HD RADIO SERIOUSLY DEGRADES ANALOG FM STEREO

[SUPERCASTER]

Thanks to radioskeptic for the link.

Noise Levels
HD Radio self-noise levels depend on receiver IF bandwidth, detector type, postdetection filtering, and stereo-decoding method. Since the late 1970s, a standard receiver architecture has become common. It uses two wideband ceramic IF filters, a quadrature FM detector, no postdetection filter, and a squarewave stereo-decoder IC. For such systems HD Radio self-noise typically will be about 50 dB below the level of a maximum-deviation 1-kHz sine wave. The derivation of this figure, which involves both theoretical calculation and signal measurement in a typical tuner, is here. The great majority of home tuners and receivers made in the past thirty years should exhibit HD Radio self-noise levels within a few dB of this figure. The noise level can be up to 10 dB higher for an extended-hybrid HD Radio signal.

Because RMS audio levels for stations with conservative signal processing typically are 10 to 25 dB below peak deviation, HD Radio self-noise can sound much louder than the 50-dB S/N figure might suggest. In fact, I've found HD Radio signals to be unlistenable on every home tuner or receiver I've tried that lacked a special stereo decoder, narrow IF filter, or postdetection filter. Such features are not common.

The worst case mentioned for FM HD station transmitting HD2 extended-hybrid signal:
HD noise reduces analog FM stereo signal to noise to -40 db to about -15 to -30 db below RMS average modulation.
Now you know where the added buzz on the analog stereo signal of a station transmitting HD comes from. It is the HD signal that is causing the noise.
The FM station turning off the HD signal will get rid of the noise HD Radio adds to the stereo analog FM signal.
Here is the link:
http://users.tns.net/~bb/hdrsn.htm

 

[clouseau]

From the Article...
HD Radio self-noise levels depend on receiver IF bandwidth, detector type, postdetection filtering, and stereo-decoding method.

Exactly. Out of standard radio = Out of standard performance. Non of this should matter because Analogue and Stereo FM AND SCA has a 100 khZ Bandwidth and 53 KHz is the top of FM stereo. 53 KHz. Remember that SCA runs at 67 KHz and 92 KHz right now. No HD or anything else. If you were to buy into this garbage then you would believe that RDS would interfere at about 57KHz. If your radio receives outside of 100KHz, it''s WAY "Out of spec". Stereo ends at 53HKz.

The article itself tells why this is Bravo Sierra. Rememeber an FM stereo signal goes to 53 KHz. We "SHOULD" all agree on this.

The two sidebands, [HD] each consisting of hundreds of digital subcarriers, occupy spectrum from about 129 to 198 kHz away from the channel center frequency. Each sideband is 23 dB below the level of the analog carrier. [3db=one half. Do the math]. In some modes, additional subcarriers may extend the sidebands down to about 102 kHz from channel center.

At 102 KHz bandwidth it could conceivable interfere with 92 KHZ SCA, After that it would bother the 67 KHz SCA... then the RDS at 57 KHz and then.... Please. If you have the patience, the article references that they are talking about the following...

From the article...

A stereo decoder with a squarewave oscillator waveform will demodulate noise near 190 kHz to audio with the fifth harmonic of 38 kHz.

A FIFTH harmonic? Are you serious??? Memo to the radio designers. 190 KHz is well over 3 times above anything that resembles regular audio. But it gets better...

A square wave's fifth harmonic is 14.0 dB below the fundamental, so the MP1 noise level is -14 - 33.7 dBc - 2.2 = -49.9 dBr after stereo decoding and deemphasis.

49.9 DBr below carrier. That's as bad as it can get. Got it.

You do know the FCC allows 54 DBu signals on adjacent channels right? They have for many years.

As an FM station, your "Protected contour" is allowed interfering signals on your Co Channel (or the same channel) at a level of 40 DBu. But you crabbing about 49.9 DBu in a worse case scenario NEXT to your channel. This is JUNK SCIENCE!!

