Wideband recording of historical AM radiofrequency bands

ID: 567631
Wideband recording of historical AM radiofrequency bands 
02.Apr.22 11:04

Hermann Scharfetter (A)
Articles: 49
Count of Thanks: 5
Hermann Scharfetter

Dear friends of radiomuseum.org,

With a delay of 16 years, I would like to revive a topic that Jakob Roschy had posted on 16 Sep 05, in the thread on tuning through a radio, and which initiated my project COHIRADIA which may be very interesting for some radio amateurs and collectors. The cited thread began in 2004 with Ernst Erb's groundbreaking call to make recordings of the sound when tuning through an AM-radio. Facing the rapid disappearence of broadcasting stations on MW, SW and LW due to decreasing importance of the AM technology, he stimulated the conservation of audio-documents which give listeners a sense of what AM radio "felt like" when it was alive and well - an excellent idea indeed. In a reply to his call Jakob Roschy suggested to try broadband recordings of the entire MW or LW band directly from the antenna by means of a video recorder which by default has a bandwidth of several MHz. When coupling the antenna input of an AM receiver to the video output and playing back the video tape, one can retune all stations, thus having an exact 'canned' copy of the AM band at the time of recording - an ingenious idea.

Some RM members made attempts to do so and their reports encouraged me to try the same. I managed to record a number of videotapes with MW and SW broadband signals on a portable Philips VR6711 with acceptable quality from the end of 2006 on. I captured the entire RF signal with a wire antenna, fed it through a bandpass filter (e.g. 500 - 1700 kHz) to a broadband amplifier and coupled the output to the video input of the VR. Playing back these tapes one can actually tune through and listen again to all stations received at that time. Thus I possess a nice archive of original historical radio signals.

However, VCRs do not have a particularly good signal-to-noise ratio (SNR), which means that weak stations are noisy or even drowned out. There is also a small frequency offset, about +3% on my VR6711, because tracking without a real video signal apparently causes a deviation of the tape speed from the ideal one. Furthermore, I could not copy my tapes without further signal loss and make them available to other radio amateurs, although I wished to share my 'treasures'. The suggestion by Ingo Truppel, to record digitally with 16 bit resolution was out of reach for me at that time due to the lack of a fast digitizer. My dream came true at the end of 2020 with the STEMLAB125-14 from RedPitaya, a 10x6 cm 'small' single board computer with FPGA based signal processing hardware. When connected to a PC the device can be used as oscilloscope, signal generator, network analyzer etc., but also as a software-defined radio (SDR) up to 60MHz. It allows to continuously record and playback wideband RF signals at 14 bit resolution. Thus I could digitize all my tapes and store the files on harddisks. For playback I simply couple the STEMLAB via a balun to the antenna jack of an AM radio, usually a tube device. Personally I tried this successfully on my Ingelen Columbus51, Minerva375, Atwater Kent10 and Eumigette W. For a demo see my video.

Since early 2021 I have been recording contemporary MW, LW and SW signals directly digitally in more or less regular intervals. Running a hobby lab with an outdoor antenna strung in the garden on the outskirts of Graz (Austria), I am in the lucky position of still receiving many southern and eastern European stations as well as a few North African transmitters without experiencing too strong urban interferences. The quality is very satisfying, under good conditions I can even pick up weak stations like Greek 'pirate' transmitters between 1600 and 1700 kHz. During playback on a good MW radio I cannot recognize a noticeable difference between original and recorded signals.

Since several weeks radiomuseum.org is hosting my project COHIRADIA and I would like to express my special thanks to Ernst Erb, Ueli Kurmann and Heribert Jung for their invaluable and excellent support. COHIRADIA stands for CONservation of HIstorical RAdiofrequency bands by DIgital Archiving and aims at collecting and archiving broadband radiofrequency bands from analog sources. The purpose is two-fold:

  • provide interested radio enthusiasts, collectors and museums with original radio signals from the past for playback on historical radio receivers.
  • equip amateurs with tools for making their own recordings of contemporary radiofrequency bands all over the world to build up a versatile and international signal database.

