The NRSC filter and Ham radio

[KA3EKH]

I have been giving some thought to installing a NRSC filter in the audio chain of my RCA. The filters I have give about a 50 dB cut of any audio above 10 kHz but have to wonder if I am wasting my time doing this. I know in theory the old RCA is flat out to 20 kHz and maybe my old Volumax and mixer should be their but how much energy can there be in a voice channel out above 6 or7 kHz anyway? Happen to have a couple old NRSC filters left laying around from the nineties and the old processing equipment before  replacing all that stuff with  digital optimods at a couple commercial AM stations. The digital optimods have NRSC built in so now have some spares to play with. Anyone else using NRSC filters or anything more advance than an equalizer?

[Mike/W8BAC]

Hi Ray,

Are you saying your RCA has an rf bandwidth of 40 KHz? The NRSC 10 KHz filter will do a better job in that case bringing you back to 20 KHz but I think it might be a bit wide still. Something in the neighborhood 5 to 8 KHz might be best for keeping the piece. I'm working on that myself.

Depending on the processing, compression and expansion, distortion and other factors you can have quite a bit of energy above 7 KHz. When the NRSC filters came out some broadcasters had been pushing audio out to 15 KHZ and beyond.  Best of luck.

Mike

[KA3EKH]

I don’t know, I may not be the sharpest tool in the shed but back in the old days when I had to do profs on AM transmitters we would do audio response from 20 Hz to 20kHz at 25, 50 and 100% modulation so I always thought the response for a broadcast transmitter would be relatively flat to 20 kHz, thinking in terms of sidebands imagine that may be 40 kHz wide signal. If I can get the bandwidth down to 6 or 8 kHz I would be happy, I always use the 8 kHz filter on the receiver. I have not done a full sweep of the transmitter after moving it to 1.885 so have to wonder if my tank and output are that wide to begin with but imagine this is the problem when using a old broadcast transmitter. Not uncommon to hear someone as close as 1.880 or a slop bucket station as close as 1.890 above and want to live in the 10 kHz channel so wonder if that means I have to limit audio response down to 5 kHz?

[K5UJ]

It's ham radio--no music just voice.  5 kc audio bandwidth is FB.  The NRSC mask rolls everything off around 9.5 kc.  The human ear mostly can't hear much high frequency anyway.  I bet most people can't hear a 9000 hz tone.  But anyway, an easy quick fix if all else fails is to get yourself a couple of relatively cheap dbx equalizers and chain them together in series.  They usually have a +-15 dB boost/cut.  move the sliders on all the bands to +15 dB and adjust your levels on the other air chain gear into the rig to achieve your usual modulation level.  Then take all the sliders above 5 kc and drop them to -15 dB.  You now have a 60 dB cutoff at 5 KHz.   This probably isn't a beautiful music hi fi solution but as I said earlier this is ham voice AM.  It will sound okay.   If you run any kind of broadcast peak limiter I'd try putting the eqs in the chain just before the limiter.  If you don't do any compression or limiting I'd put them as the last thing in the air chain just before the rig.

rob

[k4kyv]

But anyway, an easy quick fix if all else fails is to get yourself a couple of relatively cheap dbx equalizers and chain them together in series.  They usually have a +-15 dB boost/cut.  move the sliders on all the bands to +15 dB and adjust your levels on the other air chain gear into the rig to achieve your usual modulation level.  Then take all the sliders above 5 kc and drop them to -15 dB.  You now have a 60 dB cutoff at 5 KHz.   This probably isn't a beautiful music hi fi solution but as I said earlier this is ham voice AM.  It will sound okay.   If you run any kind of broadcast peak limiter I'd try putting the eqs in the chain just before the limiter.  If you don't do any compression or limiting I'd put them as the last thing in the air chain just before the rig.

If it's a stereo unit, it should be easy enough to take the output from one channel, run it through an attenuator cobbled together with a few 1/4w or 1/2w resistors, and feed the output into the input of the other channel, thus giving you double the boost or attenuation in the response curve.