1st adjacent (Next to, as in 93.1-93.3) is allowed by FCC rules at 54 dbu.

Now don't get me wrong. I'm sure that somewhere there is SOME poor slob using SOME badly built radio that get a little hiss from SOME FM station SOMEwhere because SOME other station is transmitting HD and the radio is junk. But then again I got one of those junky clock radios in my family room. It won't get that 2nd adjacent DX station that I can always get in my car. Maybe you we should rise up against docket 80-90. ... Never mind.. Forget it, That's another thought for another day...

On a side note...
Hey Soup... Do you really READ these articles? Do you really understand them??? Just wondering.

Clouseau.

Not an HD lover, but a JUNK SCIENCE HATER!!

 

[SUPERCASTER]

From the Article...
HD Radio self-noise levels depend on receiver IF bandwidth, detector type, postdetection filtering, and stereo-decoding method.

Exactly. Out of standard radio = Out of standard performance. Non of this should matter because Analogue and Stereo FM AND SCA has a 100 khZ Bandwidth and 53 KHz is the top of FM stereo. 53 KHz. Remember that SCA runs at 67 KHz and 92 KHz right now. No HD or anything else. If you were to buy into this garbage then you would believe that RDS would interfere at about 57KHz. If your radio receives outside of 100KHz, it''s WAY "Out of spec". Stereo ends at 53HKz.

The article itself tells why this is Bravo Sierra. Rememeber an FM stereo signal goes to 53 KHz. We "SHOULD" all agree on this.

The two sidebands, [HD] each consisting of hundreds of digital subcarriers, occupy spectrum from about 129 to 198 kHz away from the channel center frequency. Each sideband is 23 dB below the level of the analog carrier. [3db=one half. Do the math]. In some modes, additional subcarriers may extend the sidebands down to about 102 kHz from channel center.

At 102 KHz bandwidth it could conceivable interfere with 92 KHZ SCA, After that it would bother the 67 KHz SCA... then the RDS at 57 KHz and then.... Please. If you have the patience, the article references that they are talking about the following...

From the article...

A stereo decoder with a squarewave oscillator waveform will demodulate noise near 190 kHz to audio with the fifth harmonic of 38 kHz.

A FIFTH harmonic? Are you serious??? Memo to the radio designers. 190 KHz is well over 3 times above anything that resembles regular audio. But it gets better...

A square wave's fifth harmonic is 14.0 dB below the fundamental, so the MP1 noise level is -14 - 33.7 dBc - 2.2 = -49.9 dBr after stereo decoding and deemphasis.

49.9 DBr below carrier. That's as bad as it can get. Got it.

You do know the FCC allows 54 DBu signals on adjacent channels right? They have for many years.

As an FM station, your "Protected contour" is allowed interfering signals on your Co Channel (or the same channel) at a level of 40 DBu. But you crabbing about 49.9 DBu in a worse case scenario NEXT to your channel. This is JUNK SCIENCE!!

1st adjacent (Next to, as in 93.1-93.3) is allowed by FCC rules at 54 dbu.

Now don't get me wrong. I'm sure that somewhere there is SOME poor slob using SOME badly built radio that get a little hiss from SOME FM station SOMEwhere because SOME other station is transmitting HD and the radio is junk. But then again I got one of those junky clock radios in my family room. It won't get that 2nd adjacent DX station that I can always get in my car. Maybe you we should rise up against docket 80-90. ... Never mind.. Forget it, That's another thought for another day...

On a side note...
Hey Soup... Do you really READ these articles? Do you really understand them??? Just wondering.

Clouseau.

Not an HD lover, but a JUNK SCIENCE HATER!!

