Out of Order: Tracking Down a Mystery Signal
Answering my call for entries in the new Out of Order feature on RF Cafe, Mr. Denny Condron responded with a great saga of his use of finely tuned (pun intended) use of fox hunting skills to track down and ultimately resolve a bad case of interference in the 2-meter amateur radio band. Although names are omitted to protect the innocent (and the guilty) parties, it is good to know that the offending source was remedied willingly by the owner of the equipment: a manufacturing operation. This is a fine example of the term "unintentional radiation," and why ignorance is usually only bliss, as the saying goes, for unaffected and usually oblivious parties. Denny noted separately that this tale is a greatly abbreviated version of the effort required to find and remedy the interference source. Do you have a good work-related anecdote to share? Please email it to me for consideration. Thanks.
Tracking Down a Mystery Signal
By Denny Condron - K0LGI
An interfering signal had been perplexing amateur repeater operators in a Midwest city. Several that were operating above 147 MHZ in the amateur 2 meter band often experienced an intermittent, raucous sounding, drifting signal modulated by 360 Hz AM and FM components that came and went for several months. Interference was primarily in the FM band segment.
Although there had been discussions about what it could be and what might cause such an interference, little confident interference source directional information had been determined in reference to its location.
After confirming the typical daily frequency drift pattern and on-off timing characteristics, the mystery signal was exhibiting from the home location over a period of several weeks. A heading roughly towards the city center had been established and with this information in hand, a determination was made to locate it using by now well-established methods from a mobile unit.
A portable HP spectrum analyzer, a broadband HP unit preamp and heavy duty 12 volt to AC inverter was obtained and connected with the car portable DF (direction finding) antenna. On a late evening of a cold, snowy December day we set out to locate the offending signal origin.
Because of the characteristics of the signal, the mobile Doppler DF unit could not be used without constant frequency changes. The spectrum analyzer was essential as a time saver not only because of the highly variable frequency of operation and drift characteristics, but also the sporadic on-off times of this unknown interference source.
After several miles of driving - with some minor off-nominal signal arrival direction headings due to the usual city environment multipath - finally at the top of a hill it was noted that the signal had a dramatic increase in strength. That was in contrast to any previous DF location observed en route when heading in a similar direction or on the back side of the hill. The DF equipment continued indicating a straight-ahead bearing, assuring a good clean, non multipath heading was valid.
As the mobile unit continued on this street, the signal levels were soon of such high levels that the preamp was no longer needed and in fact, several 10 dB steps needed to be added to the spectrum analyzer RF input to reduce the signal power to more usable levels for DFing.
At last during on this very cold, inclimate evening, after seeing the DF indicating a hard right turn was obviously needed, I arrived in the front lot of a facility from which the signal was being broadcast. A DF read was made from the front of the building and also 90 degrees offset from the original heading at the south end of the building, confirming finally this spurious signal emission orgin.
The only obvious accessible entrance to this facility at that time of night was in the rear side of the suspect building. A contact was made with an employee after ringing a kind of alarm bell to alert staffers to activity. The employee that responded as if he had no idea or interest in what I was referring to, and suggested I contact the front office staff during the daytime hours for help with the issue.
Doing so provided, eventually, an invitation by an official at the company to bring whatever was needed to show a company representative exactly what the issue was and to demonstrate it to him.
Gladly accepting this opportunity, the same test equipment was used and I was greeted the next Saturday morning by the company representative that secured the equipment. In addition, I was taken on a tour of the manufacturing area when employees were on break, so they were not present when I was in the area.
Later as employees again entered the production area and began processing the products of a variety of types and designs, it was noticed that the RF equipment that had been seen during the initial guided tour was in now in use. It was used mainly to preheat powdered plastic which was then formed in a mold shaped and pressed into the product sizes required.
The RF equipment that this company was using consisted of over 20 manually operated RF power oscillators generating approximately 0.5 kW of RF power and one semi-automated oscillator generating approximately 2 kW of RF power.
In watching the ones that were in use, it soon became obvious that the high power semi-automated unit was showing considerable visible arcing around the chamber access doors where the products were being transported in and out. Upon seeing the arcing, it was immediately evident from where and by what means the RF was radiating.
After the employees once again left the production area, I proceeded to set up the equipment I had brought with me, including a small probe connected to the spectrum analyzer to sample each of the manual RF ovens, including the major radiator: the higher power semi-automatic unit.
It was then realized that in fact these long-sought-out interfering sources were operating on a 'grand fathered' ISM (Industrial, Scientific, Medical) band at about 74 MHz! Also verified was the same 360 Hz modulation components noticed previously in measurements that had been made in both fixed and mobile environments.
Although it was never witnessed that the RF devices were ever operating near to or at the airport marker beacon frequency of 75.00 MHz, it could have been a far more serious issue if they had drifted much higher in frequency. Previous measurements made in the mobile unit and in the field had been done on the second harmonic of these erratic signals, occurring in the amateur 147-148 MHz, 2 meter amateur band.
After several lower power RF oven units were also verified as having substantial leakage radiation, the semi-automated unit was approached with caution as it was loaded with the required amount of product plastic powder and molds in use at the time.
Although I don't recall the level of the RF that was measured emitting from this massive beast, I do remember that the previously used probe did not need to be located anywhere near as close to the oven to see major radiation from it - even with significant attenuation set on the spectrum analyzer RF input. It was this source that had the 360 Hz modulation component which had been plaguing our radio operations, and now was confirmed by the discovery that it used 3 phase AC power input for a supply.
Manual and automatic oven door RF gasket finger stock was either missing or broken on both types of RF heaters.
Austerity measures were in place on equipment repair and maintenance minimizing equipment upkeep.
Maintenance efforts were reestablished after acknowledging the RF leakage existence was substantial and causing interference.
Interference was gone within days in the 147 MHz region that had been noted before the discovery visit.
Awareness was made to the company regarding interference caused to other services, including concern of the nearby airport operating Marker Beacon frequency.
Awareness was made of the concern of RF exposure to employees who were operating deficient units due to excess leakage.
Posted February 14, 2014