The manufacturers' specs on broadcast audio equipment usually say +/- a dB or so from 20~ to 20K or at  least 50-20K, but it has been my experience with transformer coupled equipment that the response often becomes ragged, with peaks and dips, as well as odd phase shifts, above about 10K. And this is with the most expensive, highly reputed audio transformers such as the UTC LS series or equivalent.

I ran a pre-installation test with my n.o.s. UTC LS-49 class B driver transformer by connecting it to a scope, with the midtap of the secondary grounded, feeding one leg of the winding to the horizontal input, and the other leg to the vertical input.  With the gain adjusted for equal deflection in each channel, the display on the scope should be a straight line at a 45° angle to the base line.  That's what I got, up to about 11K.  Above that frequency, I began to see the line separate into a narrow ellipse that varied with frequency, indicating that one side of the winding was slightly out of phase with the other. In a push-pull amplifier this will result in distortion, and if severe enough, add bandwidth the signal.

Some of my older WW2 and pre-War era LS series and similar transformers begin to show deviation from flat response above about 5 kc/s.  Don't know if this is some kind of deterioration (can't think of what could happen to a transformer that would alter the frequency response if the windings are not open or with shorted turns).  I'm wondering if the manufacturers gave them an inflated nominal response rating knowing that very few users would notice the difference.

I had someone check my signal the other night, using a spectrum analyser connected to the i.f. output of his receiver. With my brick-wall 3400~ sharp cut-off low pass filter in line, he said my signal was about 7 kc/s wide, and the brick-wall high frequency cutoff was obvious, although it showed up a little higher than what my audio response checks indicate it should be. Then I switched in the filter I normally use most of the time, which has a nominal 6 kc/s cutoff, but which actually begins to roll off at 5K, with nothing visible on the scope beyond 7.5K.  He said his receiver display showed just that.  Then I cut out the filters altogether, and he said the signal extended out slightly wider before the roll-off became visible, maybe closer to 6K, but that everything was still pretty much gone at 7.5K, with nothing visible above about 8K.  Running a response check with a signal generator, my transmitter tests flat out beyond 12k.  I suspect the microphone may be the limiting factor; I was using a D-104 with an additional high frequency boost of about 9 dB above about 2 kc/s.  But if the mic doesn't respond above 7.5 K, a boost in the response curve in the pre-amp will not bring the signal up from zero. With all the low-pass filtering gone my signal was not "40 kc/s wide" as some slopbucketeers insist about every AM sigal they hear.

[flintstone mop]

WOW Ray
That was long ago to proof a TX to 20kc. There was a myth that a properly adjusted AM transmitter could sound as good as FM.
That would be good for a 50KW station wishing to serve a community about 15 miles from the transmitter.
HD radio took all of that away and with today's crappy auto radios and stereo receivers, you're lucky to get an I.F. to pass more than 5kc.

But I agree with Don and Rob. Just some simple EQ boxes. I bet you have some nice processing gear hanging around from your engineering dayz, that you glommed on to. So stick in that Orban box you have and the NRSC and it will be FB OM

Fred

[W7TFO]

On transformers, specifically those by United in the "linear standard" series, lots of them won't proof up to the published specs.

Sometimes I found the only difference between the Special series, Commercial grade, and the LS is the case they are potted in.

Old Doc Hoyer back in Wisconsin has wound transformers for almost 80 years now, and he agrees wholeheartedly with this.  Same diatribe on lots of Thordarsons, Western Electric, SNC and Thermador.

Higher marks across the board for Triad, Stancor, and Merit.

YMMV.

73DG

[W8IXY]

The AM broadcast NRSC filter specifies an audio cutoff that must be at least 25db down at 10.2 kHz.  Plus, the AM broadcast NRSC "curve" specifies a 75 microsecond pre-emphasis applied to the audio as well.  That 75 microsecond pre-emphasis curve results in approximately a +10db boost of audio in the 10 kHz range.  So, if you apply 10 db of boost at 10 kHz, you need a very steep cutoff at 10 kHz to achieve an apparent 35db cutoff point.   Most AM broadcast NRSC filters start the rolloff about 9.5 kHz, so they can be down > 25db at 10.2 kHz.

