AM music sounds really good on all recently designed radios. By recently designed, I mean within the past 15 to 20 years. Another piece of iBiquity misinformation / outright lies was their contention that AM radios are limited to 3 to 4 kHz response. This was true in the old "All American 6" transistor radio design and its derivations. Those designs used three IF transformers. You can recognize them by color of the adjustment slug - yellow, white, and black. The fourth can, which had a red adjustment, was the oscillator / mixer / converter stage. These radios used no RF gain transistor - RF gain was accomplished by the Q of the ferrite bar antenna and antenna section of the tuning capacitor. One transistor was the oscillator, the second two were IF gain, the fourth was an audio pre-amplifier, and the last two were a class B push pull amplifier stage. There was also one, and most likely two audio transformers. This design changed little for 3 decades, surviving the change from germanium transistors to silicon, PNP to NPN.
Slight variations occurred to add an RF gain stage - usually a can with a green adjustment. Radio Shack made a variation with 4 IF cans. Transistor counts varied from 7 or 8 for tuned RF models, up to 9 for models with a more powerful audio output or voltage regulation. Transistor count was perceived as a measure of receiver quality, the more transistors the better - which led to 10 to 12 transistor units where the extra transistors did almost nothing. But the constant was the audio response at 3 to 4 kHz.
Two key technical advances changed AM radio design completely:
With the advent of IC's, a lot of the gain functions were put into the IC. Some IC's contained just RF components, such as the LA1260 from Sanyo, others also had audio output capability like the TDA1083. Initially, they still used the IF cans and therefore had the same audio response as before. Those ICs also made it very easy to incorporate FM into the radio, so there were almost no AM only radios made.
Ceramic filters were introduced about the same time, but the IF can manufacturers fought very hard to survive by cutthroat price reductions to fend off encroachment by ceramic filters. They eventually lost the battle - relegated to oscillator and FM detector sections. Along with the change to ceramic filter IF stages came cost reductions, and ceramic filters got very bad.
The bottom line is that the present state of AM radio sections of radios is pathetic. They are wideband - not by any desire on the part of radio manufacturers to produce radios with good audio quality, but due to the fact there is one very cheap, sloppy ceramic filter for IF. I have found radios with AM bandwidth of +/-40 kHz. So if you have a strong local station with nothing near it in frequency, music sounds really good! But put a station nearby, and you get a mixture or only the stronger of the two stations. My daughter's little radio, bought ten years ago, was a good example. We were in Dallas, she wanted to hear Radio Disney on 620 - a strong local station. But KSKY on 660 completely swamped it due to the wide IF response. I installed a better ceramic filter, and the radio worked just fine. I have found radios that use a ceramic capacitor instead of a ceramic filter to save money. This, of course, makes the entire selectivity of the radio dependent on the Q of the ferrite bar antenna and the tuning capacitor resonant circuit. I am not sure I'd even call them superheterodyne at this point. The tune the antenna, tune the oscillator, but the IF is just a broadband amplifier. So they are little better than all tuned RF design, which I have also seen in novelty radios. If the tuning capacitor only has a single section, and there is no oscillator coil, it is pretty obvious you are only one step above a crystal set.
Where iBiquity lies came in - they carefully hand picked a set of radios that had the old all American 6 reference design, took the response, and announced to the world that typical AM radios are limited to 3-4 kHz, and nobody would miss broadband response so they could brick wall low pass AM audio so their system would work. The problem is - no all American 6 reference design radios have been produced in almost 20 years, and virtually all new AM radios are broadband. Not only will consumers notice a change to station audio quality, they will also hear sideband hiss from the portions 10-15 kHz, and also from 5-10 kHz unless the radio is perfectly on frequency. iBiquity has perpetuated this lie for years and been virtually unchallenged. It makes you wonder what other lies they told us. It is not just bad science, it is corrupted science, altered to support their pre-conceived publicity no matter what the truth is.
FM sections of most radios aren't much better. People care about FM more than AM, so you see a true superhet design with two gang antenna / oscillator tuning, one ceramic filter (accompanied by one IF can about 50% of the time), and either ceramic or coil discriminator. The result is a credible unit for local and some rim shots, with the added IF can doing a lot to eliminate images in strong signal locations.
I've been able to improve almost any of these new radios by putting in better ferrite bars for the AM, putting in better ceramic filters for AM and FM, and actually aligning the thing with instruments - a step that seems to be omitted by manufacturers. Doing so yields radios with decent selectivity about +/- 6 kHz AM, and decent sensitivity. FM radios with one 150 kHz ceramic filter instead of 280 kHz can tune second adjacents next to locals, and in some cases first adjacents if they have some good signal strength relative to the local. I wouldn't call them "DX models", but there is a lot of good design put into those IC's and it can lead to decent performance if the surrounding components give it half a chance.
The newest IC's are offering a lot of promise. They take even more components into the IC, directly sampling AM, and using integrated filtering for FM IF. I've even seen chips from people like SiLabs that put everything except the antennas into the IC. It is a sad end to hardware hacking to improve the radios, because the IC does everything with no handles for the hobbyist to change things, but if the trend continues even the $5 radios at Walmart will have the same capability as a top of the line component, because they use the same IC that integrates everything and leave no room for cost reduction.
What happened to analog clocks 25 years ago is happening to radios - single reference designs that are going to be "as good as it gets". It will be a few years yet, but it will happen. Smaller, cheaper, one IC, small, low power consumption, and with great performance. It will happen. I just wonder at this point whether AM will even be included on the IC because AM antennas are bulky.
.)