Time for new Speed Limit? 150% AM

[amfmsw]

Is it time to revisit the 125% mod limit on AM? In the new century, is it time to raise the volume to fight the new noise floor? 150%+ ?

 

[kyscott]

Is it time to revisit the 125% mod limit on AM? In the new century, is it time to raise the volume to fight the new noise floor? 150%+ ?

I'd be willling to bet some folks are already doing 150%.

 

[Tom Wells]

You could do this, but there are a lot of radios out there that already have a hard time with +125%.

They block up and become spatter-y sounding. GM's Delphi car radios were pretty bad about this.

The carrier swing envelope fools the AGC time constant into weird audio distortions.

If the time constant were longer, or the AVC circuit better controlled, this would not be a problem.
I think all my radios would be OK with it except maybe my Sangean 803.

While we're fighting noise, let's pull the NRSC filters and sound like real radio.

 

[CW]

You could do this, but there are a lot of radios out there that already have a hard time with +125%.

They block up and become spatter-y sounding. GM's Delphi car radios were pretty bad about this.

The carrier swing envelope fools the AGC time constant into weird audio distortions.

If the time constant were longer, or the AVC circuit better controlled, this would not be a problem.
I think all my radios would be OK with it except maybe my Sangean 803.

While we're fighting noise, let's pull the NRSC filters and sound like real radio.

AMEN, have the RECEIVERS fixed instead......DSP, user selectable filtering (the consumer is a little more tech savy today than 25+ yrs ago), etc.......and run HI FI AM.....Back to 15....also preemphasis and INVOKE AMAX rcvr requirements!! (DSP though would help make a BIG difference...and the chips are cheap nowadays even to go in Walkmans and easier on the batteries than a IBUZ decoder will do)....
Its amazing that they try to "fix" AM by doing it at the transmitter....but the RECEIVERS have been the problem all along...

 

[OKCRadioGuy]

LOL! It's amazing how much trouble we go though and how much money we spend to go to audio that's not quite as good as good 'ol AM from a good AM radio. Personally I'd rather have a full-sounded AM without the NRSC filtering on a good radio any day over the crappy IBUZ coded audio.
Talk about re-inventing the wheel!

 

[littlejohn]

I'm more the other way. Sunset analog AM, and give each channel 20KHz or so in which to broadcast ones and zeroes. It would work fine.

 

[gunterm]

I'm not old enough to remember how it used to sound but based on what 10khz filtered audio sounds like through a test bed its no wonder music AM radio died! Who wants 5khz limited audio, yuck.

 

[bossbill]

I don't recall having a receiver problem back when there was no positive modulation limit. There of course is a gigantic problem if the carrier cuts off, which is the absolute limit on negative modulation. And that's usually the problem. Before the absolutely arbitrary 125% positive ruling, it was not uncommon for some stations to modulate in excess of 125%. We were doing 180% at one station on occasional peaks, but the tubes weren't lasting all that long.

The cool thing about AM is that higher modulation does improve coverage, unlike FM. Why FM broadcasters screw up the audio in an attempt to be louder is beyond me.

 

[rrsounds]

Why FM broadcasters screw up the audio in an attempt to be louder is beyond me.

Why indeed. I've read (and even written) a number of papers/presentations on this subject:

Up to a point, processing helps FM audio overcome background noise.

Beyond that point, it's anyone's guess. It seems to me that once the distortion is more noticeable than any potential noise, you're going backwards, not forwards. But that's just me.

As for AM, it's (once again, my opinion) that since detection is simple (wire-diode-headphone), the medium should continue to exist in some form, rather than be phased out altogether. It is ideal for dispersing emergency information over long distances (e.g., Katrina and WWL).

If the number of stations was reduced to, say 10% of those presently on the air, we could dispense with positive mod limits and bandwidth filtering, while also setting aside a portion of the MW spectrum for digital-only broadcasts. But as long as thousands of AM stations are shoe-horned together as they are, the status quo should probably prevail.

Kind Regards,
David

 

[BobOnTheJob]

While we're fighting noise, let's pull the NRSC filters and sound like real radio.