To make this work best in the real world, the TRANSMITTER should have as wide and flat of a frequency response (as well as low phase shift and low distortion) as possible, and all audio processing and filtering should be done in the audio processor before application of the audio to the transmitter.  Most of today's modern AM audio processors feature several different low pass filter slopes, from the widest @ 9.5 kHz, to as low as 4.5 kHz.   Also most shortwave broadcasters filter out all audio information above 5 kHz.  Those AM processors also include high pass filters, selectable from below 20 Hz to 60 Hz.  Those processors also usually include the option to switch in a "phase rotator" in the audio path to remove asymmetry in voice waveforms.

Over the years I have found that trying to minimize IM distortion as well as THD results in really good sounding on-air audio, if that transmitter is fed with a pristine input signal.

These days, I use a commercial broadcast audio processor with the high pass filter set at 60 Hz, and the low pass filter set at 4.5 kHz.  I also use a Symmetrix microphone preamp/processor to allow as much control of my audio as possible.  The radios include a K3, several older Kenwoods, a Valiant, and a DX100.  Limiting the frequency response to run audio only between 60Hz and 4.5 kHz allows those transmitters to not have to strain too much if power output is kept at our below their design specs.

73
Ted  W8IXY

[W2VW]

To make this work best in the real world, the TRANSMITTER should have as wide and flat of a frequency response (as well as low phase shift and low distortion) as possible, and all audio processing and filtering should be done in the audio processor before application of the audio to the transmitter. Ted  W8IXY

It's amazing to me how many people who seem be intelligent otherwise just don't seem to be able to grasp this concept.

[KA3EKH]

Yea, back in the old days when you had to d a yearly proof I would dread that thought of trying to get the stuff to work. There was a special flat output from the modulation monitor that was used to drive the distortion analyzer and at the end of the day you would waste two or three nights trying to get the transmitter to look good on paper and later it got that more the numbers would magically start to fall in line. These days the only thing I look at annually is the NRSC compliance and critical field points. The two AM commercial stations I have both are using Optimods digital processors that bring up another question. Corporate has us set all our AM processors to 5 kHz bandwidth. They are insistent this produces better coverage; I cannot see how reducing bandwidth for an AM station from 10 kHz to 5 kHz would make any difference in coverage, kind of like the myth that turning off the stereo pilot on a FM broadcast station provides more coverage. Something about their being more energy packed in a smaller space or something like that. At this point I am going to install the NRSC filter before the audio processor and if nothing else it will make me feel like I am doing something to restrict bandwidth.

[k4kyv]

That was long ago to proof a TX to 20kc. There was a myth that a properly adjusted AM transmitter could sound as good as FM.
That would be good for a 50KW station wishing to serve a community about 15 miles from the transmitter.
HD radio took all of that away and with today's crappy auto radios and stereo receivers, you're lucky to get an I.F. to pass more than 5kc.

Back in the late 50s, before there was such a glut of AM stations jammed and  crammed into the MW band, the old TRF receivers were said to be capable of better reception of local signals, than superhets. 

The only problem I could see with that is that many of those ancient receivers had a "square law" detector that produced high distortion.

The main problem with sky-wave reception is the envelope detector and  selective fading distortion.

Perhaps with a flat passband out to 30 kc/s, with good phase shift characteristics (which immediately excludes mechanical filters), combined with a synchronous detector, AM reception could be as good as monaural FM.

In practical terms, considering the hearing limitations of our ageing ham population,  the vagaries of ionospheric propagation and the limitations that exist on 99% of the receivers used for reception of amateur radio signals,  I don't see much point in transmitting audio beyond about 5 kc/s (on the ham bands).

But

Quote from W8IXY
To make this work best in the real world, the TRANSMITTER should have as wide and flat of a frequency response (as well as low phase shift and low distortion) as possible, and all audio processing and filtering should be done in the audio processor before application of the audio to the transmitter.

UTC, Thordarson and other transformer manufacturers recommend a flat response that extends at least one octave above and below the intended frequency range of the audio to be passed through an amplifier. Assuming voice frequencies going down as low as 100~, to maintain that  standard, the undistorted flat frequency response with minimum phase shift of the audio chain in the transmitter should extend at least from 50 to 10,000 cps.

