Overview for Managers:

Occupied bandwidth and RF harmonic measurements are required on all AM stations annually.  The FCC allows a maximum of 14 months between successful measurements to stay in compliance.  This kind of measurement, otherwise known as NRSC, replaced annual "Audio Proof of Performance" measurements and was required starting on June 30, 1994.  

The first year saw many stations making equipment changes to bring facilities into compliance.  As a result, interference on the AM broadcast band has noticeably decreased.      

Details for Engineers:

AM NRSC compliance testing kept a lot of other people, including me, busy these days.  As a C-QUAM AM Stereo installer and technical consultant to 40 stations, I was well aware of the problems stations could face in getting transmitters to pass the occupied bandwidth and spurious emission tests. 

A legal loophole was inserted in the FCC rules in 1990 when the FCC temporarily suspended the requirement to do performance testing if the stations involved had installed an NRSC compliant audio lowpass filter. 

AM RF harmonics are required to be suppressed at least 73 DB below carrier on a 1 KW station and 80 DB on a 5 KW or higher power station. All AM stations have been required to meet those specifications since the 1960�s.  Over the years, I repaired many RF harmonic problems by redesigning antenna coupling networks and installing RF traps.

Of the 47 stations I personally measured the first year of NRSC, about 20% did not make specifications on the first try.  About half of those had RF harmonic levels in excess of that allowed by the FCC.  The other half had occupied bandwidth problems. 

I found two stations with significant RF mixing products resulting from proximity to another AM station.  Both of the troubled stations had Harris MW-1A transmitters although the problem can happen with any transmitter.  It is just a matter of degree.  For one station, I built and installed an RF trap to keep the unwanted station from getting into the transmitter.  The trap lowered the level of the unwanted station until the mixing product was transmitted at a level that was low enough to make FCC specifications.  In the other case, the two stations involved were 60 KHz apart.  The mixing product appeared 60 KHz on the other side of the client's carrier at about 68 DB down.  It fell within the NRSC mask.  If it had not, the mix would have been illegal.  So, the client was lucky in that he did not need to purchase an RF filter even though the mix was 5 DB hotter than would otherwise be allowed.  

I have carried a Delta Splatter Monitor since they were first introduced.  It makes a great precision demodulator for tuning AM transmitters.  For occupied bandwidth measurements, I carry a spectrum analyzer.  It has an RS-232, which links to a notebook computer.  The screen data is stored on floppy disk, which is carried back to the office and printed by a desktop computer along with the rest of the report.

 If a transmitter sounds good on the air then chances are it will pass the occupied bandwidth testing.  I ran into a few stations where the processing badly smashed the audio before the NRSC low pass filter.   The station sounded bad, but passed the test. 

Tube type transmitters are most likely to have bandwidth problems when the tubes have low emission.  Almost no tuning will fix that.  Remember that harmonic products of audio may easily fall outside the 10 KHz bandpass if the harmonic distortion is created in the transmitter after the NRSC filter.

I ran into several pulse duration modulated transmitters with problems.  One was a Continental Power Rock where a PDM filter coil went out of tolerance allowing more than the legal amount of 70KHz modulator switching frequency through.  A Nautel AMPFET 5 had an open capacitor in one of its 70 KHz notch filters. 

It seems all of the Harris MW-1 and 1A transmitters I checked used up more of the dial than their high level plate modulated counterparts.  In most cases, adjusting the RF Driver tuning made a big change in how clean the transmitter was.  Harris recommends running the RF driver at about 1.5 ampere of DC current. They also have a modification kit to clean up much of the splatter.  Apparently transmitters on the high end of the AM band are most likely to occupy more bandwidth than is allowed.  The kit takes at least three hours to install.

See you further down the road.  I'll leave the soldering iron on for you. 

Mark Persons, W0MH, is certified by the Society of Broadcast Engineers as a Professional Broadcast Engineer with more than 30 years experience. He has written numerous articles for Radio World over the years. His Web site is www.mwpersons.com

From the Radio World April 7, 2003 issue  http://www.rwonline.com