You are confusing the typical receiver's detected FM stereo subcarrier analog audio signal to noise ratio, with FM RF signal levels. Totally different things. The maximum detected received stereo audio signal to noise ratio would apply even at the transmitter site, provided the radio was not totally overloaded by the strong RF.
The article uses typical popular FM stereo radios as examples, not the "out of adjustment out of standard" radios you claim.
The extended hybrid HD mode offers only -40 db signal to noise ratio below peak 100% modulation. Typically, the RMS average level is -10 to -25 db lower depending on audio processing settings of the FM station.
You are comparing apples and bananas, confusing FM RF field strength, with maximum stereo audio subcarrier signal to noise ratio as decoded by the radio's stereo decoder.
Typical analog stereo FM radios have an RF bandwidth of over 200 kHz, the total FM channel bandwidth.
You got it?
If you hate junk science so much, don't peddle it.

 

[clouseau]

You are confusing the typical receiver's detected FM stereo subcarrier analog audio signal to noise ratio, with FM RF signal levels. Totally different things. The maximum detected received stereo audio signal to noise ratio would apply even at the transmitter site, provided the radio was not totally overloaded by the strong RF.

I'm not confusing it, I'm using it to give a frame of reference. This article is using a bunch of Mumbo Jumbo numbers that the public doesn't understand to scare people and promote a false adgenda.

The article clearly states...

The digital sidebands occupy spectrum beyond the normal bandwidth of an analog FM signal.

Agreed.

However, they still lie outside the receive passband of the alternate FM channels 400 kHz away, which may be locally allocated. Although closer than the analog signal, at one two-hundredth the power the digital sidebands are not likely to interfere.

Correct

But they can easily interfere with reception of distant stations on the adjacent channels 200 kHz away since the signal spectra overlap. FCC rules do not protect nonlocal reception from interference.

That's right there is no protection for these stations. There shouldn't be. They're already very weak (Below 60 Dbu contour)

The article uses typical popular FM stereo radios as examples, not the "out of adjustment out of standard" radios you claim.

If you look at the Spectrum analyzer picture in the article, you can clearly see the digital sidebands or HD. They're at 129 to 198 kHz (Above and below) and are well outside the old allowable bandwidth of 100 KHZ. (Above and below) At a reduced power level they should not have an effect. Actually HD radiio isn't IN Band ON channel. It's In Band Adjacent channel. IOW, an old FM stereo station that broadcast on 93.1 actually had a bandwidth (or signal width ) of 200 KHz. (That's effectively .2 MHz for those of you in Rio Linda ) and would take up a "Range" on the dial from 93.0 MHz to 93.2 MHz. The part of this used for FM stereo is up to 53KHz so the same analog FM station uses essentially from 93.047 mhZ to 93.153 Mhz. An FM radio should never have received anything outside of this range EVER. The HD stuff is at 129 to 198 kHz. (or 92.902 MHz to 92.971 MHz on the low end and 93.249 MHz to 93.298 MHz on the high end.) Obviously the world is not perfect and we could use a little elbow room between the low power digntal and the High power analog. That's why we don't put FM stations every 2 MHz.

But there IS a good deal of space in there already. In fact so much there's other stuff in there as you know. Like RDS right next to stereo at 59KHz SCA at 67 and 92 KHz. There's already subcarriers allowed ON THE SAME CHANNEL.

The extended hybrid HD mode offers only -40 db signal to noise ratio below peak 100% modulation. Typically, the RMS average level is -10 to -25 db lower depending on audio processing settings of the FM station.
You are comparing apples and bananas, confusing FM RF field strength, with maximum stereo audio subcarrier signal to noise ratio as decoded by the radio's stereo decoder.

I'm not confusing anything. I was pointing out that most people would respond with "Gasp! only 40 Db. That must be bad. BTW what's a Db?"

Fact is, -40 Db is a lot. I don't have a calculator handy, but in the range of .002% I would guess. (not a bad ratio for bad to good)

Typical analog stereo FM radios have an RF bandwidth of over 200 kHz, the total FM channel bandwidth.

This is a little overstated as there is nothing for them to receive beyond 53 KHz. Otherwise they'd get interference from SCA and RDS and... Well you get the idea... but even at 200 "WIDTH" that's one hundred Above and One hundred Below. NOT in the range of HD.

