|
|
This is an exploration of some anomalous shortwave transmissions that this listener first noticed in the broadcasts of Radio Habana Cuba in the late fall of 2003. See below ("Conclusion") for more information and links to other opinions.
It all started when we began to notice a steady erosion of transmission quality on RHC; we sent Glenn Hauser's "DX Listening Digest" the following letter:
AM I IMAGINING THINGS OR IS CUBAN RADIO DESTRUCTING?
Glenn: I have been listening with morbid curiosity, off and on, to
Cuban SW since the time of Castro's revolution, though it's not now as
weirdly amusing as in the heady days of the cold war and height of
"the Red menace."
For a while, immediately prior to 9/11, Radio Habana Cuba almost
seemed to be getting warm and fuzzy; and the fidelity was superb, with
powerful and rich audio and effective modulation...
But that era has ended suddenly, and the anti-western, anti-democratic
propaganda campaign of the Castro regime seems now to have risen to a
fever pitch, though -- ironically -- we really can't seem to hear more
than a faint whiff of it! For Radio Habana has deteriorated to a puny,
unintelligible, distorted mess.
It strikes me that Arnie Coro, who seems to have pretensions to be a
radio technology expert, should be ashamed of himself, if he has
anything to do with the status quo of Cuban shortwave transmission.
-- Steve Waldee (retired broadcast engineer)
Here are screen shots of the strange squealing RHC transmissions, picked up in San Jose, CA. between 9:50pm and 9:59pm, PST, 11/18/03, using an Icom R-75 receiver and 350-ft dipole antenna, with Wide filter, AM detection (bandwidth at +/-6 dB: 6 kHz.) Software used for digitizing and analyzing signal was CoolEdit 2000 (recording and measurements by S. Waldee.)
In this 8-minute recording above, a few seconds of a female news announcer on Radio Japan (11760) was captured at about -1 dB level for reference; this appeared to be "normal" full modulation. After a blank space of a few seconds of silence, the long section to the right at much lower modulation is RHC's signal (no change of audio or RF gain was made: this is the relative audio modulation level of the two stations, which had approximately the same received signal strength.) RHC's audio consisted of very faintly heard music and speech, with the steady background whine being much louder than the voices, which were more than 20 dB softer than R. Japan's announcer. The "blips" of louder signal during the RHC audio segment were sudden bursts of volume caused by very rapid selective fades/reinforcements of carrier level and sidebands.
You will note -- on the Radio Japan signal if not clearly in the RHC signature -- that there is a DC offset, caused by all the recording equipment used to make these sound samples, that resulted in a slightly higher amplitude in the positive direction of the baseline. This is quite often observed in "quick-n-dirty" digital recordings unless great care has been taken to compensate for small direct-current leakage through coupling capacitors in the audio chain.
Smoothed Fast Fourier Transform, RHC "squealing" noise
An FFT measurement of a short two-second segment of the continuous squealing noise on RHC shows several noticeable frequency blips between 100 and 800 Hz, with a very strong general region of signal around 1000-1200 Hz. To the ear, the noise consisted of more than one tone: a very complex audio signal that varied slightly in pitch and volume, with faint, distorted program material underneath.
Unsmoothed higher resolution FFT plot showing jagged frequency
contour
Noise Profile Plot Derived by Alternative Analysis
Individual Cycles of Noise Signature
Three-Dimensional Plot of Audio Spectra
Having spent many years as the chief engineer of numerous AM and FM stations, I am well-acquainted with the use of composite FM STL's for supplying program audio, and for feeding transmitter control telemetry. Usually the telemetry subcarrier is at a higher frequency than the baseband audio, and is frequency-modulated by series of tone carriers derived from control signals for operating transmitter functions. THIS SQUEAL SOUNDS LIKE A TELEMETRY SIGNAL! In fact, it might signify that a microwave studio-to-transmitter link that supplies both program audio and telemetry is off-frequency or drifting. The distorted quality of the faint program audio furthermore sounds remarkably like the sound of an FM radio that has been badly mis-tuned. Ergo, the resulting combination of tone signals AND weak, fuzzy program audio (plus noise) could be a mistuned STL receiver that is picking up BOTH carriers -- program and telemetry -- and not properly isolating them; or -- equally possible -- a microwave STL transmitter whose frequency control has lost lock, or whose reference oscillator has drifted.
Suppose that the RHC transmission system is semi-automated, and that the transmitter operator does not always check the demodulated off-air signal (particularly if he or she might not understand English); a bad STL in the English feed could conceivably get on the air without being noticed. It is likely that the programming is recorded, or is delayed or rebroadcast in the evening (the announcers, newscasters, and hosts MIGHT NOT even be in the studio at the actual time of broadcast, which could well explain some of the strange moments of dead air, wrong carts and feeds, skewing tapes, and other gaffes that one has noticed from time to time in their night-time English programming.) So, this writer assumes that the errant STL (which may work correctly PART of the time, but not always) happens to get on the air in its "broken" state on certain occasions, with no one at RHC the wiser.
But is this the only possible scenario? What if the strange artefacts, noise, and distortion in the program were intentional? What if they were due to a misguided, possibly incompetent test of digital broadcast transmission?
