RF Column 48 - November 1995
Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED
TOPICS: R.F. items from the 1995 World Media Expo/SBE in New Orleans:
ENG Safety!
Underwriters' Laboratories (UL) listings for transmitters?
Effect of towers on side mounted UHF slot antennas - Harris
8-VSB HDTV signal Under Non-Linear Amplification - Harris
Dealing with IOT Arcs with HV Crowbars - Larcan / TTC
PS*2 HV supply eliminates need for HV crowbars - Comark
60 KW UHF Diacrode has no need for 30 KV supplies - Acrodyne
Acrodyne 60 KW Diacrode transmitter passes 24 hour full load test
RF Web Page changes and updates
--------------------------------------------------------------------
TV Broadcast RF engineers attending this year's 1995 World Media Expo
and Society of Broadcast Engineers (SBE) Engineering Conference in New
Orleans found several sessions devoted to RF topics. I'll cover the
highlights this month and tell you how to obtain copies of the
conference papers. As the deadline for this column approached,
Acrodyne's Au60D single tube 60 KW Diacrode transmitter successfully
completed the 24 hour black picture RF stress test. I feel the Diacrode
and Comark's PS-Squared power supply technology were the two most
significant developments introduced at NAB this year. Both Acrodyne and
Comark presented papers at SBE. Read on for the latest news on both
technologies. I'm planning a new project for my Internet Web page to
coincide with a major project I'm working on with Telemundo. If you have
an interest in UHF transmitters, you might find it interesting.
This year's SBE Engineering conference schedule was similar to those
from past years -- a session focusing on F.C.C. regulations, a session
on transmission and sessions on new technology. The final session this
year was a bit different. It focused on an area that deserves more
attention - safety!
Any speaker would have difficulty keeping an audience awake and
interested (after three or four days in New Orleans) during SBE's final
session on Saturday afternoon. Mark Bell, a consultant and contributing
writer for Television Broadcast magazine, succeeded. A condensed version
of Mark's talk "ENG Safety and Liability", minus the video inserts, is
included in the SBE Engineering Conference Proceedings. If your station
has ENG trucks, his paper alone may be worth the cost of the
Proceedings. Mark started the presentation with a video of an ENG truck
smoking and sparking as its mast contacted an overhead power line. In
the past two months - July and August - three similar accidents
occurred. ENG truck technicians and talent are being killed and maimed
with horrible regularity in this business.
Mark pointed out state and federal laws make stations and managers
liable for employees' acts. So far, there have been no large lawsuits
from a member of the public injured during these accidents. It could
happen. When the truck attempts to elevate itself to several thousand
volts above ground, what happens to the cables connected to that truck?
Federal laws require equipment used in the vicinity of high voltage
(over 600 volts is considered high voltage) power lines be kept at least
ten feet from those lines. What is the answer? Mark used AFLAC Broadcast
Division's plan to show one approach. The plan is based on mandatory
training and certification of ENG truck operators and is backed by the
support and recognition of the top management of the company. Not only
are the operators held responsible for any violation of safety
procedures, their supervisor and any other levels of management involved
are held responsible as well. If a news director insisted on a dangerous
setup, he or she would be just as responsible as the operator and
subject to the same or greater discipline.
Mark Bell finished the paper with an plea for station engineers and
managers to know the codes, know the laws, set policies and stick to
them. Training is most important. Legal sanctions are hard to deal with.
They pale compared with the personal anguish that would result from the
loss of a friend and employee because training and enforcement weren't
adequate. Don't let it happen.
Terry Baun, from Criterion Broadcast Services, had another interesting
talk Saturday afternoon on "UL Listing and Other Studio Transmitter
Safety Issues". As engineers upgrade their transmitter sites, some have
been surprised to find the local electrical inspector looking for a UL
Listed label on their high powered transmitter. Terry pointed out that
electrical wiring codes are based on the National Electrical Code. The
Code's premise is that wiring should be sized and installed so that
fires don't result. Fire, not safety, is the focus. On the other hand,
Underwriters' Laboratories was founded by manufacturers wanting to prove
their new electric appliances were safe. Underwriters' Laboratories is a
private, non-profit company. There is one UL standard that applies to
broadcast equipment, UL 1419, which mentions professional studio
equipment. Terry pointed out that "wall-warts" those wall mounted power
supplies, are a way manufacturers can avoid UL certification. If the
wall-wart is certified and puts out less than 30 volts, any equipment
connected to it falls below UL's criteria for certification. I've used
this provision of the UL process myself when custom building equipment.