If HD interferes with Stereo then either the HD signal doesn't meet spec or the radio doesn't meet spec. The Old spec. If some fifth harmonic in the radio's decoder get annoyed because of a signal outside the original signal mask, (One example in the article), that's called substandard. Notice some people have reported first adjacent reception issues. (ie 93.1 - 93.3).. Where's the outcry about HD interference to the mains? You think Broadcasters want interference to the main channel? They're greedy and stupid, but not THAT greedy and stupid.

Clouseau

 

[pullitin]

hahahahaha! that was a joke right?? lol
thats funny because I dont notice ANY diffrents!!

Thanks to radioskeptic for the link.

Noise Levels
HD Radio self-noise levels depend on receiver IF bandwidth, detector type, postdetection filtering, and stereo-decoding method. Since the late 1970s, a standard receiver architecture has become common. It uses two wideband ceramic IF filters, a quadrature FM detector, no postdetection filter, and a squarewave stereo-decoder IC. For such systems HD Radio self-noise typically will be about 50 dB below the level of a maximum-deviation 1-kHz sine wave. The derivation of this figure, which involves both theoretical calculation and signal measurement in a typical tuner, is here. The great majority of home tuners and receivers made in the past thirty years should exhibit HD Radio self-noise levels within a few dB of this figure. The noise level can be up to 10 dB higher for an extended-hybrid HD Radio signal.

Because RMS audio levels for stations with conservative signal processing typically are 10 to 25 dB below peak deviation, HD Radio self-noise can sound much louder than the 50-dB S/N figure might suggest. In fact, I've found HD Radio signals to be unlistenable on every home tuner or receiver I've tried that lacked a special stereo decoder, narrow IF filter, or postdetection filter. Such features are not common.

The worst case mentioned for FM HD station transmitting HD2 extended-hybrid signal:
HD noise reduces analog FM stereo signal to noise to -40 db to about -15 to -30 db below RMS average modulation.
Now you know where the added buzz on the analog stereo signal of a station transmitting HD comes from. It is the HD signal that is causing the noise.
The FM station turning off the HD signal will get rid of the noise HD Radio adds to the stereo analog FM signal.
Here is the link:
http://users.tns.net/~bb/hdrsn.htm

 

[Tom Wells]

Using a Sansui TU-7700 as reference, I find no objectionable increase in analog reception noise.
This is 7 miles from downtown Chicago. None of the IBOC FMS have enough dead air to judge the noise floor.
The one station that pauses to take a breath, WFMT 98.7 does not run IBOC, and in fact pulled SCA after MAJOR listener complaints.
So maybe I have only the city-grade opinion.
IBOC definitely makes tuning mushy and blots out the previously clear windows.
Degrades,.... yes, by numbers and charts. By listening, I think "seriously" is too strong a description.
If capture ratio weren't so characteristic, this might be true.
I'm sure the perspective of this post must stem from a fringe listener, where the "mushiness" overrides a carrier.

In close listening with headphones, I find only a negligible detriment in FM stereo.

I think the sound of music is best in wideband hi-fi AM or AM stereo, and convert all listening at home to AM hifi regardless of source.
There is inherently less distortion in AM, because it is Easier to vary output vs frquency without introducing phase distortions.
Linearity of result is the bottom line, and the ear is VERY sensitive to altered phase and harmonic content.
You can add harmonics if they are even-order, as occur in nature, and the ear and brain say "OK".
You alter phase or introduce odd-order harmonics, even in low doses, you strain both ear and brain.

None of the local FM IBOC stations have anything I would tune in for.

 

[radioskeptic]

Tom Wells is correct when he says, “Linearity of result is the bottom line, and the ear is VERY sensitive to altered phase and harmonic content. You can add harmonics if they are even-order, as occur in nature, and the ear and brain say ‘OK’. You alter phase or introduce odd-order harmonics, even in low doses, you strain both ear and brain.”

