Jed Margolin's Simulated AM Radio


United States Patent       5,978,488     Nov. 2, 1999

A realistic illusion of listening to an actual AM radio is produced by providing a number of audio program signals and by simulating the process of tuning across the radio dial where each station is playing a different program. In a superheterodyne receiver using envelope detection, tuning a station off-channel results in a changing mixture of distortion and noise. The distortion function is caused by a variety of processes, among them being the filtering action as parts of the signal spectrum are cut off by the bandpass filter in the IF amplifier. Another is caused by the reduction in the carrier which causes the foldover of negative peaks in the envelope detector.

Other Patents That I Find Interesting

U.S. Patent 1,342,885  METHOD OF RECEIVING HIGH-FREQUENCY OSCILLATIONS   issued June 8, 1920 to Armstrong.

The superheterodyne was invented by Edwin H. Armstrong.

The architecture of Armstrong's receiver is shown in FIG. 13a. The broadcast signal is received by Antenna 1301 and multiplied in Mixer 1302 by the signal from Local Oscillator (LO) 1303. The resulting difference signal is amplified in Intermediate Amplifier (IF) 1304 and envelope detected in Detector 1305. The listener hears the demodulated signal in Headphone 1306 (referred to by Armstrong as a telephone).

Note that Armstrong did not use an RF amplifier. The reason for this is that the triode vacuum tubes of the day were not capable of operating as amplifiers at the high frequencies (above 1 MHz.) which were Armstrong's goal. The parasitic capacitance between the plate and the grid caused amplifiers to oscillate when operated at high gain. (The higher the frequency, the greater the positive feedback through the parasitic capacitance.)

By converting high incoming frequencies to a low frequency, Armstrong's method allowed IF Amplifier 1304 to operate at a frequency where triodes could be used at a much higher gain without oscillating. In addition, by tuning Local Oscillator 1303, all frequencies were converted to the same IF frequency so that IF Amplifier 1304 could be optimized to operate at a single frequency. This made it possible to tailor the gain and filter characteristics. Armstrong's '885 patent even teaches a double conversion superheterodyne receiver operating at 10 MHz.

U.S. Pat. No. 1,501,543  VARIABLE CONDENSOR issued Jul. 15, 1924 to Hammarlund.

The main object of this patent is to provide a variable capacitor that can be accurately and precisely adjusted.

What I find most interesting is that FIG. 4 shows the axis of rotation of Rotor 9 and Stator 8 offset from the center of the midline.  This provides a capacitance curve that linearizes the dial so that the stations are not bunched up at one end.

This became the standard method used in the construction of AM radios and is discussed in my patent (5,978,488).

U.S. Pat. No. 1,604,552    CONDENSOR   issued Oct. 26, 1926 to Hammarlund.

The main object of this patent is to provide a method of construction that is efficient and reliable.

Another object, which I find more interesting, is to provide a capacitance curve to linearize the dial so that the stations are not bunched up at one end.

FIGS. 4 and 5 show Stator 17 with a curve along the midline. As Rotor 31 is rotated, the capacitance will be modified by the shape of the curve.

Dial linearization is discussed in my patent (5,978,488).



I started this work sometime in 1998, maybe before that. I had been to a flea market for oldtime radio and had seen a beautifully restored Atwater Kent radio. The asking price was about $4,000. (Atwater Kent:


I didn’t buy it.


But on the way home I thought, suppose someone bought it, brought it home, and plugged it in. What would they hear?


They would have heard Dr. Laura. (She was on 450 radio stations back then.)


You might like Dr. Laura, and that’s ok, but Dr. Laura is no Jack Benny.



I wondered how to simulate an oldtime AM radio. When you tune between stations you get a changing mix of noise and distortion. What causes it?


I consulted some old radio engineering books. Nothing. I also did an Internet search but didn’t find anything about it.


So, I did my own research. The results are in the above patent.


Once I figured it out I made a demo program. Actually I did several demo programs.