A M E N

 

[CW]

I don't recall having a receiver problem back when there was no positive modulation limit. There of course is a gigantic problem if the carrier cuts off, which is the absolute limit on negative modulation. And that's usually the problem. Before the absolutely arbitrary 125% positive ruling, it was not uncommon for some stations to modulate in excess of 125%. We were doing 180% at one station on occasional peaks, but the tubes weren't lasting all that long.

The cool thing about AM is that higher modulation does improve coverage, unlike FM. Why FM broadcasters screw up the audio in an attempt to be louder is beyond me.

IF done right, FM is not screwed up...but the loss of dynamic range on highly processed stations supposively improves S/N at the fringes, etc....and listeners in tests decided that louder was better....though the quality was not as good...Some formats just dont cut it with high processing (Classical, etc) but Rock, AC, etc almost demand it...problem is most of the material is already processed now on the CDs....so add that to a high processed FM signal and it sounds like garbage......PLUS the threat of listener fatigue (which Robert Orban has always cautioned against).

I have some FMs with Classic Rock, etc that sounded GREAT.....clean, clear and loud.....yet no distortion...then I worked at one station where they brought in a "golden ears" consultant for $1000/day...he played with the 8200 for 3 days....and it still sounded like crap...the PD asked me later to look at it...one thing I found was the input level of the Optimod was looking for +10 with a clip at +14!!! I changed that to a +4 with a clip of +18....then increased the AGC range and release time....plus some other settings...the PD stopped me later in the hall and gave me a BIG thumbs up...he liked what he heard....and so did I (and this was on a Hispanic format Playing with a spare 8200 at home years ago gave me some insight into what to do and how to make it louder without distortion...there are proper ways to do it...but the CONsultants and most PDs out there dont know what they are doing with the processing...all they know is MORE is louder.....

 

[mr_elbows]

Tom wrote:

"The carrier swing envelope fools the AGC time constant into weird audio distortions"

Are you talking 'carrier shift' ?

There is no reason to have carrier shift with modern transmitter design.

Gary

 

[Tom Wells]

Tom wrote:

"The carrier swing envelope fools the AGC time constant into weird audio distortions"

Are you talking 'carrier shift' ?

There is no reason to have carrier shift with modern transmitter design.

Gary

Not carrier frequency shift. I refer to rectified carrier DC voltage off the detector in the receiver.
If the voltages developed (audio plus carrier DC) exceed the dynamic range of the subsequent circuit, there is audio distortion
from overloading. It is all too common in cheaper modern digital design.

All has to do with each radio's individual design approach to AVC control (how many stages are controlled) and the time constants chosen.

If you really want your head to spin, investigate tube AVC action vs solid state.
There are some real advantages to the variable mu in tubes vs fixed hfe in transistors.
The net effect is much less intermodulation distortion in the presence of high AVC voltages for tube circuits.
I can't cite source, but know these are facts I've put away somewhere in the past.

 

[rfry]

Carrier shift sometimes refers to the regulation of the average amplitude of the carrier during amplitude modulation. The average value of a carrier amplitude-modulated up to +/- 100% by a sine wave is the same for all modulation levels.

Once the carrier is modulated above 100% positive while maintaining 100% negative swings then asymmetry is required, and the average value of the carrier no longer is constant.

All transmitters (even solid state) will show variations in average carrier value with modulation under these conditions.

And of course, changing the audio waveform as needed for these modulation conditions adds audio distortion at the output of the transmitter -- not just what is produced by a receiver.
//

 

[fm-engineer]

The first thing a good consultant would do is to recommend replacing the 8200 with a new 8500 or Omnia 6. What's out there now is light years ahead of the 8200. If you were charged $1,000 to tinker with an old 8200, you were taken for a ride. People like that hurt the legitimacy of our business. As for your AM, I've seen a few 160% and 170% rigs, but never 180. That must have smoked! One common thing I have found here in the northeast is old antenna systems and ATU's failing when new transmitters are installed. They try to modulate 150 % out of the box, and very quickly the weak links are exposed. Also, most of the stations I have found above 125 are also shutting off the carrier with the negative side.

