never answers) during the last
three visits through DFW airport. My recommendation to have two Internet
providers stands. One of them might be your older on-line service, now
that they are adding Internet World Wide Web browsers.

If you are on Compuserve you now have World Wide Web capability using
the latest version of WinCim and their "NetLauncher" software. I'm late
submitting this article because I wanted to see how it worked. The
package includes Spry Mosaic, a well known Web browser. You can,
however, use any WINSOCK compliant software with it. I tried it out the
evening of May 1 and found that while the system responded nicely to the
Compuserve Web pages, as soon as I tried to connect with one of my
favorite weather sites it slowed down to the point I gave up on the
transfer before it finished. The same site, when accessed using Netcom,
responded at top speed with no delays. Netcom's "Netcruiser" Internet
access had the same network bottleneck problems a couple months ago, but
not nearly as bad as I saw on Compuserve. At this point I'd recommend
Compuserve for Internet connections only if you use it in the middle of
the night or are very patient. They do have a special plan that gives 30
hours of Internet access for a monthly fee of approximately $25. This
includes access to Compuserve basic services, so it isn't a bad deal. On
the other hand, Netcom's under $20 for 40 hour prime, unlimited
non-prime access is tough to beat.

As I was writing the article, I took a break to see how America On Line
(AOL) was coming along with their Web software. They were offering a
preview version of the new software for downloading. The information
page indicated you could visit only those sites they had selected for
the preview. I found that it was possible to visit other sites merely by
typing in the proper URL. When I tried my favorite weather server, it
came up quickly without hesitation. AOL's browser didn't show graphics
unless you clicked on them, which makes comparison with other browsers
that do show graphics unfair. To AOL's credit, I was able to quickly
download a satellite map and surface analysis with no problem. On the
speed front, they seem to be winning the battle over Compuserve. On the
other hand, the Compuserve software allows use of third party software
and is much more flexible, even if it is somewhat more difficult to get
up and running.

One final Web site worth checking out. Try URL
"http://bongo.cc.utexas.edu/neuroses/cwsapps.html" for a review and
listing of "Consumate Winsock Applications". You'll find links to
Netscape, Eudora, Free Agent and many other useful Winsock applications.
Netscape, of course, is the standard reference for Web browsers. Free
Agent is available as a beta-test. It is a program for downloading and
reading Internet news groups. If you are used to using an off-line
reader like Tapcis or Ozcis for Compuserve messages, Free Agent will let
you do much the same thing with Internet news groups. Strongly
recommended for Windows users.

That's it for this month. As usual, I've gone over my allocated space
and still didn't cover all the topics I promised last month. So, let me
say that next month I hope to include a comparison of the Acrodyne and
I.T.S. 1 KW solid state LPTV transmitters along with more NAB items and
some items I promised last month but was not able to fit in this month.
All those that requested copies of the path calculation spreadsheet via
e-mail or who sent me a disk and mailer should have received them by
now. Please let me know if your request fell through the cracks or got
buried on the desk.

The best way to reach me is via e-mail. Send mail to me at "dlung@gate.net".
Compuserve users should use my PPN "70255,460" instead of my Internet
address to avoid getting charged for the Internet mail. I'm hoping to
have a home page and possible FTP link up sometime in June. My plan is
to have links to the sites I mention here as well as copies of some of
my more popular columns, programs and graphs. You can also phone me at
Telemundo, (305) 884-9664 or fax (305) 884-9661. If you call, please do
so after 6 PM as I tend to be very busy before that time. If you must
use conventional mail, I still maintain a mail service at 2265 Westwood
Blvd, Suite 553, Los Angeles, CA 90064. Note that because of my travels
it may take 6 to 8 weeks for conventional mail to catch up to me. Thanks
for all your comments. I appreciate it!



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