I always use a UL certified power supply and leave it outside the
home-brew unit.
Some inspectors have said the FCC requires UL listing. Terry found only
one case where this applied. FCC Rules Section 73.508 is peculiar to
non-commercial FM stations with 10 watts transmitters. UL listing isn't
mentioned anywhere in conjunction with commercial stations.
Terry Baun stepped away from the UL issue for part of his time to
outline a few safety guidelines we as transmitter engineers should
follow when working around high voltage. Don't work on high voltage
circuit by yourself. Know when you are too tired to work safely. After
fifteen hours with no sleep you can get sloppy and desperate. Terry
related that he never had an argument with station owners or management
when he told them he would have to have a second technician around when
working on high voltage equipment. When a transmitter is off the air,
our first priority becomes getting it back on the air. It seems to be in
every broadcast engineer's genes. If you insist on working by yourself,
make sure someone else knows where you are and keeps track of your
condition (perhaps the guy at the studio). Most of us know someone who
has been injured or killed working on a transmitter. Terry advised us to
share those experiences with new engineers so they don't duplicate them.
Don't ignore safety. Work smart. Work rested. Avoid "freak" accidents.
I've mentioned before that I felt SBE offered a better selection of
papers on TV RF topics than NAB did. This year transmitter manufacturers
reinforced that feeling. Harris/Allied presented three papers and sent
Karl Black to run an Ennes Workshop on "Solid State Power Amplifier
Troubleshooting and Repair". I wish I'd been able to make that workshop.
Acrodyne, Comark and Larcan-TTC also presented papers. Their papers
represented six sessions devoted to TV RF related topics out of a total
of twenty non- FCC/Regulatory SBE sessions.
Because Harris had the most papers, I'll start with them. Two of
Harris-Allied's papers consisted mostly of the results of tests done at
Harris' facilities. When I saw Rex Niekamp's paper titled "The Effect of
Supporting Structures On UHF Slotted Antenna Patterns for NTSC and HDTV
Applications" I thought for a moment I was going to get a replay of Oded
Bendov's (from Dielectric) paper on the same topic at NAB 1995. I had a
full report on Oded's paper in an earlier column. Rex's paper didn't get
into the complicated frequency dependent nulls caused by reflections
from nearby tower members. Instead, his paper focused on the over-all
effect on the antenna pattern as measured at Harris's obstruction free
test range. Rex spent some time describing the antenna test range,
located on a three mile flood plane of the Mississippi River. Most of
the antenna engineer's I've talked to agree this is one of the best
antenna test ranges in the country. It is important to avoid
obstructions within the test range because they will cause reflections
which tend to fill in antenna nulls and distort the pattern. At the
Harris site a small transmitter drives a tower mounted Yagi
(directional) antenna pointed at the antenna under test on a turntable
on a hilltop three miles away.
If you can visualize how direct and reflected waves interact, you can
probably guess the basic results of the tests. Here are some of the
conclusions. Varying the spacing between the antenna and the tower
changes the pattern. In some cases, the radiation in a particular
direction may be greater than that indicated by the free space pattern.
The nulls in the vertical polarization are different than those in the
horizontal polarization. This indicates that elliptical or circular
polarization is one way to minimize the effect of the tower. Perhaps the
most surprising result of the study was that the change in amplitude
versus frequency was only a few percent (1 or 2 percent). This appears
to contradict Oded Bendov's analysis which showed deep, significant
frequency dependent nulls caused by structures near the antenna.
I'll describe one test which showed how important the spacing from the
tower is to the pattern of a side mounted antenna. An omni-directional
antenna, mounted three feet from leg, showed significantly greater
relative field in the direction in line with the center of the tower,
the leg and the antenna than the free space pattern indicated.
Increasing the spacing to four feet reduced the relative field in that
direction, but increased it in directions either side of the center
line. When the spacing was increased to 6.4 feet, the nulls decreased
and the measured field pattern more closed matched that of the free
space pattern, except, of course, in the direction of the tower. You'll
have to purchase the SBE Proceedings or contact Harris Corporation
(+207-222-8200) for a copy of the measured antenna patterns.
