Enabling NRSC pre-emphasis by any means available is pretty much a requirement to match the de-emphsis that is supposed to be in place in the consumer radios. Depending on whether the peak limiter is before or after the pre-emphasis is important to know. In the AMT5000, the internal pre-emphasis is located after the peak limiter. This means when you enable pre-emphasis, the limiter (MODULATION control) must be adjusted downward somewhat to ensure the high-frequency peaks (after pre-emphasis) don't exceed 100% modulation (by much anyway) to prevent over modulation distortion on the boosted high frequencies.
The Q of the AMT5000 output is such that the 3 dB audio bandwidth is about 11 kHz assuming a "typical" antenna ground resistance of 30 ohms. A lower ground resistance will increase the Q and a higher ground resistance will decrease the Q.
The pre-emphasized high audio frequencies are filtered out by the output Q, so there is no direct need to reduce the bandwidth of the audio to match the RF output bandwidth, but it is important to not over modulate the transmitter at high audio frequencies. High frequency over modulation is not easily perceptible. It usually manifests itself as distortion of the "S" sounds.
It may seem counter productive to have to reduce the modulation level of lower frequencies to achieve the NRSC boost curve. But, that's the price paid for good high frequency fidelity. When a broadcast station decides to reduce its audio source bandwidth to say 5 kHz, you can see on Rich's NRSC curve in post #21 above that the pre-emphasis only needs to be 7 dB instead of 10 dB, so the lower frequencies are brought up 3 dB relative to the 100% modulation limit.
Step 1. Manufacturers reduce the high-frequency audio.response at the outputs of AM radio receivers to eliminate or reduce nighttime interference by channels adjacent to the one the radio is receiving -- the purpose being to eliminate buyer complaints and (probably) to reduce the cost of manufacturing those radios.
Step 2. Broadcast stations and industry groups recognize the loss of fidelity in AM reception as a result of step 1, and develop NRSC pre-emphasis in an attempt to overcome it.
But narrowband receivers now have more interference to their audio in the band from 20 Hz to 5 kHz from AM stations on adjacent channels, when all of them are using the NRSC curve. The (freehand) illustration below shows why this is true.
NRSC is useful in improving the fidelity of AM stations in narrowband receivers tuned to stations having relatively no interference from adjacent channels. However that improvement is limited to areas close to the transmitter where field strength is very high compared to the signals of adjacent channel stations.
Slightly reduced modulation level of the lower audio frequencies to avoid overmodulation of the pre-emphasized high-end is less of a problem because the lower frequency region is less susceptable to noise.
But we wonder whether there is reason to chop off or roll-off part of the low end.
For example, according to Dave Moulton's audio spectrum chart (TV Technology 4/1/10) the 1st octave of sound, 31.25 to 62.5Hz, contains little musical content, zero voice content, and is not reproduced by most loudspeakers. Perhaps this octave should be removed so it cannot waste unheard energy.
Note in the graphic I posted in Reply 33 that it is the audio producing the highest modulation percentages near ~9.5 kHz on the adjacent channels that produces low-frequency audio interference on the desired station.
The receiver demodulates interfering signals with respect to the desired carrier. Interfering signals produced by 9.5 kHz modulation of the adjacent channel signals are located 500 Hz from the desired carrier in the desired channel, and therefore produce an interfering audio frequency of 500 Hz in the receiver output -- which is very audible even in inexpensive AM receivers.
This is all irrespective of whatever low-end cutoff is used to modulate the desired station.
The fine arts of AM radio hold many secrets.
I'm glad you repeated the business about LF artifacts being produced by way of signal mixing under the stated circumstances. I'd missed that point on first reading.
Given that situation, it appears that low frequency audio below the range of reproduction (by the radio) does not cause any problem in itself.
It's probably not very scientific/accurate but there was an idea advanced on Page 15 of The Low Power AM Handbook about using the WinAmp and other equalizers to approach boosting the highs.
Druid Hills Radio AM-1710- Dade City, FL. Unlicensed operation authorized by the Part 15 Department of the FCC and our Resident Hobby Agent.
I pretty much go in expecting nothing.
Every now and then something pops up.
On the 6 local channels all together, I've
gotten 32 stations. My best distance so
far is 399 miles.
If something REALLY interesting happens,
I'll let you guys know.
NOISE AND STATIC RADIO
Using the EQ in Winamp is very handy, and indeed could provide a quick trick for AM pre-emphasis.