As

I don't recall having a receiver problem back when there was no positive modulation limit. There of course is a gigantic problem if the carrier cuts off, which is the absolute limit on negative modulation. And that's usually the problem. Before the absolutely arbitrary 125% positive ruling, it was not uncommon for some stations to modulate in excess of 125%. We were doing 180% at one station on occasional peaks, but the tubes weren't lasting all that long.

The cool thing about AM is that higher modulation does improve coverage, unlike FM. Why FM broadcasters screw up the audio in an attempt to be louder is beyond me.

IF done right, FM is not screwed up...but the loss of dynamic range on highly processed stations supposively improves S/N at the fringes, etc....and listeners in tests decided that louder was better....though the quality was not as good...Some formats just dont cut it with high processing (Classical, etc) but Rock, AC, etc almost demand it...problem is most of the material is already processed now on the CDs....so add that to a high processed FM signal and it sounds like garbage......PLUS the threat of listener fatigue (which Robert Orban has always cautioned against).

I have some FMs with Classic Rock, etc that sounded GREAT.....clean, clear and loud.....yet no distortion...then I worked at one station where they brought in a "golden ears" consultant for $1000/day...he played with the 8200 for 3 days....and it still sounded like crap...the PD asked me later to look at it...one thing I found was the input level of the Optimod was looking for +10 with a clip at +14!!! I changed that to a +4 with a clip of +18....then increased the AGC range and release time....plus some other settings...the PD stopped me later in the hall and gave me a BIG thumbs up...he liked what he heard....and so did I (and this was on a Hispanic format Playing with a spare 8200 at home years ago gave me some insight into what to do and how to make it louder without distortion...there are proper ways to do it...but the CONsultants and most PDs out there dont know what they are doing with the processing...all they know is MORE is louder.....

 

[CW]

The first thing a good consultant would do is to recommend replacing the 8200 with a new 8500 or Omnia 6. What's out there now is light years ahead of the 8200. If you were charged $1,000 to tinker with an old 8200, you were taken for a ride.

True, I agree...but this was 5 years ago and I wasnt the one making the buying decisions. At the time, all the stations had were 8200s....(one with V2 software in it!!)...now IF I could have changed it, I would have to an 8400 or later Omnia...and you are right, the CONsultant should have suggested that...but his "golden ears" (especially at the age he was at THEN) werent so golden (except in the price he charged)........what he did on the 8200 made no sense either...

 

[LA_Guy]

The 125% limit was put in place in the '70s to reduce splatter. Prior to that limit, there was NO limit on AM positive modulation, and some stations modulated to 200% or more positive.

NRSC-2's Mask and 10 kHz audio filter has made this rule redundant and obsolete-just like many rules and laws on still on the books that need to be removed.

 

[rorban]

While we're fighting noise, let's pull the NRSC filters and sound like real radio.

A M E N

Pre-NRSC, "real radio" at night was plagued by splatter onto second-adjacents, caused by audio modulation from 10 to 15 kHz that hardly one in 10,000 radios tuned to the interference generator could in fact reproduce. We were transmitting 15 kHz audio into 3 kHz radios, which receiver manufacturers made narrower and narrower as recording styles changed and energy about 10 kHz increased. I'm sure anyone old enough to remember pre-NRSC radio remembers the "cch...cch...cch...cch..." interference caused by a rock-formatted second-adjacent playing a cymbal-heavy song.

I recall being the one at the NRSC who originally suggested the 10 kHz filter as a compromise between broadcasters who wanted to retain full bandwidth and receiver manufacturers who were heavily lobbying for a 5 kHz filter to kill first-adjacent interference. I still believe that the 10 kHz filter was a good idea during the day and, moreover, that broadcasts should be limited to 5 kHz at night, which would kill first-adjacent skywave interference once and for all.