Robert Davis' paper "Time and Spectral Characteristics of the Grand
Alliance 8-VSB HDTV signal Under Non-Linear Power Amplification" was
similar to Rex Niekamp's in that the paper focused on tests,
measurements and results and not on specific products, leaving the
listener to reach their own conclusions. This paper is also included in
the SBE Proceedings and, as before, I'll refer you to that publication
for the details. Here are some of the highlights I noted during the
session. Robert started with an ideal amplifier and then added
distortions, showing the amount of power back-off from saturation needed
to keep sidebands outside the 6 MHz. channel 50 dB below the average
power at the center of the spectrum. An ideal amplifier limiting at 0
dBm will generate -50 dB sidebands with an average power 6 dB below the
peak power (6 dB back off). With a 3 dB back-off, the error rate will
always be greater (worse) than one error in a hundred (10E-2). If cubic
non-linearities are present in the amplifier (these lead to third order
intermodulation), the average power must be backed off by 14 dB to
maintain the -50 dB sideband specification. A back-off of over 20 dB may
be needed to keep the sidebands at or below -50 dBm when the amplifier
has a peak ICPM (Incidental Phase Modulation) of 10.5 degrees.
Mr. Davis summarized the results by stating that spectral sidelobes 1)
increase with power amplifier drive levels, 2) cause adjacent channel
interference and 3) depend on the AM/PM (Amplitude / Phase Modulation)
non-linearity in the amplifier. Amplifier non-linearities will have to
be corrected in order to reduce interference and allow the receiver to
handle signals in the presence of noise.
Although the Inductive Output Tube (also referred to as an I.O.T. or
Klystrode) has become the tube of choice for most high power UHF
transmitter applications, the IOT has always had one serious drawback.
The relatively small size of the vacuum, short element spacing and very
high voltage (30 KV or more) make the tube susceptible to arcing. Such
arcing, if not quickly stopped, leads to the catastrophic failure of the
tube. Three of Friday afternoon's sessions focused on this problem from
three completely different directions. It made for one of the liveliest
afternoons at the show.
Gordon Allison from Larcan-TTC described the most traditional way of
dealing with IOT arcs -- a crowbar across the 30 KV line to short the
energy to ground, immediately quenching the arc and protecting the tube.
A typical IOT power supply is capable of delivering a couple amps of
current at 30 KV. EEV, an IOT manufacturer, requires the protection
circuitry stop the arc fast. How fast? A 40 g. (British - SWG) wire must
not melt when placed between the power supply and ground! If a
transmitter's protection can't meet this requirement, the EEV tube
warranty is voided.
Gordon described the characteristics of the three types of crowbar
devices used to protect the IOT: the thyratron, the vacuum gap and the
spark gap. I won't cover the technology behind each of these designs.
Gordon's paper does a good job of it and it is in the SBE Proceedings.
He compared each of the three based on cost (over the life of the
transmitter), size, complexity, warm-up time and several other
parameters. All the devices had to meet certain basic requirements
necessary to protect the tube, otherwise they couldn't even be
considered. Over a twenty year period, Gordon estimated a crowbar would
fire around 1,800 times. The thyratron is fast and reliably protects the
IOT, however, it is expensive --costing around $5,000 plus another
$5,000 for associated circuitry, complex and it takes up quite a bit of
space. It also requires a seven minute warm-up time. Vacuum gaps are
less complex, cost around $5,000, require no maintenance, have no heater
and thus no warm up time. Spark gaps are even less complex, cost less --
around $2,000 and reliably protect the tube from arcs over 22 KV. Under
that voltage, additional circuitry is needed.
While the printed paper lets the reader draw their own conclusions,
during this session Gordon made it clear the spark gap was his device of
choice, combined with a fast acting SCR controller on the primary of the
HV transformer. While lacking the complexity and cost of the other
options, he explained it was capable of providing reliable and cost
effective protection.
Nat Ostoff, President and C.E.O. of Comark, suggested eliminating
crowbars altogether in favor of Comark's new PS-Squared high voltage
power supply. The PS-Squared supplies, based on high voltage modulators
used in super power AM transmitters, use eight 4 to 5 KV switching power
supplies wired in series to supply power to the IOT. The IGBT (Insulated
Gate Bipolar Transistors) used in the supply are inherently current
limited and incapable of damaging the tube. Crowbars are required for
conventional HV power supplies because the stored energy in the power
supplies' filter capacitors is capable of supplying enough current to
damage the tube even if the primary power is turned off quickly enough
(we're talking microseconds here).