I have been using the DSP Winamp Plugin called StereoTool, which has a low and high pass set of filters, and can be spectrum shaped with EQ settings.
My new Spring project will be designing a High Q AM antenna that is self filtering for around 6 or 7kHz, which is an overshoot just half-way into an adjacent channel.
My new Spring project will be designing a High Q AM antenna that is self filtering for around 6 or 7kHz, which is an overshoot just half-way into an adjacent channel
Make it narrow in diameter.
Yes, Dade City, but how narrow?
I am going to start with a hair thin wire antenna. What kind of bandwidth will that be?
Wow, that's some good reading, to keep us busy for the next month.
See you sometime in March!
I'm a bit late to this thread, but have some comments regarding bandwidth/frequency response.
As has already been noted, there is nothing inherent in the process of amplitude modulation to limit bandwidth. Current FCC regulations set the upper frequency limit for audio content on the AM broadcast band at 10KHz. But, and this is a big BUT, that has not always been the case, and in fact that rule does not apply to us under Part 15.
In days of yore, when radio began, the state of technology was the limiting factor in fidelity. By the 30's or so, microphones and speakers were the limiting factor in frequency response. By the mid-late 50's true HiFi was coming of age, and 50-10K response was common, and 30-15K was attainable. By the time the 70's had rolled around, the 'gold standard' for frequency response in the audiophile world had settled as 20Hz-20KHz, which is also assumed to be the extreme limits of human hearing.
Anyhow.... at some point in all this, probably in the 50's or early 60's, the FCC designated 15KHz as the limit for AM Broadcast. A tad short of 'audiophile" quality, but nonetheless still excellent. By the late 60's or early 70's, stations had begun to discover the "benefits" of simple compression and equalization in attempts to make their station sound "louder" or "better" than the competition. That practice has, of course, continued to this day, with ever-more-complex audio processing being applied for virtually all broadcast audio (am, fm, digital, analog, tv, internet, you-name-it).
At some point, and without trying to research it I'll again take an educated guess and say perhaps late 80's, and maybe coincident with the appearance of C-QUAM stereo, the FCC cut the bandwidth on AM BCB down to its present 10KHz.
But we, operating under Part 15, are not subject to that, and therefore can offer (to those listeners who can receive and appreciate it) a wideband 15K or even 20K signal, provided our transmitters are capable. And most transmitters sold for Part 15 use are indeed capable of at least 15K or better, as they do not employ any steep or "brick wall" filtering in the audio path.
One exception to that would be Radio Systems and LPB transmitters that may have been used in as TIS stations: If you repurpose one for Part 15 use (carrier-current, most likely) that was previously used for TIS, you'll want to bypass the "TIS Filter" which limits bandwidth down to 3KHz. I'm not sure if this filter is there for practicality or if there is an FCC rule limiting TIS stations to 3K audio bandwidth, but you'll definitely want to bypass that filter (if present) unless your station is to be talk-only.
OK... 'nuff rambling on my part.
No shirt, no shoes, no pants, no gods -- no worries!
At the present time the rules for TIS (Traffic Information Stations) do indeed specify an audio bandpass limit of 3kHz.
No reason is officially stated in the rules, but insiders believe the NAB (National Association of Broadcasters), in one of their smaller efforts to protect their corporate members, used their lobbying power to restrict the TIS frequency response as a means of making such stations "hard on the ear" and thus of no possible competition to the Membership, who in anycase sabotage themselves by their poor programming, loud and clear as it may sound.
MRAM 1500 streams the TIS station from his Ohio town, and it's actually better than 95% of the stations on my home dial.
Our TIS station down the street was
installed several years ago, and I heard
a few things about it, which I think are odd.
The installer starts out with 10 watts and whatever
transmitting antenna they use. 10 or 20 feet long?
I don't know. But then the field strength has to
be brought down to meet FCC reguirements,
(I forgot the specs). And then the station isn't strong
enough in certain parts of our town (in my opinion.)
Also it's running an audio loop that hasn't been changed
since Novermber. But I happen to know that the
person who records it has to do about a zillion other
things, and this is just one more thing for him to do.
Bruce, DOGRADIO, you go over to the licensees of your TIS station and tell them you will take over the business of updating the message and caring for the station and tell them you won't charge much.
I'd say, $5000 a year in one annual installment.
An interesting thing one can do with TIS is low frequency roll-off and still maintain the 3KHZ bandwidth requirement. For example, roll off the low at say 300 hZ. The 3kHz BW high end is now 3300 hz, thus improving the quality of sound.