Most "communications quality" audio equipment doesn't cut it.  That's why the space shuttle sound so dominates amateur radio phone signals, even on AM as well as SSB. Just because the modulation transformer measures flat down to 30~ doesn't mean that significant voice modulation components actually extend down that low, or that the transformer would even be suitable for music bass tones below about 60~.

Most "bassy" and "tinny" audio results from an unbalance in the response curve, not a frequency response that extends too high or too low. IOW, a bassy signal is usually a result of deficient high frequency response, not too much bass, and vice versa. If you get reports of a bassy signal, it is counterproductive to reduce the coupling capacitances between audio stages.  That just gives you even more of a "space shuttle" sound with telephone quality audio. Likewise, if your audio sounds tinny, the last thing you want to do is to add capacitors across the audio transformers, plate resistors and grid resistors to attenuate the highs.  The cure for most cases of tinny audio is to boost the bass, and for most bassy audio is to boost the highs.

As for splatter, the vast majority of cases is the result of distortion products, not the frequency response of the system, or even the bandwidth of the transmitted audio.  If you want to cut down the bandwidth of a clean signal by chopping off the highs, the place to do it is in the low level stages, as close to the microphone as possible.  Any low-pass filter in the audio chain for this purpose should cut off the high frequency response as sharply as possible.

A perfectly flat response out to some high frequency cut-off point is apt to sound dull and lifeless over the air.  In combination with a low pass filter (or equaliser adjustments), or a microphone with inherently limited high frequency response, the frequencies in the upper range just below the nominal cutoff should be boosted up to 9 or 10 dB or so, in order to add a pleasant degree of "presence" to the audio. If the highs are cut off sharply at 3 to 4 kc/s, this presence rise compensates for the lack of highs beyond the cut-off point, and to many listeners the audio still sounds "broadcast quality".

Actually, Astatic did a crude version of that with their old D-104 microphone with the peak in audio response somewhere around 2500 cps. That's why those old mics had such a reputation for adding "punch" to the audio, particularly since the response curve and distortion characteristics of most of the older ham transmitters was so atrocious.

Posted by: KA3EKH
Corporate has us set all our AM processors to 5 kHz bandwidth. They are insistent this produces better coverage; I cannot see how reducing bandwidth for an AM station from 10 kHz to 5 kHz would make any difference in coverage, kind of like the myth that turning off the stereo pilot on a FM broadcast station provides more coverage. Something about their being more energy packed in a smaller space or something like that. At this point I am going to install the NRSC filter before the audio processor and if nothing else it will make me feel like I am doing something to restrict bandwidth.

Although I had amateur radio audio in mind, maybe some of the above suggestions would work in your case, too.

[K5UJ]

<<Corporate has us set all our AM processors to 5 kHz bandwidth. They are insistent this produces better coverage; I cannot see how reducing bandwidth for an AM station from 10 kHz to 5 kHz would make any difference in coverage,>>

Sounds like CC.  If it is any comfort, most consumer AM bc rx these days don't even make it out 5 kc from zero beat.  I've heard a lot of them only have 6 kc passband.   I have one rx I regularly use that is so wide I get IBOC hiss in the audio.  The other two along with the car radio don't, which tells me they go out at most 5 Kc. 

I do have one extremely wide rx in the ham shack and when tuned to a strong analog station using the full NRSC standard it sounds wonderful but I'll bet the vast majority of consumers listening to AM don't have that.

[WA3VJB]

I'll bet the vast majority of consumers listening to AM don't have that.

Not that the word "vast" means a lot of people listening to AM>

[WB4AIO]

Yea, back in the old days when you had to d a yearly proof I would dread that thought of trying to get the stuff to work. There was a special flat output from the modulation monitor that was used to drive the distortion analyzer and at the end of the day you would waste two or three nights trying to get the transmitter to look good on paper and later it got that more the numbers would magically start to fall in line. These days the only thing I look at annually is the NRSC compliance and critical field points. The two AM commercial stations I have both are using Optimods digital processors that bring up another question. Corporate has us set all our AM processors to 5 kHz bandwidth. They are insistent this produces better coverage; I cannot see how reducing bandwidth for an AM station from 10 kHz to 5 kHz would make any difference in coverage, kind of like the myth that turning off the stereo pilot on a FM broadcast station provides more coverage. Something about their being more energy packed in a smaller space or something like that. At this point I am going to install the NRSC filter before the audio processor and if nothing else it will make me feel like I am doing something to restrict bandwidth.