Some background is perhaps required at this point. The original impetus to check this apparent problem came to me from reading comments by noted shortwave guru and DX expert Glenn Hauser, who wrote in his "DX Listening Digest" of 12 November 2003:
I had been hearing the squeal and distortion artefacts for some time, and was not altogether in agreement with the explanation, posited in Hauser's guide by WWCR manager George McClintock, that it was caused by the failure of one or more low-level digital modulator modules in the final stage of a pulse-modulated high powered SW transmitter; my extensive followup remarks about that suggestion were reprinted by Glenn in his DX Listening Digest for 14 November 2003. However, subsequent comments by Dave Frantz about "WWRB testing in full digital" in the 16 November 2003 issue -- in which he explains that the station was testing their 'own' process for low-cost implementation of digital shortwave transmission, made me imagine that the Cuban transmissions might not necessarily be "mistakes". For Frantz observed that WWRB was using independent sideband mode, with a digital data stream from an AOR 9800 Encoder on one sideband, and normal analogue audio modulating the other sideband, rather than using the expensive and complicated DRM exciter.
Could RHC be doing something like that, too? Well, my first test was to see if their transmission was in ISB mode: nope, surely not. There was no audible change in sound when putting my Icom receiver into SSB mode and selecting either upper or lower sideband; the combined noise/squeal/distorted, weak program audio seemed about the same either way. Was the squealing artefact frequency modulated? My Icom can do FM demodulation, and no sense at all was made of the transmission in that mode. Could the program consist of some form of digitally-encoded audio superimposed over analogue audio, a test of some kind of compatible digital/analogue mode? Well, if so, the test surely is a failure for listeners with standard AM receivers, for the program quality takes such a severe "hit" that it becomes useless.
Furthermore, on the tiny island of Cuba, with a Communist government and Marxist economy, the cheapest DRM-type radio would cost -- probably -- much more than an average citizen's individual monthly income; compare that to the price of an inexpensive Chinese shortwave radio, which might be ordered from various Internet entrepreneurs for as little as $10US! The idea of using digital broadcasting -- NOW -- for Cuba will certainly not help their own impoverished citizens; maybe, however, some weak-minded "authority" in the Cuban broadcasting elite has come to the silly notion that by testing a form of digital broadcasting, the country is doing something "progressive".
If so, the "progress" we hear consists of unlistenable, uselessly distorted, noisily unintelligible radio that seems utterly pointless, a waste of electricity. In a strange and ironic way, though, we might sneer approvingly at this, since most of RHC programming is now so riddled with Communist agit-prop and anti-US rhetoric -- consisting mostly of re-writes and bombastic inflations of criticisms of US foreign policy that one can find here in the United States on our own news and commentary websites in this land of free speech and democracy! -- that the noisy transmissions might help to "improve" the quality of the programming by rendering their totalitarian propaganda virtually unintelligible as reproduced by a conventional SW receiver.
Therefore, traditionally the United States has not felt the need to suppress dissenting foreign broadcasts that do not support its policies; we feel our citizens have a right to hear alternative views and are politically mature and can make up their own minds. Furthermore, the political criticism of America's foreign policy and President, heard regularly on the so-called "newscasts" of RHC -- essentially propaganda programs in this writer's opinion -- are often mild compared to what US citizens can read or hear in their own country. For instance: this writer has heard far more pointed, even outrageous, criticism via the talkshow programs on KGO-San Francisco, and has read much more disturbing stuff on the Internet, free, unfettered, and available on any American's computer desktop with a few mouse-clicks. So, why, then, jam the mere prattling of RHC? (For the opinions of Cubans regarding America's transmissions to them, click this link.)
Again, there is much debate -- especially here in the USA -- about the nuances of what consists of "suppression of ideas" when it comes from ideological pressure from commercial entities, news media, or political groups and their often rancorous criticisms of opposing views. But there is no institutional United States government department with the delegated task of jamming foreign broadcasts, or suppressing political ideas that emanate from other cultures (although it has been argued by some civil libertarians that such governmental activities as outlined in this Guardian article are somewhat related -- though the American government denies this, stating that there is a difference between a specific criminal investigation, and an attempt to suppress political information or disagreement.)
That is obviously not true in the Communist totalitarian states (often having political monocultures, one ruling party, oligarchies of elite leaders who are not elected, and no institutionalized process of changing governing structures at a grass-root political level.) So, it's obvious WHO is jamming WHOM: the Cubans jam America's Radio Martí, not the other way around (as we generally understand from our history.)
However, when one hears the "squealy" RHC transmissions, their sound quality and efficiency is abysmal. As shown in the graphs above, the audio program modulation level -- in this instance -- is at least 20 dB below another equal-power SW station, Radio Japan (and, we should add, Radio Japan used far less audio processing than, say, VOA, Radio Nederland, or DW: and sounds a bit softer in volume, though their modulation density is maximal, due to psychoacoustical effects of different styles of audio processing.) RHC's "squealy" programs are, however, almost INAUDIBLE with terrible weak voice or music modulation. Yet at other times, without the "squeal", RHC programming that is listenable is processed with much higher audio levels -- even some obvious audio peak clipping -- with an apparent program volume increase of 10 to 20 dB over the "squealy" programs. External jamming can NOT cause that.
Ergo, the problem exists AT the facilities of RHC -- if it indeed is a "problem" rather than an intentional transmission.
There is proof, of sorts, that these Spanish-language "numbers" transmissions emanate from Cuba: instances of mistakes or incompetence that has led to a switch over from the numbers recitation to RHC programs, or faint RHC audio being heard in the background during the numbers recitations (as if a radio speaker was playing.) Unless these are weird "spoof" broadcasts, one must use Occam's Razor to conclude that they are evidently Cuban intelligence operations.