That's the way to do it!
Sec. 90.242 Travelers' information stations.--(8) Each transmitter in a Travelers Information Station shall be equipped with an audio low-pass filter. Such filter shall be installedbetween the modulation limiter and the modulated stage. At audiofrequencies between 3 kHz and 20 kHz this filter shall have anattenuation greater than the attenuation at 1 kHz by at least:
60 log10 (f/3) decibels.
where ''f'' is the audio frequency in kHz. At audio frequencies above 20kHz, the attenuation shall be at least 50 decibels greater than theattenuation at 1 kHz.
Messengers who deliver bad news could have been undertakers but sometimes pursue radio as the place to ruin it for people in what little happiness they might accidentally have.
At least we get to be clever and smart until we get that message of gloom.
Musn't weep, or the text will start running off the page.
Quotes from the FCC rules can be discouraging but necessary to dispel misinformation.
by MRAM 1500
Charter Member - Association of Low Power Broadcasters
Chairman - ALPB
Services for the burial of failed ideas are delayed because of terrible weather. Full coverage on your local Part 15 station.
I was just thinking out of the box based on a tech note from TenTec that reads as follows:
Description of Jupiter Transmit Roll-Off ControlThe TX Roll-Off control was added to the user menu in version 1.17. This control allows the user to position the lower edge of the selected transit filter. The lower edge setting affects thetransmitted audio range. The control range is 0 to 1270 Hz with a default value of 200 Hz.When adjusting this feature it important to remember transmit bandwidth and TX Roll-Off together determine the audio quality. The TX Roll-Off determines where audio starts on the lower side and transmit filter selection determines the passband. For example: if you have a 3 kHz filter selected and set a 400 Hz TX Roll-Off the transmitted audio would be 400 Hz to 3400 Hz. The large control range will allow digital mode users to place narrow filters at any position theychoose. A large TX Roll-Off with a wide filter is not possible as other factors limit the usablerange. Likewise, the low frequency audio response is affected by hardware. So there is some inherent roll-off of frequencies below 150 Hz.
While trying to find a post where Rich recently displayed audio bandwidth figures for available AM receivers, I chanced upon this highly informative thread from earlier in the year,
Being reminded of the NRSC Preemphasis capability built into every SSTran transmitter, tests will begin immediately to employ this feature as per the opening post by PhilB.
For several years our transmitters have not employed this feature, but it's there, and I'll tell you what results.
Excuse me, I'm out here in the back adding Jumper S22 on the AMT5000.
EDITOR -- Correction... PhilB's comments appear later down the list, not Post # 1.
NCRS Jumper in place, the FCC approved AM pre-emphasis is turned on, and I'll tell you why I like it.
But first, some preliminaries.
My radio playlist is processed with Hans Van Zutphen's "Stereo Tools". a highly professional software that allows setting of just about every parameter of an audio path, including the ability to set the audio bandpass.
My encoding system feeds two different types of transmitters: AM radio transmitters and internet streaming encoder/server, a form of internet transmitter.
My previous bandpass setting favored the AM side of the operation, being set for 80 to 10,240 Hz, because that is a full seven-octaves of sound.
But with the NCRS PreEmphasis turned on, and according to remarks previously made by PhilB, there is no need for equalizers or audio bandpass filters, since the AM system filters its own audio, considering the loading coil, antenna and receiver as having their own bandwidth characteristics.
Given the freedom to do so, I've reset the "Stereotools" processor for a bandwidth favoring the outgoing stream server, which itself is fixed at 22,050 Hz sampling rate, stereo.
Thus the new "Stereotools" bandpass is 50% of that: 80 to 11,025 Hz.
Now I will tell why I like the sound of the NCRS PreEmphasis.
The "sharper" sound from radios, there is one in every room, "projects" farther into the room space and is much more listenable at distances away from the receivers.
The "flat" ordinary non-NRSC sound is fine while near the radio, but becomes "duller" with greater distance from the loudspeaker.
While this is, of course, a subjective evaluation, we will leave it to every future operator who tries NRSC PreEmphasis in his/her SSTran transmitter to arrive at their own conclusion.
In general, NCRS sound should improve highway listening in automobiles, where road noise often makes radio listening difficult. This same rule holds true in garden patio settings, where neighborhood powertools, barking dogs and aircraft above can compete with radio listening.