FCC Commissioner's Comments on AM

[Kmagrill]

Where the boat was really missed was when the NRSC standard was put into place. The radio stations were mandated to comply with the NRSC mask but the receiver manufacturers were not. If the NRSC standard would have been executed all the way through to the listener, the AM listener would be hearing programming that sounds almost as good as what we hear on our air monitors....chock full of 7-9khz high end which is a mere 1 octave away from FM quality. And those radios would be in every home by now as the NRSC mask was put in place in the mid 1990's.

From the receiver manufacturers' perspective, the two main issues with making widebanded AM receivers were noise and cost. At the time, it was more expensive to make a quality receiver than a poor one (this is, unfortunately not true for FM). Today, with DSP, it should actually be much easier to design a really good AM for about the same cost as a bad one. Second, they rightly pointed out that making a widebanded AM receiver means it's going to be a lot more prone to impulse noise which was already intense in most cities and has gotten worst since. If you're a 50kW station on the low end of the dial, you can probably be heard pretty well on a widebanded AM radio, but if you are a 500W station on 1220, you're going to start having noise problems 4 or 5 miles away from the transmitter.

My '86 Chevy had an AM stereo radio with a manual Wide/Narrow button. Narrow was about 3.5kHz and "wide" was about 7kHz. I ran tests on the radio and found that, although it was down by 10dB, there was still significant output even at 15kHz. When we switched in the NRSC mask, the radio sounded noticably duller. I ran A/B comparisons with NRSC on and off and everyone listening in wide mode chose the unmasked signal as the better sounding. After NRSC, the wide mode still sounded better than narrow, but it really never sounded anywhere close to as good as FM. It never was really FM quality even without NRSC. NRSC just made it worse. Wide mode certainly was far more prone to noise. I recall gritting my teeth driving under power lines.

So, receiver makers said there's no reason to make widebanded radios because people won't like them and there's no market incentive to spend the money. Given the physics of AM and its inherent limitations, I'm not sure that mandating widebanded AM radios would have helped very much, but we'll never know for sure.

 

[satech]

Second, they rightly pointed out that making a widebanded AM receiver means it's going to be a lot more prone to impulse noise which was already intense in most cities and has gotten worst since. If you're a 50kW station on the low end of the dial, you can probably be heard pretty well on a widebanded AM radio, but if you are a 500W station on 1220, you're going to start having noise problems 4 or 5 miles away from the transmitter.

The "AMAX" receiver standard promoted by the NAB and CEA required impulse noise blanking circuitry, except on portable battery powered radios. The AMAX standard required a frequency response of at least 50 Hz - 7.5 kHz, no more than 2.5% THD, wide/narrow bandwidth selection (either manually switchable or automatically variable depending on signal strength), 10 kHz adjacent carrier whistle reduction of at least 25 dB, and impulse noise blanking.

For battery powered portables, the requirements were relaxed to 50 Hz - 6.5 kHz response, < 5.0% THD, and no noise blanking required -- mostly so Sony could sell their cheap little $30 SRF-42 AM Stereo Walkman and have it be "AMAX Stereo" approved.

 

[Kmagrill]

Noise blanking helps some, but doesn't eliminate the problem.

With today's DSP, I'll bet it would be possible to design an economical AM radio chipset that would exceed AMAX standards and, most importantly, eliminate a lot of the impulse noise in the software. Possibly, decoding multiple digital systems like CAM-D, DRM and HD would be possible, too.

 

[KeithE4]

Noise blanking helps some, but doesn't eliminate the problem.

With today's DSP, I'll bet it would be possible to design an economical AM radio chipset that would exceed AMAX standards and, most importantly, eliminate a lot of the impulse noise in the software. Possibly, decoding multiple digital systems like CAM-D, DRM and HD would be possible, too.

It's possible now, but who would buy it? AM has comparatively few listeners, so why would any company take the time to develop a product that would generate very few sales? It might work in car radios, but I can't see it being practical or financially doable anywhere else.

 

[Kmagrill]

Cost certainly has been a major factor until now, but part of the 'magic' of software defined radios and DSP is that making a great radio doesn't have to cost more than a bad one. Most of the fuctionality is in the firmware.

 

[satech]

A lot of factory equipped car radios are already using DSP based tuners with AM noise blanking. The problem is, they're being lazy and designing them for the European MW band, where the stations are tightly packed on 9 kHz channels, and thus the audio cuts off like a cliff at 4.5 kHz.