Any other suggestions? You might listen to our sound sample, check regularly the broadcasts of RHC on different frequencies, and then email Glenn Hauser and give your opinion: check the "Contact" link at the top of his World Of Radio webpage.
Tonight, I listened for more than 3 hours to the BBC, while working on computers in my shack, and heard the news many times over the World Service, usually featuring a repeated story out of Cuba that is very encouraging: the announcement of a new vaccine which, if effective, could greatly help treatments of pneumonia and meningitis (see this link on the BBC News website.) However, as if to prove my point, expressed above, that the RHC "news" editors are eschewing practically any positive story, even ones that would point with justified national pride to the Cubans' own genuine accomplishments, the RHC "prime time" English newscast broadcast by Langston Wright to the USA tonight -- on 27 November 2003 at 0500 -- once again dealt only with criticisms of the US...and, for a change, of Israel, and the US's support of Israeli policy; not one syllable about this potentially important Cuban medical breakthrough! This is sure confirmation of our belief that RHC programming has become stuck in a groove, like a broken record. WHY did they not even deign to announce the big story about their country that was one of the top items being simultaneously broadcast on the BBC?
Furthermore, the broadcast, on the new transmitter occupying 9550, was (as usual) terrible sounding: low modulation, mushy sound, and with a constant grind. Further investigation determined this was isolated mostly to the USB, and while tuning in ECSS mode, we located the 'center' of a modulated carrier at 9551.1. It had only the "grind" noise, no intelligence; we'd guess that it was actually being transmitted on the RHC main carrier as a kind of spur. So, standard AM-envelope detection of the newscast was so unpleasantly noisy and muddy that it was scarcely worth hearing, except in the spirit of scientific curiosity.
It was also disappointing to try to listen to the Mailbag program (9820, after 0545) and hear the distorted, clipped audio with RF levels that were more than 20 dB lower than we obtained on this frequency last week (not, we think, propagation problems as 9550 was strong and all other signals we tuned in this evening were normal. No, 9820's rig was probably at lower-than-normal power.) The embarrassing thing about this state of affairs was that one of the program hosts started talking about "RHC's engineer Arnie Coro" -- not an auspicious program to mention him by name, considering the lousy transmission! I shuddered, thinking how I would have felt, being a retired radio engineer, to have a particularly bad carrier and audio being 'attributed' to me!
However, tonight I had to say something about the broadcast at 9600 which came on the air at 0330. I tuned past RHC as the anthem was playing, and during the opening announcements it was obvious that I was hearing the WORST gawdawful 120 Hz hum that has been caught on the air in ages. It sounds as if you had plugged only the center of your RCA pin plug into the jack: just dreadful. In addition to the big loud roaring 120 Hz there is an overlay of higher harmonics that wobble around: must be the most appalling ground loop of all times. The modulation level is about 30%, if that. What annoys me most is that at about 0335 on came a program about classical piano music -- hey! -- but the hum is louder than the instrument's loudest notes. BWWWAAAHM! (tinkle tinkle) MMUMMBAAAHMM! That is unacceptable. What gripes me, as I've said before (being a retired broadcast transmitter and audio engineer) is that this Coro fellow, ostensibly the engineer for RHC, has a show in which he touts his own ideas about making antennas, receivers, transmitters, test equipment, etc., and SOUNDS like he knows something. (I note also that he has a fancy website, which does not seem to have been updated since around May, 2002.) Oh, well: must be philosophical. Guess this reveals what happens in "The Workers' Paradise" when the inmates run the asylum.
On 1 May, they celebrated the 43rd glorious anniversary of RHC, which I forgot about and failed to tune in; but Señor Arnie Coro Antich was only too happy to contribute a bombastic essay to Glenn Hauser's DX Listening Digest No. 4074, which outlined the alleged "improvements" to their transmission systems. So, having read his glowing promotion piece, I naturally assumed that when I tuned in, I'd be rewarded with "professional quality" audio and full distortion-free modulation. (pause...beat, beat, beat.) WRONG!
Here's what I heard, sometime between 0440 and 0450 on the evening of 3 May 2004. The transmitter at 9820 was off the air (it came on again after 0500, with the same lousy distorted sound as usual.) The brand-new facility at 9550 was on the air, but modulated with program material -- very mushy and indistinct -- about 35 dB below normal. Above the level of discernible program material there was a nasty audio hum, which I would guess must have caused their modulation monitor needle to "hang" at about 25%. And 9655 at least had intelligible modulation; but it was tinnier than a toy paper-cup-and-string telephone, modulated about 30%, tops. I had to document this silliness to counter the disinformation in Coro's boilerplate agitprop. So I prepared a sound clip lasting about 16 seconds, the first five seconds being the signal at 9550; then seven seconds of 9665; followed by a reference signal obtained from the BBC World Service at 5975 so that you can judge the relative program modulation levels and quality differences. I might add that the RF level of the RHC transmitters was actually a bit stronger than the BBC signal, so the relative volume differences (with my Icom's AGC working hard) are those of the actual modulation, not differences in received RF signal level.
BUT, dammit! This is one of the audio files that I can't find, now that I have taken the trouble to return this article to the net. Oh, well...my description is pretty vivid.
SO, we have the BEST that the glorious heroic people's broadcasting workers can achieve, after spending $200 million in communist Chinese funds: and THIS is the ridiculous result! As I said, the joke continues. And while logging the lower shortwave bands on 5 May, I made note of this new interference problem:
I have only one thing to add: "Professor" Arnaldo Coro Antich: for shame! - SRW
No, Waldee: it is not incompetence. For the Cubans have now become so adept at jamming, as witnessed by what they are doing to R & TV Martí, and have done for the Iranians with their satellite jammer, that surely THEY ARE PRE-JAMMING THEIR OWN RHC BROADCASTS TO THE UNITED STATES!
Plausible? You be da judge...
In the period of "classic old time radio" of the forties and early fifties, before the advent of modern audio processors, while the rest of the "broadcasting chain" was delivering nearly hi-fi performance, the bottlenecks where sound quality suffered -- at the transmission end -- were generally two important factors: (1) the equalized telephone lines (only 5 kHz bandwidth for network shows, though local studio-to-transmitter links were 8 kHz bandwidth types that sounded much better), and (2) perhaps even more important in reducing listenability, the crude and rudimentary audio limiter or clipper used to control modulation levels. For these latter devices were, only slightly earlier than just a generation ago, in a rudimentary state of conceptualization, at the very same time that almost every other part of the broadcast chain could be adjusted to sound nearly as good as the finest analogue equipment extant in 2004.
A large part of the giant leap in AM radio broadcast sound that has occurred in the past two decades may be attributed to the final evolution of modern audio processors, such as the amazing Orban Optimod [tr], which may be adjusted to produce both a loud and a clean transmitted signal (basing its remarkably sophisticated audio waveshaping on an advanced application of psychoacoustical principles, a comprehensive analysis of neurophysiological human aural perception, and an appreciation of of the physics and aesthetics of musical sounds: which I know about first-hand, as a former assistant and consultant to the developer, Robert Orban, back in the 1970's during his design and development of the device, as I handled much of the earliest west coast field-testing; click
here for an interesting account of Bob's own personal version of what he recalls of some of the events that led to its development.)
But the earlier audio processors were, frankly, TERRIBLE. Sound of high quality went IN; and CRAP came out: weakened, tinny, "pumpy", jumpy, jerky, congested, and often harshly distorted audio was produced by even the finest brands of processors made by Gates, RCA, CBS, or Raytheon: then this ugliness was fed into the transmitter, not the beautiful, rich, sweet sound that could be heard at the output of the mixing board.
Listening tonight, on 10 May 2004 at 0410 UTC, to Radio Habana Cuba's transmission at 9820 kHz -- yes, they seem to have fixed it once again, and have put back on the air a (partially) usable, and at least intelligible, signal -- I was taken back to the "bad old days" of fifties' radio at its worst. When I rekindled my interest in SWL activities about three years ago, and tuned in RHC broadcasts at 9820, I very frequently heard very good, pleasantly full and rich sound of the incomparably superb Cubano music that I love to enjoy. But, as I have observed in the main thrust of this web page, that status quo is long past, and we have now the bad audio that I personally spent more than twenty years of my professional career as a broadcast audio specialist trying to banish (in part, by working with Orban and later with Inovonics, both manufacturers of modern processing gear; but mostly as a dedicated radio broadcast technician and chief engineer.)
For, even after the expenditure equal to 200 million US dollars -- supplied by China -- for the upgrading of RHC facilities and Cuban television, no one has thought about the audio! It is evident that RHC, at 9820 (whether or not employing at any given moment, an old Brown Boveri, or a new modern and efficient transmitter) is STILL utilizing what radio engineers called -- quaintly and euphemistically back around 1958 -- trapezoidal modulation. The term is generally used in two ways: one, to illustrate a particular means of measuring a modulated AM carrier; and two, a type of audio processing that creates clipped, tilted waveforms; it is the latter phenomenon that I am attempting to describe in this present discussion.
In this second context, "trapezoidal modulation" means, simply, that a couple of diodes have been shunted across an audio circuit, forming a clipper and holding the output amplitude peaks to a certain vaguely ambiguous "ceiling" value. For those old soft clippers, used on transmitters from the late forties until, in many cases, the late sixties (up to the advent of the famous CBS "Volumax" and competitive processors) just simply caused DISTORTION, of a type that is remarkably similar to the unpleasant sound of over-driving an amplifier or radio by "turning it up too loud." They "compressed" or processed the audio signal merely by clipping it; and the topology was so rudimentary that anybody could make one with a few resistors and diodes in minutes. Sometimes these "trapezoidal" devices were built into the transmitter inputs -- Harris Corporation still put them into AM transmitters even at the time that I retired from radio in 1989 -- and they did their damage by instantly clamping the peak audio levels, introducing HUGE amounts of harmonic and intermodulation distortion -- but also by making the audio seem louder, harsher, and (in effect) more "punchy". If you could STAND to listen, that is...the fatigue factors were huge!
The term "trapezoidal" was added, as a fancy buzz-word, to 'spin' one of the negative drawbacks of the circuit: that the clipped waveform ceiling would be phase-shifted and distorted and altered by the transmitter's audio and modulator stages -- and even the output and antenna system -- from a nice tight squared off waveform, to a jaggedy sawtooth shape: a trapezoid! This meant that "trapezoidal modulation" was at once, loud and blasting, and at the same time, wasteful and inefficient: for a lot of potential modulation power was lost by having to compensate for the the phase-shifts by lowering the modulation level. You see, the engineer has to adjust the tippy-top of the tilted peak of the trapezoidal waves up as closely as possible to 100% modulation; but depending on the whole Rube Goldberg system, the lower amplitude end of each half-cycle of the tilted waveshape might drop from 1 to 3 dB or even more at some critical frequencies. A careful, considerate engineer -- understanding how to truly employ a proper peak limiter -- could get a louder -- and INFINITELY CLEANER -- air signal without the "trapezoidal clipper", by merely adjusting the time constants properly and then modulating just right, with the controlled -- but undistorted -- peaks at 95 to 99%. This sounded often just as loud, during the short term momentary program audio peaks, as the most squashed, distorted, and congested "trapezoidal" modulation; but over the long term, the trapezoidal processing SEEMED a bit louder. Again: if you could STAND to listen that long...
Left Waveform: Low Frequency Sine Wave;
Right: Heavily Peak Clipped Wave Becomes Squared
Bottom graphic shows what happens when this square wave is applied to the input of an old, deficient amplitude modulated transmitter with a significant amount of phase shift, primarily caused by poor low frequency response, resulting in baseline shift and square wave tilt. Note that you MUST set the peak modulation values at the "tips" of the highest amplitude of the tilted waves, lowering the average level compared to full modulation of a perfect square wave. At some point, the compensation factor is so great that you lose all "loudness enhancement" over a NON-clipped input signal. Thus, "trapezoidal modulation" is an old, inefficient, discredited processing technique.
Trapezoidal, Phase-Shifted Clipped Waveform:
Tilted Waveform Requires Backing Off Level
THIS is precisely the audio quality right now on RHC, and what the transmitter at 9820 has been cranking out for some time now: months if not more than a year or two. It sounds to me as if an ancient, crude "trapezoidal clipper" is almost the only audio processing (or perhaps it is preceded, as I hear from day to day, with varying amounts of gross audio limiting effects from an equally crude, old-fashioned "broadband" audio limiter. This varies, as RHC's audio levels are totally inconsistant from moment to moment, program to program.)
The problem with trapezoidal modulation is that -- according to the antenna current meter -- it works! If I recall my old textbooks rightly (this is not a topic I've thought about much since around 1975, when considering ideas being proposed for the Orban AM-Optimod) the antenna RF current will rise about 22% with maximum undistorted amplitude modulation of a sine wave. And the old rule of thumb in AM radio was "your signal goes farther as you increase the modulation": a pernicious half-truth that was not a proper understanding of the complexities of the effects of loudness factor in narrow-band AM.
As any radio engineer knows, when you measure the power output of your AM transmitter, you must wait for a lull in the audio, so that the thermocouple meter movement can relax and fall back to carrier output level, not mistakenly indicating carrier plus sideband power. The infernal thing about using cheap & dirty trapezoidal modulation (heavy clipping) is that it looks great on that one meter! In conformance to the "loudness equals distance" alleged rule, clipping must, therefore, improve the signal, its coverage area, and its overall efficiency. But, transcending those issues, it sounds bad.
Heavy brute-force clipping produces both intermodulation ("IM") and harmonic distortion products; the highest harmonics are not necessarily going to be heard on a typical radio, and may not even be radiated by the antenna system; but the IM products are the killer! For, typical program material is mostly strongest in the lower middle frequencies, where the average energy is highest. Those are the frequency peaks that slam hardest into the clipper, and produce the richest IM sideband products; and those spurious signals "fill up" the audio passband with junk, tending to do exactly the OPPOSITE of what is desirable for narrow-band AM transmission: it causes the sound to be muddy and indistinct and jumbled, not crisp and intelligible so that it rises above background noise while maintaining clarity.
Before the AM Optimod was designed and built by Orban, aside from one or two recent limiting devices (like the "DAP" or the Modulimiter) most of the conventional transmitter limiters relied heavily on brute clipping, by diodes (after the fifties, using germanium or silicon rectifiers, avalanche zeners, or transistor junctions); there were even earlier ones that used fussy critically-biased tube clippers. Yes, the duty cycle could be increased by clipping so that the RF ammeter rose up nearly to the theoretical maximum steady-state modulation level of a tone; but the station's sound did not necessarily become easier to listen to, understand, or enjoy: often it drove away listeners, stressed the equipment, and in some cases even shorted out or broke transmitter and antenna parts! The spec for, say, a 1952 RCA 5-kW AM transmitter that I used to work on was "average continuous modulation level=30%"; above that, and you could blow out the modulation transformer (as one of my "hot" early custom audio processors actually did, once, at KEEN Radio in the 1970's, even though the peak modulation level was precisely legal.)
What I perceive in the RHC modulation techniques is that the station SELDOM reaches full modulation; but often sounds as if the signal is heavily clipped. I intuit that the RHC techs do probably know, from rueful experience, that their old rigs can't modulate steadily at 100% negative carrier peaks. So, they use EXTREME amounts of clipping but feed the transmitters so that the peak negative modulation level is rather low: maybe 75% or less -- some days very much less, depending on how inattentive they are to details. The heavy clipping is probably evidence of the use of merely old junk that is hopelessly obsolete, rather than newer (post 1970) processing gear.
One of the other frustrating things about trapezoidal modulation -- heavy clipping -- is that it is so damnably indefinable. No two engineers agree on what sounds about right: one fellow thinks it's distorted, while the other one can't hear anything wrong; there is a constant tug-of-war about how much or how little to clip. Even if one tries to rationalize how much to use, then you have to worry about how much will degrade the signal; reduce its clarity; drive away listeners who don't know WHY it sounds bad, but still perceive that something is wrong.
The amazing thing about this is that in the United States, almost ALL the old timey deaf idiot radio engineers I used to know and hate are all gone...dead, mostly. The new crop have read something about audio processing, and are familiar with the post-Orban theories and understanding; and the managers and program directors aren't fools, and won't tolerate deaf and grindingly stubborn tin-eared transmitter engineers any more.
And I have discovered that American RADIO HAMS are also aware; years ago, amateur radio was abysmal sounding. Today, it's crisp and clear. Some are experimenting with illegal wideband "hi fi" sound; some refuse to use peak limiting or clipping at all, accepting a lower modulation level in trade for "purity" and clarity and smoothness; and even single sideband HAMs are sounding very good. But this knowledge, apparently, is not drifting down to much of Latin America and the middle east, whose commercial broadcasting and SW stations tend to sound badly distorted most of the time.
Remember: the information is out there and in references and obtainable resources; the proper equipment has been manufactured for so many years now that used versions are selling cheap and can make a station sound very nice indeed for dimes on the original dollars. If you hear bad, distorted sound on shortwave radio, and it is demonstrably NOT caused by propagation problems or your radio -- well, then, the problem is at the STATION, and is in the hands of the engineering staff, who are acting as if it's 1910 and they are using carbon-button telephone sets rather than 21st century radio equipment.
The great-sounding BC and SW stations employ intelligent processing, using techniques devised since at least the mid-70s, if not in the '90's. They needn't be over-processed, or over-clipped, to be able to sound loud, clean, punchy, and to be able to cover up at least some of the atmospheric noise (except in periods of severe fading.) I don't imagine FOR ONE MOMENT that VOA, or BBC, or DW are relying on "trapezoidal modulation" clipper gadgets consisting of simple shunt diode circuits as their essential audio processing. But surely that's exactly what RHC's transmitter at 9820 reveals, by its pitiful sounding and distorted audio -- that is, when it manages to stay on the air and not sound like the mp3 file in the above section of our article.
In fact, information specifically describing the Brown Boveri transmitter used for RHC's 9820 (at least up to the very recent time, and possibly at this moment) states that "a peak clipper may be connected at the input for trapezoidal modulation", as you may read at this link in the "Transmitter Documentation Project" that describes the BB model SK-51, the old 100 kW transmitter used at Cuba's Bauta site for the 9820 service. QED.
Well, enough of this ramble for tonight. For as the distorted Cuban jazz droned on in the background, the music program drew to a close just as I was finishing up this section of the web page. The Cuban anthem came on, but then at about 0459 = POW! = there was a loud electrical snap, and the audio level dropped at least 20 to 25 dB: now, the signal is barely audible, though the RF level hasn't changed a bit. I checked my radio by tuning up and down a few channels. Yep, not MY problem; RHC's. As usual.
In fact, as I would expect from this, the harmonic distortion and splatter being generated by this terrible signal pretty much cover up the audio up at 9840 of Radiodifusao Portuguesa, Lisbon, and traces of the lower-sideband snap, crackle, and pop could be detected at the "dead" frequencies, where I could not hear any stations, as low as 9775! This is, of course, considerably worse than the bandwidth of the DRM hiss that Coro complained about from interference to Cuba's signal at 6000 kHz, a few months ago. This surely is the same phenomenon I described above, in discussing Radio Rebelde's splatter all over the region above and below their frequency of 5025. This, of course, may be almost totally avoided by doing TWO little things, neither one of which requires a Ph.D. in electrical engineering, but just the TINIEST bit of good old capitalist horse-sense: (1) installing a simple audio lowpass filter to limit transmitted bandwidth to something reasonable; and (2) NOT cranking up the negative modulation til the carrier pinches off. (We know this here in the 'evil empire', but it seems not to be understood by the great and heroic peoples' engineers in Cooo-bah.)
ASIDE: Why would I deign to give these "socialist heros of da pipples" my advice? Perhaps it's best to leave this ludicrous state of affairs alone, so that we can continue to be amused at the pretension of these dupes as they carry on their little charade of being "real" broadcasters. (And, after all: let's give them credit where it is due: they sure know their jamming techniques!)
This was, of course, intended as a joke; but things have a way of coming true. Tonight, I tuned across the upper end of the 31M band while looking for signals to log, as I checked each frequency against the ILG Radio database. I noticed that ILG are now listing the specific frequencies for "Cuban jammers", at the same time and frequencies that R. Martí signals are on the air. It has been noted in the past that sometimes these Cuban jammers stay on the air after the frequencies have been vacated by the "dangerous propaganda" (as they would allege) of Martí, from Florida. So, apparently, ILG felt that they would differentiate the jamming signals from the Martí signals, since -- depending on where one was located in the world -- the two overlapping signals would somewhat differ in strength and discernibility (if you can "discern" in any real way that horrible bubble noise being emitted from Cuba, fueled -- one assumes -- from an almost endless supply of sugar cane, denied to the Cuban populace in a useful form.)
Well, as it turned out, R. Martí was on the air at the frequency of 9805, and the Cuban jammer was happily "bubbling away" at the same wavelength, to drown it out. But I noticed that the bubbler was wetting down a nice swatch of bandwidth, from about 9793 all the way up to 9815: in other words, right on top of the higher audio frequencies of the upper sideband of Radio Habana Cuba itself, on 9820! So: no joke; the Cuban bubbler was intersecting with RHC's own English transmission to the USA.
As Lawrence Welk might have said, while conducting a bouncy mambo:
ASIDE: The following thought just occurred to me, while listening to Arnie Coro's "DX'ers Unlimited" (but not unclipped) show:
After BPL is established here in the USA, making RCH's signal sound like a QRP transmitter from Pluto, will Cuba announce to radio listeners in search of faint signals, free of the terrible digital interference from the BPL carriers all across the HF bands, that they should move to Cuba to resume their hobby?
Imagine the travel brochures: "The socialist island paradise of low-QRM shortwave, Cubissimo Libre: Free Territory, the Americas!"
Of course, the requirement for unfettered tuning would likely be:
Ironic, eh, in light of Castro's brush with death in 2006, and his continued poor health!
ASIDE: take two. While composing the above remarks, and preparing to paste them into the web page, I was entertaining myself with the pleasant Cubana music on RHC, which was supposed to continue its English program to 0700 UT. But -- and my attention was divided, so I didn't get the exact time -- suddenly the program DIED. Right in the middle...it simply expired. Dead carrier. I left it on...five minutes, ten minutes, fifteen minutes...a half hour. Then I tuned to the parallel frequencies of 9550 and 9655, which had been simulcasting the same show. Silence -- except for the usual trace of Cuban residual background hum. Good carriers, though.
Eventually, somewhere around 0642 UT, a different announcing voice began right in the middle of a syllable at 9655...and also on 9550 and 9820. What happened? Did Habana, but not Bauta, have a power failure? No backup tapes at the tx site, one presumes. (Heck, even my old 1 kW Spanish-language station KOFY in San Mateo kept a backup studio room at the transmitter, with a Gates board, nice ITC triple deck cart machine, mikes, and open reel transport at the ready, as well as backup non-EQ'd phone line link, in the ancient days of the early 1980's. Ah, but of course: I forgot! You see, we had commercials on the air, to pay for these bloated capitalist treasures...)
9655 R Finland, Pori, Finland. female, Finnish, ID, heard under strong dead RHC cx 0429 9600 R Habana Cuba. Dead carrier, //9550, dead cx, //9820, dead cx with no mod! 0428
11760 R Habana Cuba. EE, as below; however barely receivable, covered by R Japan 0635UT 9550 R Habana Cuba. EE, Isabel Garcia, Cuba news, GREAT SOUND! //9820, a bit clipped 0633UT 9655 R Habana Cuba. "Dead" carrier, with bad hum, Isabel Garcia about -40 dB, sigh. 0633UT
Finally: if I wanted to be really picky: amigos, that "National Anthem" cart is SHOT. It's a rather stirring anthem, as a matter of fact, as anthems go...but -- whoops! -- the pitch shot UP and DOWN over and over when you signed off the English transmission; made me sea-sick. (What would "el lider mas grande" think?) Little "touches" like that set RHC apart and bring loads of laughs to "el muchacho de radio en San Jose".
But, as it turns out, at least the difficulties we observed in the very short period of time comprising the last weekend of May 2004 were due to a serious fire at the broadcasting facility, though happily no one was seriously injured or killed. These bizarre events seem to have provided a kind of dramatic climax to our narrative.
It seems as though there must have been a case of rather egregious human error by a workman, whose torch ignited some oil in the basement of the radio building. So my presumption of "power failures" was a bit inaccurate and at least incomplete; something of this nature could happen anywhere, but seems in light of the rather appalling inconsistency of the Cuban broadcast quality variations to add nevertheless to the impression one has of a rather poorly managed outfit. Still: we had a similar fire here in the city of San Jose, California, within the last two years, almost demolishing a new shopping center; and one can only think also of the disastrous fire in Buckingham Palace during the Queen's "annis horribilis" in June of 2002 to remind us that something of this nature is always a possibility when repairs or construction are being effected.
I have cobbled together a very rough translation of some of the items in Glenn's DX Listening Digest, quoting the original Spanish reports that were often oddly conflicting, to aid readers who aren't fluent in that language:
A fire of "medium proportions" without victims was reported today [May 27] in a building of Havana where they operate four official radio stations, interrupting transmission, according to authorities. According to reports to the press by Colonel Mario Alvarez Martinez, second in command of the firemen of Havana, the fire originated in the cellar of the building. "It is not great, we have it totally controlled". Although the report said that "it is necessary to investigate properly to be able to give accurate information", it was suspected the wreck had as its possible cause "some negligence" by workers who were making a repair using an open flame. The official Latin Press news agency stated that the transmissions were interrupted because of the damage. (via Dino Bloise, Flowery South, The USA, Digital connection via DXLD) FIRE AFFECTS BUILDING OF The SEAT OF IMPORTANT RADIO TRANSMITTERS IN CUBA (ChronicAFP, May 27) (2004-05-27) A fire, apparently of considerable effect, damaged on Thursday a central 5-floor building in Havana where were located the radio transmitters Progress and Habana-Cuba Radio, stated the AFP. According to witnesses, the damage began around 0900 local time and caused the evacuation of numerous people who work in the building, located in a large center of transit through the Cuban capital. Firemen used stairways to evacuate several people from the upper floors of the building, while ten ambulances waited a few meters outside the building in case their intervention was necessary. One of the firemen said to the AFP that "there are no wounded", but that one person had suffered respiratory disorders and had to be removed from the building by means of a stretcher tied with cords. The causes are not known, but the presence of several cars from the electrical services suggested to witnesses that the origin of the fire had been a short circuit. The police closed street transit access for 300 meters around the building, where several tankers waited to assist the firemen. Using bullhorns, the police requested hundreds of persons in the immediate neighborhood to clear a zone to allow the work of firemen, police paramedics, and specialists. Radio Progress is one of oldest, greatest, and most popular neworks of Cuban radio, while Habana-Cuba Radio is the international shortwave transmitter service of the country. Both work in the ill-fated building, constructed in the 1950s. FIRE OF GREAT PROPORTIONS IN RADIO BUILDING PROGRESS the extraida News of CubaNet HAVANA, 27 of May (Ana Leonor Diaz, Group Honor/ http://www.cubanet.org) - a fire of great proportions, attributed to an electrical failure, affected the building of Radio Progress and another two transmitters in a central zone of the capital Havana. At the time of the fire sixty people, employees of Radio Progress, Radio Habana Cuba, and the provincial transmitter the COCO, were caught in the conflagration in the building, and had to be evacuated up through the roof due to the toxic smoke. The recording studios of Radio Progress had been under repair for several months to change the air conditioner system, dating from the 1950s. Though the studio dated back 50 years, it had been equipped with modern digital recording technology. All the personnel of the Provincial Commando of Firemen of Havana and troops of the Revolutionary Armed Forces were mobilized until dawn, under the supervision of the minister of the Interior, Abelardo Ibarra; he declined, according to Arnaldo Slaen, to calculate the damages caused by the wreckage (Argentina, May 28, Digital connection via DXLD) ...R. Progreso is orginating from the R. Rebelde studios now; [the article] mentions RHC but nothing specific about their situation: EXTINGUISHED FIRE OF MEDIUM PROPORTIONS IN RADIO PROGRESS Commandos of Firemen of the Capital. Photos AIN --- By Reynold Rasse de Granma The transmissions of Radio Progress originate from the Pan-American studios of the closely related Radio Rebelde. A fire of medium proportions took place yesterday morning in the cellar of the building of the radio transmitter Progress in the capital. It was brought under control and did not cause fatalities or injuries. Colonel Mario Alvarez Martínez, second head of the Body of Firemen in the capital, explained that the wreck began in the morning between 9:00 and 10:00, when an oil spill in the work area in the basement of the building caught fire from the flame of an oxy-(acetylene?) torch being used to install electrical generating equipment. Despite the immediate action of those who were in Radio Progress, and also operating the other transmitters in the same building, including Radio Habana Cuba, National Musical Radio (CMBF), and the COCO (that transmits from there provisionally), the fire spread. Several commandos of the Body of Firemen of the City of Havana took part quickly with specialized fire-fighting techniques. They contributed to the evacuation of the personnel, aiding the Revolutionary National Police, the population, and different organs of the province and the municipality, which had planned for such cases. The corps commander Abelardo Ibarra, minister of the Interior; Esteban Lazo; and Pedro Sa'ez Montejo, first secretary of the Party in the capital -- all members of the Political Bureau -- as well as other leaders of the Party and the Government in the territory, went to the location of the wreckage. The authorities are making the necessary investigations to determine the causes this fire, and to evaluate damages.
As Glenn Hauser observed, depending on whether one was sympathetic to the Cuban government, the fire was either "not very serious", or "serious" indeed.
Surely the articles are all slightly misleading at least in one particular: the transmitters of Radio Habana Cuba are not in the center of Havana, but 40 km away at Bauta (described in this article as "a rural town outside of Havana.") One assumes, however, that there are microwave links or other means of sending the signals to the remote SW transmitter facility, and that the studios of RHC are co-located with the other stations mentioned: "Progress" and CORO and CMBF. Were these studio-to-transmitter links to all the various high powered international SW transmitters at Bauta damaged or disrupted?
On the evening of the fire I logged the transmitters of RHC operating with full carriers but no modulation, inexplicably to me at the time. Then the following night's broadcast employed Radio Rebelde simulcast, not by a normal program tie but by a means that I speculate above to be arguably the connecting of a distorted radio to the RHC transmitter, to pickup the relay -- doing it poorly and with embarrassing results (R. Rebelde's studios and broadcasts were not affected, the service emanating from a nearby building that was not damaged by the fire.) Perhaps the confusion is understandable given that, surely, the most experienced RHC engineering personnel would probably have been called on to assist at the damaged studios; but, still, the whole terrible sequence of events, and the dreadful transmissions heard over the air, must give one considerable pause.
The unspoken but irresistable thought is, of course: could there have been sabotage?
As we considered near the start of our article, above: the months and months of hideously bad transmissions, with squeals and noises and hums that almost obliterated the intelligence (to use a perhaps slightly inappropriate term) of the programs must make us wonder, too. How could any broadcast organization be this ill-fated? To what extent could their obvious problems be attributed to poor training, institutional lethargy, and incompetence...or through more sinister causes?
I think this gives me a convenient place to conclude for at least the present my chronicle of the variable and unreliable transmissions from RHC. I'll tune back in, after a few months have passed to allow a mop-up and general settling of this turn of affairs, before resuming my critical evaluation of their broadcasting techniques.
R. Moscow's Greatest Transmitter Overloads |
R. Tirana Blows A Final |
World's Funniest Sounding RCI Relay Noises
And, seriously: if "radio spook stuff" fascinates you, DON'T MISS:
Sefton Delmer, the British WW2 Radio "Black Propagandist".
The author, and little swl helpers.