When I was the chief engineer at WEAM in Arlington in the 70s and 80s, we also had the obligatory annual audio proof of performance -- the system, from console microphone preamp to modulation monitor output, had to be flat from 100 Hz to 5 kHz within a dB, and less than 2 per cent. distortion as I recall. The WEAM main (RCA BTA-10U2) and auxiliary (Gates BC-5E) transmitters easily met this standard, with less than 1 per cent. distortion and response from 30 to almost 15,000 Hz. As good as many FM stations, and, on my McKay-Dymek AM5 tuner, sounding just as good over the air.

It's sad to see the trashy, overprocessed audio on many broadcast stations today. The 5 kHz filter you mention will take an octave or more away from music and do little for coverage, despite what your management says.

A 5kHz filter would be a good idea for amateur AM in an active band, however.

Also, turning off the stereo pilot will indeed increase coverage for FM broadcast stations. Not so much because of the few per cent. gain in audio modulation, but because the listeners' stereo decoders will turn off, greatly increasing the signal to noise ratio in weak-signal areas. But reception will be monaural, so you might as well turn off the multiplex generator entirely and gain a few more dB to boot!

The whole "NRSC mask" idea was and is absurd, by the way. The interference problems on AM are the result of shoehorning in four times as many stations as there should be, not the result of excessive high frequency response. (The IBOC digital buzzsaws they do allow are far worse interference generators than the most boosted and clipped supermodulated highs ever were.) The reason the NRSC mask was forced on us was because both the Kahn and Motorola AM stereo systems could not meet spec at high frequencies, so their solution was to lobby to make it illegal to transmit at the frequencies where their systems broke down. Only in America.


With good wishes,


Kevin, WB4AIO.

[flintstone mop]

During my volunteer days at Georgetown University in Wash DC. We had an excellent engineer/hippy/underground radio guy that kept a Gates FM3A sounding better than the local stations, just using the Volumax/Audimax and AT&T landlines for the stereo audio.Homebrew audio console he and an associate built to FCC specs and sold to the university. The secret was very high quality magnetic phono preamps. Typical 'broadcast quality' boards had really crappy mag phono preamps or none at all.
And that was not a myth about turning off the stereo carrier. Maybe in today's receivers there is no difference but in the 70's ,,,,yes. He experimented to prove that by asking listeners who were desperately trying to hear the 'free speech' voice of WGTB in Delaware if they could hear any difference with the stereo turned off. And it was consistently......... yes...there was less hiss and more audio.
Engineer speak it was supposed to be 6dB. The killer for FM was, and still is that DAM pre-emphasis. To get around that, he installed a Dolby unit, which was allowed at that time for FM, and you could select 25 micro seconds instead of 75 and that allowed for a little more modulation.
In the commercial two-way bands there is another narrow-banding scheme coming and another lowering of modulation to accommodate more licensees. Digital is the the only answer to get back some fidelity to the received audio.
BTW the stations in the expanded band sound very good. There is more music in that part of the band. IF you can get through all of the hetrodynes and listen to the dominating station.

[k4kyv]

It's sad to see the trashy, overprocessed audio on many broadcast stations today. The 5 kHz filter you mention will take an octave or more away from music and do little for coverage, despite what your management says.

Considering the programming on >90% of the AM BC stations these days, I don't think it makes much difference; they could just as well be on slopbucket, if only the "consuming" public would be able to figure out how to tune it in. A few notable exceptions include WSM in Nashville, if you are a fan of their brand of music, or CJBC @ 860 kc in Toronto (jazz and classical), if you live close enough to pick them up clearly.