And they probably don't want to open up the audio bandwidth any higher than 5 kHz, because then they'd get complaints about annoying hiss on IBOC stations. (The stock radio in my dad's 2002 VW Golf uses a 6 kHz audio bandwidth on AM, and the hiss on IBOC stations is very audible.)

So, at this point we have two choices: either damn the torpedoes and go full speed ahead with IBOC, or kill IBOC on AM and then do everything we can to try to save and promote analog AM. But at this point receiver manufacturers will refuse to make wider-bandwidth analog AM tuners as long as IBOC is still on the air.

 

[KeithE4]

Cost certainly has been a major factor until now, but part of the 'magic' of software defined radios and DSP is that making a great radio doesn't have to cost more than a bad one. Most of the fuctionality is in the firmware.

True, but somebody has to write and debug that code. Good programmers don't come cheap. Even free software like GNU Radio has to be modified.

 

[BobOnTheJob]

Impulse noise...as in the kind heard in automobile radios should not be there. If I start getting AM noise, quite often the problem is that the antenna is no longer grounded well at it's mounting point. If we can insist on Part 15 devices being regulated, is it too much to ask that the auto radio's operating environment be up to snuff as well? Far more people listen to AM in the car than at home where the Part 15 noise is concentrated. I respectfully don't think that a poorly grounded antenna or a defective spark plug wire is a valid reason for not having manufactured AMAX car radios. If these radios were commonplace, AM IBOC would have never seen the light of day as nearly 100% of car radios would hear AM HD's dirty little secret.

 

[Kmagrill]

Cost certainly has been a major factor until now, but part of the 'magic' of software defined radios and DSP is that making a great radio doesn't have to cost more than a bad one. Most of the fuctionality is in the firmware.

True, but somebody has to write and debug that code. Good programmers don't come cheap. Even free software like GNU Radio has to be modified.

Agreed. There are development costs, but when compared to selling millions of chipsets or radios, development isn't that bad.

A lot of factory equipped car radios are already using DSP based tuners with AM noise blanking. The problem is, they're being lazy and designing them for the European MW band, where the stations are tightly packed on 9 kHz channels, and thus the audio cuts off like a cliff at 4.5 kHz.

And they probably don't want to open up the audio bandwidth any higher than 5 kHz, because then they'd get complaints about annoying hiss on IBOC stations. (The stock radio in my dad's 2002 VW Golf uses a 6 kHz audio bandwidth on AM, and the hiss on IBOC stations is very audible.)

So, at this point we have two choices: either damn the torpedoes and go full speed ahead with IBOC, or kill IBOC on AM and then do everything we can to try to save and promote analog AM. But at this point receiver manufacturers will refuse to make wider-bandwidth analog AM tuners as long as IBOC is still on the air.

DSP filtering could possibly eliminate IBOC buzz where it exists. Also, since AM can have symetrical sidebands (or intentionally asymetrical ones), a DSP hybrid approach could eliminate buzz by comparing the sidebands. This creates issues for some of the AM cost control methods and for Kahn Powerside products, but there may also be other solutions not yet considered. Dynamically adjusting the audio bandwidth based on signal strength and/or interference would also be an option.

Impulse noise...as in the kind heard in automobile radios should not be there. If I start getting AM noise, quite often the problem is that the antenna is no longer grounded well at it's mounting point. If we can insist on Part 15 devices being regulated, is it too much to ask that the auto radio's operating environment be up to snuff as well? Far more people listen to AM in the car than at home where the Part 15 noise is concentrated. I respectfully don't think that a poorly grounded antenna or a defective spark plug wire is a valid reason for not having manufactured AMAX car radios. If these radios were commonplace, AM IBOC would have never seen the light of day as nearly 100% of car radios would hear AM HD's dirty little secret.

Noise, for whatever reason, can potentially be mitigated, though not necessarily eliminated. But the point is that today's technology starts to show signs that it could spawn a whole new breed of extremely powerful radios with major performance advances where AM could be close to its theoretical maximum perfomance, where analog AM could begin to rival FM. Of course, the same is true of FM, too. I presume that such a chipset would include DSP processing for the analog FM signal, thus improving its quality in poor signal conditions, too. These chipsets would be pretty marketable, I am guessing.

 

[kenglish]

Anyone at the NAB or FCC following this thread?

 

[VelvetR]

Anyone at the NAB or FCC following this thread?

Better question:

Anyone at the FCC following anything other than the buck?