But I have to take issue with what he said immediately before that: “There is inherently less distortion in AM, because it is Easier to vary output vs. frequency without introducing phase distortions.” (I assume he means it’s easier to vary the amplitude of a carrier than to vary its frequency.)

In fact, it’s possible to get excellent linearity in the relationship between frequency swing and input voltage either in direct FM (with a variable reactance in the oscillator circuit) or with phase modulation, a.k.a. “indirect FM.” So there's no difference in distortion.

But receivers are another matter. Diodes – tube and solid state alike – are inherently non-linear devices. A diode displays a very high resistance to a reverse voltage (until the reverse voltage reaches the diode’s breakdown threshold). But when the voltage passes the zero point (the same potential as the other terminal) and becomes a forward voltage, the forward resistance is relatively high, but decreases rapidly at first, then more and more slowly until it reaches about 20 % of the diode’s peak operating voltage, after which there's no appreciabe change in its resistance.

That’s of no consequence in FM, where two diodes, in either a discriminator or a ratio detector, are dealing with voltage levels that never drop too low (as they don’t in carefully designed circuits). But trouble, in the form of excessive second harmonic distortion, is inevitable in most AM circuits with simple diode envelope detectors when negative modulation peaks are greater than 70 percent. (This explains why the audio on some shortwave international broadcast stations often sounds better than that on the domestic AM band: the shortwave stations use much more conservative modulation levels.)

With careful design, this effect can be nearly eliminated. How? By running the diode into a very high load impedance. The Radiotron Designer’s Handbook had a circuit in the back that showed the cathode of a diode being directly coupled to the grid of a triode used as a cathode follower stage. That circuit, designed by Selsted and Smith, was claimed to have very low distortion at negative modulation levels upwards of 90 percent.

Of course, you couldn’t duplicate that circuit with a bipolar transistor, which is an inherently low-impedance device requiring forward bias, but you could do it with an FET, which, like a tube, is a high-impedance device requiring reverse bias.

I’ve never seen this circuit employed commercially (has anyone else?), though I have seen the next best thing, a triode alone as an infinite impedance detector, used in a least one ancient hi-fi AM tuner.

 

[Tom Wells]

Yes, the point I was making is that it is easier to amplitude modulate. The simplicity of the mode and the
"naturalness of such minimal yet effective encoding" still stand as a model of well applied engineering.

FM is easy to do, but harder to keep linear because the modulation is swinging a parameter in a device acting as a tuned circuit.
This creates a "slew rate" consideration where the widebandedness of the circuit in the oscillator circuit is less important than what it really needs, high acceleration of the osc control. This is precisely why TV used AM for the video. The video deficiency of FM modulation was sharpness. In audio it is a phase purity loss, just as happens with lots of negative feedback in early transistor amps.
Of course, it still sounds good, overall. It just has its own sound.

This need not happen in AM as low modulaton can occur in linear portions of a tubes operation, or high level, where the input modulation
to the PA is applied through the magic of magnetic hysteresis , or with newer Pulse Modulation, even.

There is (can be) a lot less "active" circuitry to affect the tonal result.
Of course it can sound as bad as anything when poorly done.
A lot of the radios with IC designs lately have poorly implemented AGC, which leads to AM peak distortions mentioned by radioskeptic.
The low input impedances of the following audio sections are part and parcel of the same problem.


It is clear that FM is much less subject to degradation of the analog host by the introdction of IBOC than AM, at least in strong signals.

 

[SUPERCASTER]

You are confusing the typical receiver's detected FM stereo subcarrier analog audio signal to noise ratio, with FM RF signal levels. Totally different things. The maximum detected received stereo audio signal to noise ratio would apply even at the transmitter site, provided the radio was not totally overloaded by the strong RF.

I'm not confusing it, I'm using it to give a frame of reference. This article is using a bunch of Mumbo Jumbo numbers that the public doesn't understand to scare people and promote a false adgenda.

The article clearly states...

The digital sidebands occupy spectrum beyond the normal bandwidth of an analog FM signal.

Agreed.