Some of my FM receivers have a stereo/mono switch, and when the signal is marginal I can usually get far better reception in monaural mode.  I haven't checked, but I'd guess most sets made these days don't even have a stereo/mono switch, though. I don't recall the exact figure, but I  seem to recall that the receiver sees something like a 20 dB reduction in ERP while in the stereo mode.

[Steve - K4HX]

Yes, it's a -23dB SNR for stereo versus mono, assuming the recived signal is above threshold. But that's only at the transmit end (the BC stations turns off the 19 k pilot and the 38 kHz L-R). There is an SNR improvement if the RX can switch to mono (when the BC station is still transmitting steroe), but not that much, something like 5 dB (15 kHz BW verus 53 kHz). Many receivers today automatically switch to mono when the carrier level drops below a certain level.

[W1DAN]

Hi:

NRSC pre-emphasis (generally any pre-emphasis that helps to overcome a receiver's IF downward slope) will help for ham AM, but the LPF in NRSC really does no good for ham AM and voice.

After seeing that Stu AB2EZ modified a SCAF receive audio filter unit for use as transmission filtering, I built an 8 pole elliptical low pass filter using a Maxim MAX294 switched capacitor filter IC for my AM rig and set the cut-off frequency to about 6kc. Easily tuneable and works like a charm. I feed it processed audio so the relatively limited S/N is not an issue.

http://www.maxim-ic.com/datasheet/index.mvp/id/1443

Highly recommended and easy.

73,
Dan

[WB4AIO]

After seeing that Stu AB2EZ modified a SCAF receive audio filter unit for use as transmission filtering, I built an 8 pole elliptical low pass filter using a Maxim MAX294 switched capacitor filter IC for my AM rig and set the cut-off frequency to about 6kc. Easily tuneable and works like a charm. I feed it processed audio so the relatively limited S/N is not an issue.

http://www.maxim-ic.com/datasheet/index.mvp/id/1443

The only problem I see is that, according to the detailed spec sheet, there is a lot of phase shift in the filter's passband, so it won't preserve carefully limited peaks accurately. If that's important to you, you might want to insert it before the final peak limiter instead of after.

With all good wishes,


Kevin, WB4AIO.

[WB4AIO]

It's sad to see the trashy, overprocessed audio on many broadcast stations today. The 5 kHz filter you mention will take an octave or more away from music and do little for coverage, despite what your management says.

Considering the programming on >90% of the AM BC stations these days, I don't think it makes much difference; they could just as well be on slopbucket, if only the "consuming" public would be able to figure out how to tune it in. A few notable exceptions include WSM in Nashville, if you are a fan of their brand of music, or CJBC @ 860 kc in Toronto (jazz and classical), if you live close enough to pick them up clearly.

Sadly, you are right about many AM stations -- and listeners' tin ears -- these days. Actually, the digital audio feeds of the larger talk networks don't sound bad (and there is really no such thing as the "EIB" network, by the way -- it's just a time slot on ABC).

But the sound is often trashed by station management that is willing to sacrifice all pretense of quality to the point of earbleed-inducing clipping, distortion, and atrocious tinniness for all listeners just to get that last largely imaginary decibel of loudness in the fringes. Why they think this increases their listenership is beyond me.

Still, one occasionally hears good audio or interesting independent programming on the AM dial. It's a rare and kind of nostalgic experience when that happens.

Some of my FM receivers have a stereo/mono switch, and when the signal is marginal I can usually get far better reception in monaural mode.  I haven't checked, but I'd guess most sets made these days don't even have a stereo/mono switch, though. I don't recall the exact figure, but I  seem to recall that the receiver sees something like a 20 dB reduction in ERP while in the stereo mode.

It depends on signal strength (and receiver characteristics). Above a certain threshold, there's little perceptible difference in noise. Below it, the difference is huge.


With my best,


Kevin, WB4AIO.

[K5UJ]

<<<But the sound is often trashed by station management that is willing to sacrifice all pretense of quality to the point of earbleed-inducing clipping, distortion, and atrocious tinniness for all listeners just to get that last largely imaginary decibel of loudness in the fringes. Why they think this increases their listenership is beyond me.
>>>

the thinking I'm told is usually that the TSL is assumed to be low, like maybe 15 minutes so listeners will tolerate a lot of compression and high density audio.

