A Messy Ground Radial System Can Cause Radiation of Spurious Signals
By Walter Maxwell, W2DU
In 1948 I was the consulting engineer for the proposed first AM broadcast station in Mt. Pleasant, Michigan, that resulted in a construction permit and license for WCEN, 500 watts on 1150 KHz. Using a National HRO receiver, I was performing a hands-on search for an available frequency for the new station, when I encountered an interesting and unusual signal that was entirely out of place in the AM broadcast band—a CW Morse-code station illegally transmitting five-letter-word code groups at 30 words per minute. The illegal signal was S9 +40 dB on 1297.8 KHz, producing a 2200 Hz beat-note with the 1300 KHz frequency of WOOD, Grand Rapids, thus producing a an audible CW signal with the receiver BFO disabled.
The format of the coded messages appeared to be military, IDing as NSS. We know that NSS is the flagship station of the U.S. Navy in Annapolis, MD, but in the AM broadcast band? It appeared that either an NSS transmitter was producing a spurious emission, or a station using NSS as a fraudulent call sign was operating clandestinely in the AM broadcast band. I deemed it important to find out which.
As a former FCC monitoring officer at the Allegan, Michigan monitoring station, the next step was to report the situation to the Allegan station. Although Allegan was 90 airline miles away, the monitoring personnel there could not hear the spurious signal, even though it was S9 +40 dB at Mt. Pleasant. I let the FCC monitors hear the signal through the telephone, but they still heard nothing on their receivers tuned to 1297.8 KHz. Thus the signal must be of local origin near Mt. Pleasant, and not from NSS. However, to be on the safe side, FCC notified the Navy of the spurious signals, and the NSS operators began combing all their transmitters for spurious signals, and found none. The situation is now becoming even more strange.
I then copied five minutes of the coded text and sent a copy to the FCC, who relayed it to NSS for comparison to their transmissions. The situation is now both perplexing and frustrating, because the text I copied on 1297.8 KHz agreed exactly with a transmission that had been made by NSS. How could that signal have been transmitted on 1298.7 KHz if no spurious signals were emanating from NSS? But it’s obviously not a fraudulent station. What then?
A partial answer came shortly thereafter. As I resumed the search for a useable frequency for the new station, I proceeded downward from 1298.7 KHz, going through 1280 KHz and hearing WFYC, Alma, 1000 watts, 15 miles away, also S9 +40 dB. But on continuing further downward I immediately came across another S9 +40 dB thumping CW signal. I switched on the BFO and discovered the CW was also a five-letter-word coded transmission at about 30 wpm, the same as NSS. I retuned to 1298.7 KHz and the NSS signal was also there, as before. So I cranked up a second receiver to monitor both frequencies simultaneously. Surprise! Both frequencies were showing identical simultaneous transmissions. I then measured the frequency of the lower-frequency signal—1262.2 KHz. Voila! The higher CW frequency was 17.8 KHz above WFYC’s 1280 frequency and the lower CW frequency was 17.8 KHz below WFYC’s frequency. A quick reference to the Berne frequency listing showed NSS assigned to 17.8 KHz. This situation now appeared to indicate something very wrong going on at WFYC. The low-frequency world-wide ground-wave signal from NSS was apparently somehow mixing with the signal from WFYC, and producing the 1297.8 and 1262.2 KHz sum and difference frequencies. But what non-linear device in WFYC’s system could perform that mixing? Don’t know, but I reported this new information to the FCC and that was the last I heard of the situation…until…
Fast forward now to 1955. I was now employed as an electrical engineer at the RCA Laboratories, the David Sarnoff Research Center in Princeton, NJ. An assignment took me to Washington, D.C. to attend the annual conference of the NAB, the National Association of Broadcasters. President Dwight D. Eisenhower gave the keynote address. However, one of the technical forums was presented by Jack Young, Chief Engineer of the RCA Broadcast Division. His topic was on the solution of mutual interference between two broadcast stations in Los Angeles, KFI and KNX.
It seems that in a section of the Los Angeles area it was impossible to hear one of these stations without hearing both simultaneously—when tuned to 640 KHz for KFI, both KFI and KNX were heard, and when tuned to 1070 KHz for KNX, both KNX and KFI were heard. Young was assigned the task of determining the cause of the interference and eliminating it. To make a long story short, he discovered that there were ancient and rusty oil well derricks in the affected area. Currents from both KFI and KNX transmissions were being induced in the oil well towers, and the rusty joints were acting as mixers for the two frequencies, producing both the sum and difference frequencies, as well as re-radiating both signals on their original frequencies. When the derricks were removed the interference stopped.
So how is this incident relevant to the NSS problem? Well, at the conclusion of Young’s presentation I had the opportunity to talk with him, and because of the similarity of the problems, I told him of my discovery of the NSS signals appearing in the AM broadcast band. Talk about coincidences! He was surprised and excited to learn that I had discovered the NSS problem at WFYC, because he was the one assigned to determine the cause of the problem. He had never been told how that problem originated, or how the problem had been discovered.
He then explained that he had found the ground radial system under the WFYC antenna a horrible mess. Cold solder joints throughout, and far ends of the radials hanging loose in the water of the nearby Pine River, establishing the non-linear mixer condition that resulted in the sum and difference frequencies being generated between the NSS and WFYC signals. Cleaning up the radial mess ended the appearance of the NSS signals in the AM broadcast band, thus concluding an interesting journey.