Currently COHIRADIA provides a file server and a landing page, from which the following ressources will be available:

(1) Detailed documentation of the currently available hardware and software

a detailed description for the setup of both hardware and software

(2) An installation package cohiradia_V1.2en.zip (172MB) which contains

  • a detailed description for the setup of both hardware and software
  • a Software for Windows10 (COHIRADIA_RFCorder.exe, beta-version 1.1b)  
  • video tutorials part1 and part 2 which show the practical implementation step by step

(3) a continuously growing collection of data files with recordings for download and playback. In addition, there is a short test file (480 MB) with one and a half minutes from a recording made in 2021. Currently the following data files are available:

I am very happy that Gerald Gauert acted as a first 'beta-tester' and successfully played back one of my records on a radio receiver from 1928 after buying his own STEMLAB - for his really exciting  original report see the german version of this thread.

I would be very happy if soon some more amateurs would buy a Redpitaya-STEMLAB (costs currently ca 350 - 400$) and try out my software. In the near future, I will publish the Python sources of my program as an open source project on GITHub so as to make the software available to LINUX users as well. Furthermore, other enthusiasts who ar familiar with Python can write and contribute their own code extensions.

I would also like to mention another small 'subproject' of COHIRADIA in the section 'Audio', which shows a tuning-through video from a recording of the whole MW band grabbed around midnight December 30/31, 2006. In 'Parallel Clips' the video is linked to 15 extracted audio clips representing 15 annotated stations running exactly synchronously in parallel to the video for ca. 11 minutes. The video also contains German, Swiss and French stations which have ceased their services. Most  tracks contain station announcements and news at midnight, often preceeded or followed by the federal / national anthem. I would like to express my sincere thanks to Heribert Jung  for the tireless help with uploading, revising and linking my clips !

Should viewing/listening to these recordings arouse your interest, then I warmly invite you to participate in COHIRADIA so as to make it a piece of lived history of technology. I would be happy if one or the other enthusiast started also operating a STEMLAB125-14 somewhere else in the world and enjoy the playback on old radios. However, I would be particularly excited if several international amateurs  tried to make own recordings. If people located in different sites worldwide could contribute recordings to radiomuseum.org, then together we could build up a unique and highly diverse supraregional collection of still active AM bands, thus preserving for future generations a small cross-section of what is currently still being broadcast.

Best regards

Hermann Scharfetter


To thank the Author because you find the post helpful or well done.

Wideband recording of historical AM radiofrequency bands 
30.Sep.23 17:20
586 from 1127

Hermann Scharfetter (A)
Articles: 49
Count of Thanks: 3
Hermann Scharfetter

Dear friends of COHIRADIA,

Since the launch of the COHIRADIA landing page, more than one and a half years have passed and a lot has happened. Since the description of all activities would go far beyond the scope of a forum contribution, the active team (Walter Barteczek, Gianni Müller, Stan Roberts and me) has written a report 2023, in which all important basics and activities for COHIRADIA until September 2023 are documented. Interested readers will find background information, quality criteria for records, all relevant details on hardware and software used or recommended, and references.

For those who prefer an 'short version', here comes the preface:

Until December 2022, the active 'team' consisted of relatively few interested people, of whom RMorg members Walter Barteczek and Stan Roberts are especially noteworthy, who not only provided essential ideas on many technical issues but also contributed quite a few interesting shots. I was also very pleased with the use of the RFCorder on the STEMLAB by Gerald Gauert at the Staßfurt Radio and Television Museum. Since quite a few AM stations in Europe were finally switched off in the last months (RAI Italia and Absolute Radio on MW, RTL Beidweiler and RTE Ireland on LW ...), we had our hands full with recordings to capture the last hours of transmissions of these stations together with all other frequencies of whole MW and LW bands for posterity. In response to a few calls for help on mediumwave.info from Ydun Ritz, Gianni Müller (Lucerne, Switzerland) joined the COHIRADIA 'family' in January 2023.

Gianni Müller has contributed several recordings made with PERSEUS back to 2015, all of which are of very good quality and document an even much greater station richness than we now have on MW and LW. His LW recording from 2019 with the stations Europe 1 and RMC Info, which were switched off shortly afterwards, is a real rarity.

At this point I would like to express my gratitude to all mentioned colleagues, because due to their very careful and time-consuming cooperation the archive of COHIRADIA has meanwhile reached a respectable size with typically very good data quality.