I did a version that ran in DOS using either a 40 MHz 386 or a 120 MHz Pentium (I forget which.) I used Borland Turbo C which had a nice IDE.


Another version ran in a DSP board I had made that used an Analog Device ADSP-2105 and which talked to a CD-ROM drive which contained the radio program data.


This was a completely stand-alone unit that used a modified replica radio. I used an optical encoder to read the dial position and send it to the DSP board, and sent the audio back to the radio’s volume control, amplifier, and speaker.


The other projects I used the DSP for were things like:


a)  Reversed Speech (See and take it with a grain of salt, but not too much because too much salt is bad for you.)


b)  a Voice Stress Analyzer (See I got it working a few weeks before the end of the OJ Simpson trial (October 1995). I came to the conclusion that using Voice Stress Analysis to determine if someone is telling the truth is akin to examining the entrails of a chicken. And there are lots of things that can produce stress in a person, not just lying. And there are people who can lie and not be at all stressed about it. (See Also see this:


The advantage of using the entrails of a chicken is that afterwards you can eat the rest of the chicken.


c.)  a simple NTSC character generator that used the ADSP-2015 for everything (sync and video).


Nowadays you can do a) and b) on a PC or a smart phone without any problem. Back in 1995 you couldn’t. PCs were too slow and smart phones didn’t exist.


The last version of Windows that could run the DOS program was Windows 98SE. After that, Windows would not run DOS programs that used graphics.


So, I bit the bullet and ported it to Windows.


The Windows program runs in Windows 98SE, Windows Vista, Windows 7, Windows 8/8.1, and Windows 10. Frankly, I am amazed that the program (developed in Windows 98SE) runs in the versions of Windows that have come afterwards.


Here it is:



            Pfile.bin  (40 Mbytes)


Download both files to the same directory and run Radio.exe.


1.  By all means use your virus checker to verify that it is not malware.


2.  Down this program only from my Web sites ( and


3.  The program runs in place. It will not install anything on the computer or make any changes.


4.  If you tune the dial slowly you can hear the distortion and noise just like a real oldtime AM radio.



When I developed the original program my PC was not very fast.


1.  Decoding several streams of MP3 was out of the question so I used u-law encoding, which is an old telecom compression method that compresses 14 bit to 8 bits.



2. I created one radio program file (Pfile.bin) that interleaves eight radio programs.



When I made the original DOS program I thought it would be a good product for PCs and video game systems and would be sold on CDs. There would be a number of CDs with different sets of radio programs.


Although the Internet existed, most (maybe all) of the users were on dial-up.


Cell phones existed but were not Smart. And were analog.



I think this could still be a good product for PCs, Pad PCs, and Smart phones.


The radio program data could be either streamed or downloaded to a hard drive (PCs) or a memory card (Pad PCs and Smart phones).


It would require that:


1.  The radio program files be created and served.

2.  Many (maybe most) oldtime radio programs are in the public domain but care must be taken to make sure that they are.



Why would someone want this app?


When I listen to radio nowadays all I hear is crap. Royal steaming boring crap. A few companies own most of the radio and television stations. With declining listenership they conclude that they need more technology like HD radio. But crap broadcast in HD is still crap.


Oldtime radio is still funny and interesting. Many young people have never heard it and I think they would find it funny and interesting too.



If you are interested in partnering with me on this (or just buying the patent and doing it yourself) you can contact me at:

The patent expires 20 years from the filing date of September 3, 1998 which makes it September 3, 2018. (Where did the time go?) After that you would still have enforceable worldwide copyrights for the software. You might also have enforceable copyrights for the radio programs as derivative works assuming you process the material in some meaningful way, such as by cleaning it up. You would need a copyright attorney’s advice on that.  


The deal would come with the domain name .



BTW, my first Web hosting service gave me 60 Mbytes of storage, and the demo radio program file is 40 Mbytes. Things sure have changed.



Jed Margolin

Virginia City Highlands, Nevada


September 23, 2015



Copyright 2000, 2015 Jed Margolin