RFI from WNTS

Details of the problem

Dec 20, 2020: While tracking down radio frequency interference I (Don, WD8DSB) often listen on 1710 KHz for power line interference using my cars AM radio while driving around and on Sunday Dec 20th, 2020 I noticed massive splatter on 1710 KHz that was being generated from AM radio station WNTS in Indianapolis that transmits on 1590 KHz. The splatter was not present 10 days earlier when I was down in the same part of town, so the problem was new. Here is a link to a youtube video I made capturing the splatter: youtu.be/VaunFwVAyUI

I could hear the splatter on my cars AM radio from at least 10 miles away from the WNTS transmitter site. Mike (W9RE) who's located 6.7 miles from WNTS reported he could also hear the splatter on the lower portion of 160 meters, and Dan (KB9AX) located 9.8 miles from WNTS could also detect the splatter on 1.83 MHz during the day and he could also see spurs approximately 50 KHz each side of the WNTS carrier. I was also able to hear the WNTS splatter during the day on 1650 KHz from my house which is located 18 miles from the WNTS transmitter site.

Note: WNTS runs 5KW during the day, and 500 Watts at night.

Dec 22, 2020: I was able to contact Bart Johnson who is the WNTS station manager who then put me into direct contact with their Chief Engineer. I was then invited down to the transmitter site to work with Kim Hurst (WB9PRS) who is the Chief Engineer for WNTS and Mike Sienicki (WB9YPA) who is a contract RF Broadcast Engineer out of South Bend Indiana who Kim uses a lot.

Dec 23, 2020: I met with Kim and Mike at the WNTS transmitter site and we could easily hear the spur at 1650 KHz on a portable AM radio with RF gain greatly reduced, and then Mike fired up his spectrum analyzer but we were not able to connect it directly to the transmitter monitor jack because Mike did not have adequate attenuators to protect his spectrum analyzer, so Mike connected a small 6 inch diameter un-tuned loop to the spectrum analyzer input (just using some scrap wire). We then turned the transmitters modulation off so we could just observe the carrier when running 5000 watts, and while we still observed a spur at 1650 KHz from the transmitter it was approximately 65 dB below the carrier and the spur and carrier looked clean (no signs of splatter). We then turned the modulation back on and that's when we saw all kinds of splatter peaking around 1650 KHz which overwhelmingly convinced Kim and Mike that we indeed had a sick transmitter (the splatter at 1650 KHz was much stronger than the unmodulated 1650 KHz spur we previously observed). Based on Mikes experience with the Harris Gates Five transmitter the problem was with the Pulse Duration Modulator. Mikes recommendation was for Kim to order the capacitor repair kit for the Pulse Duration Modulator and this would involve replacement of approximately 26 capacitors within the transmitter cabinet. Sounds like the transmitter was built around 1995, and we saw evidence that a few of the capacitors had been previously replaced.

Note: Kim then ordered the capacitor repair kit which had a 12 week lead-time, and as of early March some of the capacitors had not yet been received. While waiting for the transmitter to be fixed WNTS ran 24 hours a day at their reduced nighttime power level of 500 watts which greatly reduced their spurious emissions.

December 23, 2020 Spectrum Plots Shown Below

I kept wondering how the 1650 KHz spur that we saw in the spectrum analyzer screen shot got generated (saw it when we turned off modulation), and think I figured it out. I found a manual for the Harris Gates Five transmitter and it's a pulse width modulated AM solid state transmitter (they called it pulse duration modulated which turns out to be the same as pulse width modulated). The pulse width modulator operates at 60 KHz in the Harris Gates Five transmitter, and the PWM filter should filter out the 60 KHz component and just let the base band audio pass, but if the filter has bad capacitors then the 60 KHz signal could easily mix with the main carrier of 1590 KHz to produce a signal at 1650 KHz. We also noticed that the hash peaked at 1650 KHz. I now can see how bad capacitors in the pulse duration modulator and especially in the pulse duration modulator filter would eliminate the 1650 KHz spur we saw along with the associated hash that peaked at 1650 KHz (I suspect they really just need to replace the capacitors in the filter but looks like they will replace all of them in the PDM circuit which I guess is a good insurance policy).

Dec 28th, 2020: With WNTS running at reduced power I went back down to the WNTS transmitter site parking lot on Dec 28th which was 5 days after I worked with their Engineers to isolate the problem, and using my SDR receiver with a 3 inch untuned loop antenna the WNTS carrier was 75 dB above my displayed noise floor and I could not detect/see any splatter so they certainly appeared to be in compliance. It appeared reducing power caused a huge reduction in splatter, not just a linear reduction in the level of splatter.

Problem Fixed

Oct 4th, 2021: Kim (the Chief Engineer) at WNTS reported that all of the Electrolytic Capacitors in the WNTS transmitter have been replaced and he sent the following Spectrum Analyzer screen shot showing the WNTS transmitter is now clean, including the1650 KHz spur which is now over 70 dB below the main carrier.

Interesting WNTS Trivia

David Letterman began his career as a weatherman at WNTS in 1979 before he moved to LA.

Website created and maintained by Don Kirk (wd8dsb), October 25th, 2021