I won't go into the details and benefits of PS-Squared technology.
You've read enough about it here in my NAB column and you'll see more
about it here in the future as the first PS-Squared powered transmitter
is tested. While I could give you the results that Nat described in his
session, I'd rather tell you my first hand experience. I had an
opportunity to see the first prototype PS-Squared IOT power supply under
test at Comark's factory. As Nat reported in this session, voltage
regulation, as expected from such a fast acting supply, was excellent. I
couldn't measure any tilt on the vertical sync and attempts at messing
things up by hitting the transmitter with varying frequencies of
bouncing APL (0 to 100 IRE units) didn't phase it. I did notice a bit
more noise on the video than I liked, but Comark's engineers explained
that the prototype supply had very little filtering on the primary side
of the supply. This was coupling a lot of switching supply frequency
energy into the primary power. Any reports of AM radios being jammed
around the Comark factory? During the testing one of the IGBT's in one
of the 4 KV power supplies caused the unit to shut off. As Nat promised,
the other supplies picked up the load and the transmitter was able to
stay on the air with the seven remaining units. I was impressed.
Returning to Nat's paper, he pointed out that crowbar circuitry can
protect the tube yet still damage other components in the transmitter
system. Shorting 30 KV to ground creates an enormous spark gap
transmitter, capable of propagating high power, low frequency RF energy
through beam power supply leads. This energy surge can find its way into
control systems, causing either outright failure or weakening of
components leading to intermittent failures. Comark used a space
consuming Faraday shield to protect circuitry in their pre-PS-Squared
transmitters. Eliminating the crowbar eliminates this potentially
damaging RF surge. I verified that the PS-Squared supply passes the
"wire-test" when I evaluated the unit at Comark. The PS-Squared supply
also boasts high efficiency - 97 percent and wide bandwidth - 10 KHz.
for better regulation. The nature of the supplies combined with a
digital control circuit are capable of generating any waveform desired
-- a gentle ramp-up for turn-on for example. Nat also felt the fast turn
off time and limited current capability should make the PS-Squared
supply much safer than conventional HV supplies in the event of
accidental human contact. He noted, however, that he had not done any
tests on this feature and did not plan to!
The last session Friday was by Dr. Timothy P. Hulick from Acrodyne.
While previous speakers were concerned about protecting the IOT, Tim's
approach was to do away with the IOT altogether! His alternative is the
60 KW Thomson Diacrode, which operates with a plate voltage of 8.5 KV
and does not require any sort of crowbar for protection. Much of the
session covered items you've already read about here. The paper is
available in the SBE Proceedings. Dr. Hulick pointed out that Acrodyne's
Au60D, based on the Diacrode, is the highest power single tube UHF
transmitter in the world. (RCA never made an internally diplexed
transmitter based on the 110 KW klystron Varian built for them.) One
thing Tim said which I hadn't mentioned before was that if aural power
is limited to 5 percent (instead of the normal 10 percent) of peak
visual power, the tube can be operated at 73 KW peak visual power.
Dr. Hulick made a comparison between the Diacrode and the IOT which drew
some criticism from the audience. Dr. Hulick claimed the IOT had
problems which were still unresolved and that the Diacrode's expected
life should be based on the data available for another high power
tetrode UHF transmitting tube -- the TH-563 25 KW tube. It was argued
that while it is true that IOT's have had their problems, it is also
true that as the technology has matured, the problems have been solved
and many IOT's are operating well beyond 25,000 hours. Indeed, many of
the problems seen with the early Eimac Klystrodes (another name for an
IOT used by Eimac and Varian) were not anticipated and became apparent
only after the tube was in the field. Even though the TH-563's
reliability has been documented, only experience will show if the
Diacrode, operating at a much higher filament current and a much higher
power level, will share that characteristic.
Early test results on the durability of the Diacrode and Acrodyne Au60D
are encouraging. As the deadline for this column approached, I got word
(plus a video tape and a full log) from Mitch Montgomery at Acrodyne
that they had successfully operated the Au60D Diacrode tube at full 60
KW peak visual power with 6 KW aural power for 24 hours with a black
(sync and blanking) video signal . While this can't duplicate five years
of field experience, I venture that many older UHF TV transmitters in
the field today couldn't handle this load that long.