BTW, the latest round of NRSC tests determined that listeners preferred 6 or 7 kHz transmitted bandwidth when listening to an average bandwidth radio. 5 kHz caused noticeably fidelity loss on such a radio, while 10 kHz sounded worse because of additional audio-processor-induced in-band IM clipping distortion. The radio filtered out the desired 7 - 10 Khz audio, but reproduced the in-band IM just fine.

The moral: don't transmit a wider bandwidth than consumer radios can reproduce. All it does is causes unnecessary interference and increases distortion in the radios.

Bob Orban

 

[Tom Wells]

OK, I'd be satisfied with 10khz response, if the 6 to 10 khz range had enough boost to "fake" the presence of even higher audio.

I well remember ch-ch-ch. It only seemed an issue if I were dx-ing, where I expected it.
And back in those days, almost all radios were continuous tuning. Every radio could tune so the detector would "see"
the full audio, but of course not all radios could produce the full 15 kc. But 90% of them could reproduce 10khz,
cause you heard the whistle on AM at night real well. Since 90% of the radios could hear the high frequencies,
I seriously think that more than 1 in 10,000 listeners was hearing the full high frequency as almost everyone tuned until it sounded
the "clearest" and with many radios this was a bit off-center, as it included the high frequency info in better proportion.
To the extent the passband included the "whole" audio passband when slightly side-tuned, and the center tuned example
discriminates against the sidebands, the proper, center-tuned example exhibits frequency distortion by definition.

It was an accepted distortion for shortwave listening, amatuer and communications where integrity of comms was of
utmost importance. In CW work, a 100 hz filter is incredible. But we're talking about music and voice on the MW broadcast band here.
No comsumers heard AM smashed flat until the GM Delco radios led a generation to believe that AM has no consonants
or sibilants to offer.

Very intersting that the 6-7 khz compromise resulted from having to accomodate the weakness of "an additional audio processor".
Were there any tests using radios that did not suffer from intermodulation distortion?
Why were the faulty designs produced? Was the clipping really in a processor in the radio or a badly designed AGC circuit?
Poor IM distortion performance is one of the finest weaknesses of semiconductors in RF receiver AVC design.
You just don't have this problem when you apply AVC to control gain in vacuum tubes.
There's a whole lot about this I've forgotten...had something to do with third-order intercept points...
It is possible to do though, I have quite a few transistor radios that have no problem with this, but they're all pre-1985.


Why were they chosen as references for the study? I could have as easily recommended a raft of wideband models to test.
I'll even throw in a free 10 khz -42 db passive notch filter so you don't have to hear the squeal on dx night signals.

I really hate to see things move to lower quality.
I cannot be convinced otherwise on this one, my pt 15 proves it continously for me.
When I check the signal driving the neighborhood, I am VERY pleased with the sound of my full unlimited hi-end compared
to commercial stations, and since my low-level modulator cannot cut off conduction completely, I have no real splatter worries.

 

[amfmsw]

rob, with all due respect, "Manufacturers who were heavily lobbying for a 5kc filter"... really? To hell with them. We, AM operators, were lured to the 9Kc (10 Kc is way to generous, and we all know it) filter with the promise of pre-emphasis, Stereo, and AMAX standards for receivers. Well, we got the first one, which made the crap GM radios clip. Then no "standard" was set for stereo, and no manufacturer was forced to convert to the new AMAX standards NRSC set. No noise blanking, no auto bandwidth, no stereo standard.

Didn't we force broadcasters and manufacturers to adopt Zenith FM stereo? RCA compatable color? Ask Bill Paley. Why not mandate the manufacturers to support NRSC AM standards?

The NRSC set the standards, but only for us. It was a one-way street. The AM operators, especially independants, saw his investment, already ailing, evaporate. I invite you to come with me on sales calls and defer the statement "No one listens to AM anymore, it sounds terrible". Who wants to listen to MUD? Even speech.

It's 2008, not 1938. Why must we be forced to regress to that? Why not remove all of our Optimod AM units and just throw our CBS Volumax or Gates Sta-Level back into the rack...no one can hear the difference. Respectfully, you've got it backwards.