[WA3VJB]

The hearing quality of the audience is at play here too.

Demographic studies I have seen on the listenership of AM indicate a lot of older men, many of whom will have a certain amount of hearing loss.

My Dad listens to some of the conservatives and "beautiful music" stations on AM, and his hearing rolls off at about 3Kc.  
Compression, distortion, and constraint of treble are not an issue.

At the other end of the audience analysis, if you have teenagers or know anyone under about 30, ask them whether they've ever tuned into the Standard Broadcast band, at all.

[WB4AIO]

The hearing quality of the audience is at play here too.

Demographic studies I have seen on the listenership of AM indicate a lot of older men, many of whom will have a certain amount of hearing loss.

My Dad listens to some of the conservatives and "beautiful music" stations on AM, and his hearing rolls off at about 3Kc.  
Compression, distortion, and constraint of treble are not an issue.

Yes, and as K5UJ says, some managers think that WWCR-style audio is okay for short-term listening. I think that some pre-emphasis and compression, and even a small amount of properly managed clipping, if artfully done, can actually sound good, even on music, while helping to increase the S/N ratio and "aiding the hearing impaired" to understand Murdoch's latest talking points.

But WWCR-style audio (God, they have to be the worst-sounding station I've ever heard whose transmitter wasn't actually on fire) just drives people away from AM radio.

At the other end of the audience analysis, if you have teenagers or know anyone under about 30, ask them whether they've ever tuned into the Standard Broadcast band, at all.

My girlfriend is 34, and she does a fair amount of listening to radio, about half AM (mainly to hear about local events and sales on a pretty good "full service" station here in Charlottesville).

My children, all teenagers, seldom listen to over-the-air radio at all, except when travelling in the car with me. (They do like Car Talk and, rarely, Garrison Keillor, and very very rarely the local news or "mix" music station.)

They listen a great deal to mp3 players, though, and a small amount of Internet radio. They've probably listened to mp3s of the Goon Show and Lum and Abner more than they've listened to the ol' Standard Broadcast band...


All the best,


Kevin, WB4AIO.

[W1DAN]

Kevin:

Yes all filters have phase shift by design. Even the consumer equalizers will have them. You are correct about the location of the Maxim filter. A good spot to have it.

The Maxim filter I built is very steep and it's phase shift is small for most of the passband, rising quickly as the cutoff is reached. I take a cheaper and dirtier way out for the overshoots. I put the filter after all processing. I add a negative clipping diode after this filter to grab the little HF-based overshoots at the expense of a very slight amount of harmonic based splatter. I then do it again-add another filter and then a diode to reduce the events. Cheap dirty and it works. Not as good as phase shift correction as Orban and others do.

WWCR-uses an Orban HF 9105A. This box is louder and carefully adds more distortion than a 9100A. When adjusted well (listen to the BBC or some VoA), it is very effective with little trade-off. When the knobs are cranked as on WWCR, it is awful. I owned a 9100A and 9105A for a while and know the sonic signatures of them.

73,
Dan
W1DAN

[W2XR]

On transformers, specifically those by United in the "linear standard" series, lots of them won't proof up to the published specs.

Sometimes I found the only difference between the Special series, Commercial grade, and the LS is the case they are potted in.

Old Doc Hoyer back in Wisconsin has wound transformers for almost 80 years now, and he agrees wholeheartedly with this.  Same diatribe on lots of Thordarsons, Western Electric, SNC and Thermador.

Higher marks across the board for Triad, Stancor, and Merit.

YMMV.

73DG

Yes, Dennis, I agree wholeheartedly.

I have always had great success with Triad and Peerless audio transformers, and I too was never impressed with the UTC LS or HA-series transformers and their measured data. One audio engineer who I knew quite well, and is now departed, used to refer to UTC audio transformers as "U Take a Chance". He had the same negative comments about their actual measured characteristics vs. the published specfications.

UTCs power magnetics (filter chokes and power xfmrs), on the other hand, are of excellent quality in my opinion, and I have used them extensively in a lot of my homebrew gear.

73,

Bruce