Reviewing the hourly long of readings taken during the test, I'm
impressed at how stable they were, even under a period of high ambient
temperature after the departing manufacturing crew shut off the air
conditioner in the test facility. Plate voltage was constant at 8.5 KV
over the period, while the plate current varied between 10.9 and 11
amps. Screen current, a problem earlier as you may recall, varied
between 160 mA at the start of the test up to a maximum of 170 mA. an
hour later. Twenty two hours into the test it was back down to 160 mA.
The temperature rise across the heat exchanger was also stable --
varying between eight and nine degrees Centigrade. Visual output power
readings varied from 100 to 103 percent. Acrodyne detects the 4.5 MHz.
different between the visual and aural carriers and uses that for aural
power measurements. Unfortunately, this resulted in the aural power
meter consistently reading up to 30 percent high during the test.
Acrodyne assures me that using a spectrum analyzer the average aural
carrier level was precisely 10 dB below peak visual carrier.
In the interest of journalistic accuracy, I do have to point out that
Acrodyne's own log shows that the sync level during the test varied from
42.5 to 44 IRE units. That means the transmitter was operating about
half way between true 75% blanking (as used for FCC power calibrations)
and the 7.5 IRE unit setup used in the United States. At 75% blanking
(40 IRE units of sync and no setup) a 60 KW peak visual power is equal
to 35.7 KW average power. The sync levels reported in the Acrodyne log
would put the average power between 33.8 KW and 34.5 KW, around 95% of
the worst case level. Given the stability of the readings, I don't think
the extra five percent would have caused the test to turn out any
different. Also, because black is supposed to be 7.5 IRE units above
blanking in the U.S., in normal operation the transmitter would never
have to deliver as much average visual power as it did in this test.
Copies of the papers mentioned here plus the others related to radio and
video production are contained in the "1995 SBE Engineering Conference
Proceedings", available from the Society of Broadcast Engineers (SBE),
8445 Keystone Crossing, Suite 140, Indianapolis, IN 46240, telephone
+317-253-1640, fax +317-253-0418. Request a listing of other books
available from the SBE Bookstore. They offer SBE members up to 20%
savings on books from CRC Press, Inc., McGraw-Hill Professional Book
Group,, Focal Press, Prentice Hall and others. The SBE's Web site
address is . (Ignore the <>`s).
I'll be monitoring the progress of both these new technologies -
Acrodyne's 60 KW Diacrode UHF transmitter and Comark's PS-Squared HV
power supply. Indeed, KVEA, Telemundo's O&O in Los Angeles, will be the
first PS-Squared installation in the United States. Working with Mario
Lazzari, KVEA's Director of Engineering, I hope to be able to bring you
a step by step description and pictures of the installation, which also
includes a new Andrew Trasar antenna optimized using the population
study techniques I wrote about a while back, on my Internet Web Site.
I recently completed adding my RF columns from 1993 to my Web page. That
year included a lot of columns that continue to generate requests for
reprints. One of them was the Cheap Remote series. I've posted the
complete set of articles, with picture, schematics and software, in the
1993 columns section. Listing all the associated graphics involved with
that series has led me to consider removing my columns and associated
files from my FTP site and converting them to HTML files with hotlinks
to other articles, Internet sites and graphics. Such a move would also
get around the 20 connection maximum my Internet provider places on the
FTP site. If you do not have access to the World Wide Web (either
graphic or text based such as Lynx) and want me to keep the files
available by FTP, let me know by e-mail. Otherwise, sometime next year
they'll be moved. Note that all the articles can be viewed from my Web
site, just remember that you may get a "not available" type message if
20 users are already accessing my ISP's FTP server. Keep hitting the
"Reload" button on your browser until it connects.
Here are all the addresses and info you'll need to contact me. Visit my
RF Page at , or my FTP site at
. E-mail me at or
. Compuserve users can reach me at <70255,460>.
Ignore the <>`s. I used them to separate the electronic addresses from
the rest of the sentence. Type the text inside the <>`s only. You can
also fax me at 305-884-9661 or phone me after 6 PM eastern time (when
things quiet down a bit) at 305-884-9664. Both numbers are at the Miami
Telemundo office, so expect a delay in a response if I'm traveling. My
mail service address is 2265 Westwood Blvd., Suite 553, Los Angeles, CA
90064. Because I'm often traveling, if time is critical please contact
me for a local address before sending items by mail.
Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED