All Digital Mode?

[fybush]

Or can digital and analog still not co-exist like that?

No matter how many times you ask what amounts to the same question, the answer is still "no, they can't."

RF is RF. Digital and analog are just modulation schemes for that RF. There's nothing magical about "digital" that would make RF invisible to an analog receiver, or vice versa. The analog receiver is still receiving plenty of RF. It only appears to be "not detecting any signal" because that signal is modulated in a way the analog receiver can't decode.

 

[radioskeptic]

What could have made anyone think that two signals could occupy the same spectrum at the same time without any mutual interference?

Perhaps a dim awareness of quadrature AM, coupled with a complete lack of understanding of it?

Two AM signals in quadrature—sans carrier—can be recovered with minimal crosstalk under good, or even fair, receiving conditions. But accurate reconstruction of the intelligence depends on a reference for the oscillator that supplies the replacement carriers—whether continuous, as in FM stereo, or in short bursts, as in analog color TV.

Obviously, on the AM broadcast band there would be no reference signal for accurately restoring the carriers, so that suggestion is unworkable.

NTSC analog color TV used two chrominance subcarriers in quadrature—specifically, two VSM signals with suppressed carriers—to transmit signals which allowed receivers to recover the red and blue video signals. These recovered signals could then be electronically subtracted from the luminance signal to recover the green signal. Carriers for the VSM chrominance signals could be could be restored with a high degree of accuracy by a keyed oscillator controlled by the color burst signal transmitted during the horizontal blanking interval.

This worked reasonably well most of the time, but multipath problems—airplane flutter, or walking too close to (or worse yet, actually touching) the “rabbit ears” on a color TV—could yield some interesting, if annoying, effects on the color.

The “FMX” system of noise reduction proposed by CBS in the 1980’s also used quadrature subcarriers (but with full symmetrical sidebands, not VSM). The second DSB/SC signal, in quadrature with the original difference signal, carried heavily compressed L‒R audio.

FMX was discredited when Amar Bose pointed out how badly the separation of the two subcarriers would be compromised by the constantly shifting effects of multipath in mobile reception. The most important point was the fact that, under multipath conditions, the presence of a second subcarrier in quadrature with the standard difference signal would seriously degrade reception not only on FMX receivers, but even on those not equipped for FMX. (See http://tech.mit.edu/V109/N7/bose.07n.html.)

Broadcast Technology Partners (the investment group that had acquired the FMX patents after Larry Tisch shut down CBS Labs in 1986) threatened to sue Bose and MIT, but it was just so much bluster. Bose must have been somewhat cowed by the experience, and that may explain his reluctance to take on “HD” radio—which, at least initally, had the support of several industry interests, all of them larger and wealthier than the speculators who had acquired the FMX technology.

Nevertheless, we can surmise his private opinion of “HD” from that fact that his firm, which was among the first to jump on the Compact Disc bandwagon, has yet to introduce its first “HD” radio product.

 

[iyiyi]

They currently broadcast simultaneous analog and digital signals on FM and MW frequencies running IBOC. Keep in mind that a TV, FM or MW "frequency" consists of a "channel" containing a specified frequency range. A TV channel is 6 mHz, FM is 200kHz and MW is 10kHz in width.

You cannot have an analog and a digital modulated signal work together because the modulations will heterodyne with each other. This creates spurs reacting in the detectors that result in gibberish outputs from both. IBOC can use analog and digital because they separate the analog and digital signals. For example: WBZ 1030 runs analog and also runs digital IBOC at the same time on the same channel. The 1030 frequency itself consists of a 50,000 watt carrier that serves as nothing more than a reference for the sidebands to hetrodyne in the detector to produce the analog output. This analog portion of the signal occupies the 1030 carrier and +/- 5kHz either side of it. This is strictly limited to those frequencies to keep any analog signals from the digital detector. The HD digital signal is currently shifted into the adjacent channels to provide space for the analog. These digital signals are sharply limited to avoid interfering with the analog. Those "carriers" that cause buzz on the adjacent channels exist only to cancel any digital degradation of the 1030 analog signal. I know that some stations have experimented by reducing the digital carriers with some success.
(I don't know if WBZ has performed that particular experiment).

Same basic setup with television. They had to perform a few cute tricks to accomplish this. They had to roll off the video signal at around 3.2 mHz to keep that information from interfering with the color I/Q signal and detection centered at 3,575,945 Hz (which I hope you notice is not a direct multiple of any video frequency). Also the I/Q signal is phase modulated with an upper sideband of 0.5 mHz and a lower sideband of 1.5mHz.

These are a couple examples of two different signals occupying the same spectrum.

Today's phase/frequency locked loops are capable of grabbing signals with a 0 (zero) dB S/N ratio. No reference signal is needed for them to lock.

-

 

[fybush]

They currently broadcast simultaneous analog and digital signals on FM and MW frequencies running IBOC. Keep in mind that a TV, FM or MW "frequency" consists of a "channel" containing a specified frequency range. A TV channel is 6 mHz, FM is 200kHz and MW is 10kHz in width.

Well, no, not exactly. The digital carriers in the FM "IBOC" system are not located within the 200 kHz "channel" used by the analog FM signal. The analog signal occupies +/- 0-100 kHz from the center carrier frequency. The digital carriers are located between +/- 100 kHz and +/- 200 kHz from the center carrier frequency, actually occupying the frequency range of the neighboring FM channels. It is more accurately an "IBAC" system - "in band, adjacent channel."

The same is partially true of the MW system, though some of the digital carriers are located within the bandwidth +/- 10 kHz from the carrier, requiring the analog bandwidth and frequency response to be reduced accordingly. Most of the digital energy is in the adjacent channels.

There is no circumstance I'm aware of, at least in the Ibiquity "IBOC" systems used in the US, in which digital carriers are occupying exactly the same frequency as their associated analog signals, though, as you note, the digital and analog signals together fit (or are supposed to fit) within the same overall "channel."

 

[radioskeptic]

Thanks for correcting iyiyi, Scott. I already explained two applications of QAM in analog in Reply # 18 on the thread “Let’s set the Buffalo News straight!”

Now would anybody like to explain to iyiyi why the color burst was an odd multiple of one-half of the line frequency? All the grown-ups here already know why, but it looks like iyiyi doesn’t. ( I’d do it myself, but I think I’ve already wasted enough time dealing with misinformation on this discussion board for one day!)

 

[iyiyi]

Sorry. I thought I had stated "The HD digital signal is currently shifted into the adjacent channels to provide space for the analog".

 

[iyiyi]

Sorry. I thought I had stated "3,579,545 (which I hope you notice is not a direct multiple of any video frequency)"

 

[radioskeptic]

As I said, it's an integral multiple -- an odd multiple -- of one-half the line frequency.

Do the math!!!!!

 

[iyiyi]

Simple! I specified a direct multiple of any video signal. Not a 'PORTION' of any video signal. Since when did "one-half" become an integer?

 

[radioskeptic]

No one ever said that one-half had magically become an integer, iyiyi.

If I had been more explicit and said, “It’s not an integral multiple of the line frequency, but it is an integral multiple -- an odd multiple -- of one-half the line frequency,” would you have found something else to quibble with?

Of course you would. You want to attack my credibility with any specious objection you can concoct, solely because, though I had been inactive on R-I for months until just a few days ago, I’ve always been one of the most persuasive critics of “HD” on R-I.

Now about why the subcarrier is an odd multiple of one-half of the line frequency: do you know why? If so, let’s see whether you can explain it. (And of course, you have the entire internet available to crib your answer!)

 

[Tom Wells]

I'll bet it has something to do with needing at least 8 to 10 cycles of color burst sync to fit on the back porch of the blanking.

I'm just bein silly there, but it's fun to say.


I'll bet the 3.579...mhz was chosen as having shown the best compromise between the most effective color modulation baseband of 0-.5 ,hz, vs the least appearance of patterning in video, for which carrier supression is also used to prevent.

The luminance signal only "went" up to 3.2 mhz so this keeps all the chroma information from modulating luminance.

This limatation was a, no, THE limiting factor in how much detail NTSC color could show until comb filters
were introduced that neatly dropped out that ringing from the the chroma info and gave markedly sharper pictures
to televisions in the mid-late 1980s.

Even free-running low-Q oscillator circuits will lock onto the phase of an incoming signal and run then free when the signal
stops.

There is nothing magical about the nature of a quadrature detector or manual injection of the signal of a
"beat frequency oscillator", it creates a strong reference of what "would be" the carrier.

If no "analog" caririer existed ( and consider the moments of zero-volt crossing instances in the waveform, that would be
2,000,000 zero-crossings per second at 1 mhz), so during one second of 1 mhz carrier, there are 2,000,000.
instances during which there is "no reference" for the data. Per second. That's a lot of gaps in a data stream.

If carrier is re-inserted to a "sideband" transmission, voice applications may still result in getting a message through,
but seldom satisfy any musical application due to arbitrary frequency distortion.
IF the reinserted carrier were at perfect frequency, so would be the pitch of the received audio.

In the case of digital data, an "assumed" or "locally derived" reference oscillator, somehow synced on a
reference pulse much like the old supressed color burst sync would be needed.

So all digital radio will need either a "vestigal" weak carrier as described by ibiquity for all digital,
or a pulse-synced, then internally generated local "carrier', as a comparator, in whatever detector follows.

Without reference, the data cannot reliably even be distiguished from noise and yet decoding goes on.
Shortwave RTTY with real teletype machines was exactly like that.
On a good day, it would be printing out somewhat readable copy from signals that were barely audible.
But if either station drifts too much, the decoded two-tone audio 2125/2295hz 170hz shift won't be right, and decode ceases.

Similarly, if I would turn off the BFO on the radio, all the teletype decoder would "hear" is light hissing, not tones.
In this case, the elimaination of the "carrier" results in there being no "data" to consider.

Some sort of reference will of needs be required.
It may not continue to be a carrier as we think of now, be it would be providing the same useful reference
for any mode of data propogated at any frequency.

Whether any highly stable and repeatable "reference" standards exist at consumer prices to make this a reality is another thing.

For now I think we'll continue to see the carrier remain because that's far cheaper than putting the
equivalent of a GPS-sourced syncing system for demodulators into every radio.