However, they still lie outside the receive passband of the alternate FM channels 400 kHz away, which may be locally allocated. Although closer than the analog signal, at one two-hundredth the power the digital sidebands are not likely to interfere.

Correct

But they can easily interfere with reception of distant stations on the adjacent channels 200 kHz away since the signal spectra overlap. FCC rules do not protect nonlocal reception from interference.

That's right there is no protection for these stations. There shouldn't be. They're already very weak (Below 60 Dbu contour)

The article uses typical popular FM stereo radios as examples, not the "out of adjustment out of standard" radios you claim.

If you look at the Spectrum analyzer picture in the article, you can clearly see the digital sidebands or HD. They're at 129 to 198 kHz (Above and below) and are well outside the old allowable bandwidth of 100 KHZ. (Above and below) At a reduced power level they should not have an effect. Actually HD radiio isn't IN Band ON channel. It's In Band Adjacent channel. IOW, an old FM stereo station that broadcast on 93.1 actually had a bandwidth (or signal width ) of 200 KHz. (That's effectively .2 MHz for those of you in Rio Linda ) and would take up a "Range" on the dial from 93.0 MHz to 93.2 MHz. The part of this used for FM stereo is up to 53KHz so the same analog FM station uses essentially from 93.047 mhZ to 93.153 Mhz. An FM radio should never have received anything outside of this range EVER. The HD stuff is at 129 to 198 kHz. (or 92.902 MHz to 92.971 MHz on the low end and 93.249 MHz to 93.298 MHz on the high end.) Obviously the world is not perfect and we could use a little elbow room between the low power digntal and the High power analog. That's why we don't put FM stations every 2 MHz.

But there IS a good deal of space in there already. In fact so much there's other stuff in there as you know. Like RDS right next to stereo at 59KHz SCA at 67 and 92 KHz. There's already subcarriers allowed ON THE SAME CHANNEL.

The extended hybrid HD mode offers only -40 db signal to noise ratio below peak 100% modulation. Typically, the RMS average level is -10 to -25 db lower depending on audio processing settings of the FM station.
You are comparing apples and bananas, confusing FM RF field strength, with maximum stereo audio subcarrier signal to noise ratio as decoded by the radio's stereo decoder.

I'm not confusing anything. I was pointing out that most people would respond with "Gasp! only 40 Db. That must be bad. BTW what's a Db?"

Fact is, -40 Db is a lot. I don't have a calculator handy, but in the range of .002% I would guess. (not a bad ratio for bad to good)

Typical analog stereo FM radios have an RF bandwidth of over 200 kHz, the total FM channel bandwidth.

This is a little overstated as there is nothing for them to receive beyond 53 KHz. Otherwise they'd get interference from SCA and RDS and... Well you get the idea... but even at 200 "WIDTH" that's one hundred Above and One hundred Below. NOT in the range of HD.

If HD interferes with Stereo then either the HD signal doesn't meet spec or the radio doesn't meet spec. The Old spec. If some fifth harmonic in the radio's decoder get annoyed because of a signal outside the original signal mask, (One example in the article), that's called substandard. Notice some people have reported first adjacent reception issues. (ie 93.1 - 93.3).. Where's the outcry about HD interference to the mains? You think Broadcasters want interference to the main channel? They're greedy and stupid, but not THAT greedy and stupid.

Clouseau

Once again, the article clearly states that an FM station transmitting HD Radio digital signals interferes with it's own analog stereo subcarrier as detected by typical consumer grade stereo FM radios.
This poor stereo analog audio performance by HD digital stations has nothing to do with signal strengths, protected contours, other FM stations, or misaligned FM stereo radios, but is largely the result of the stereo signal detection circuit used in hundreds of millions of typical consumer analog FM Radios.
No mumbo jumbo numbers here, and there have been several other studies showing similar results by top professional RF research engineers.
-40 db is not much of a stereo signal to noise ratio for audio, worse then a typical audio cassette without Dolby noise reduction.
Power and greed often trump intelligence, unfortunately that is the way of the world.