The effort is not to be underestimated, since the recordings must first be created with high-quality equipment, which requires certain planning in advance. Afterwards, all data must be resampled to a uniform file format that can also be played back by the RFCorder on analog radios. Particularly complex is the creation of the metadata, i.e. the frequency lists including information on transmitter site and reception strength, which makes it possible to identify the individual stations even later. Here the standard was set by W. Barteczek, whose template essentially forms the basis for the current documentation.

Special thanks are due to RMorg system administrator Ueli Kurmann, who built the entire IT infrastructure for COHIRADIA and is constantly working on the further development of the site with well-elaborated server programs. Thanks to his programs it is now possible to download the recordings either in the wav format necessary for many software-defined radios (SDRs) or in the dat format necessary for the RFCorder, the conversion is done automatically during the download. This means that a very powerful environment for archiving recordings is now available and it is getting better all the time.

Unfortunately, at the moment it still takes a few days from the actual recording to the release on the webpage, because each recording has to be checked by me and if necessary cut and resampled to a specific target frequency band. After that the annotation is done by the creators of the recordings, thank you very much for the effort.

The database will continue to grow, with many recordings since 2021 still pending, which have not yet been processed by me.

The highlight of the collection is the 'New Year's Eve/New Year's Day 2022/23' project, which consists of a whole package of simultaneous recordings of the MW and LW bands at 3 different locations in Europe and one location in the USA, made by Walter Barteczek, Gianni Müller, Stan Roberts and myself. The three European recordings include the shutdown of the two Slovak stations on 702 and 1098 kHz, and the shutdown of RTL on LW (234 kHz) the following day. This truly international project is unique because different transmitters are imaged differently depending on the location and directivity of the antennas used. We repeated this game again on January 19, 2023, this time on the occasion of the shutdown of Absolute Radio in England (1197, 1215, 1233, 1242, 1260 kHz). Also here a very nice diversity of the recordings can be seen, very interesting is the recording of Walter Barteczek, which is only 100 kHz wide, but has the best signal-to-noise ratio and contains all relevant frequencies.

It should also be mentioned that we have extended our recordings into the VLF range, which can already be seen on some of the recorded files. For example, we now always record when the machine transmitter SAQ Grimeton in Sweden makes its special broadcasts on 17.2 kHz. There is a first recording from 13.02.2023 on the occasion of the 50th anniversary of the UNESCO World Heritage Initiative, further recordings of this unique technical cultural monument will follow.

Of course, there were also quite a few technical developments in the context of COHIRADIA:

  • Preparation of a version 1.2 of the RFcorder, so that wav-files can be played and recorded directly as well
  • Integration of a semi-automatic tool called 'COHIWizard' for annotation (creation of metadata) of the recordings.
  • Discussion on the use of alternative SDRs for recording and playback
  • Details on file formats
  • Recording practices: antennas, preamps, filters, etc.

The last point is important in that the first COHIRADIA contributions were still based on a clearly suboptimal technology of preamplifiers, the shortcomings of which were pointed out very clearly in a forum contribution by Walter Barteczek with his rich experience (original in the German thread only). As a result of this decidedly constructive criticism, the system in Graz was completely transformed within half a year, gaining at typically 20dB in signal-to-noise ratio. The still available first videos and descriptions of the RFCorder (video and text) in the RMorg are therefore to be taken up with caution regarding preamplifier and antenna and are to be revised in the future. More about this in the chapters 3.2 and 3.3 of the report.

The active COHIRADIA family is definitely still very small, and we hope of course to attract so much attention in the next years, that as many collectors and radio enthusiasts as possible, possibly also further museums will start to use COHIRADIA and its meanwhile quite respectable data treasure.

A further step in this direction could be that an abstract submitted by us has been accepted for presentation at the IASA annual conference (for the text see conference program). The theme of the conference is '100 Years of Broadcasting and 70 Years of Television in Germany'. Here I would like to thank Mr. Gallenmiller, who pointed out the possibility to submit papers in this RMOrg thread. This will certainly give some good opportunities to get in contact with interesting people and to spread the idea of COHIRADIA.

On behalf of the entire team, I hope you enjoy reading the 2023 report.

Hermann Scharfetter


To thank the Author because you find the post helpful or well done.