A tale of five signals

I am currently finishing an interesting project involving putting up two translators on a diplexed AM tower which also holds a mobile phone/data tenant as well.  All-in-all, this seems to be a very efficient use of vertical real estate.

WMML WENU tower, Glens Falls, NY
WMML WENU tower, Glens Falls, NY

The AM stations are WMML and WENU in Glens Falls, NY.  The AM stations are diplexed using a Phasetek diplexor/ATU.

Diagram showing WENU/WMML tower with W250CC/W245DA antenna installed
Diagram showing WENU/WMML tower with W250CC/W245DA antenna installed
Diplexor diagram, WENU/WMML Glens Falls, NY
Diplexor diagram, WENU/WMML Glens Falls, NY

The translators are W250CC and W245DA which are using a NICOM BKG-77/2 two bay 3/4 wave spaced antenna mounted at 53 meters AGL.  The translators use a Shively 2640-04/2 filter/diplexor which is a broad band input port in addition to the translator input ports.  Since these translator signals are only 1 MHz apart, the higher-power Shively filter was installed because it has better rejection characteristics.  The broadband input port allows the NICOM antenna to be used as a backup for any of the three FM stations; WKBE 107.1, WNYQ 101.7, or WFFG 100.3.  Two transmitter sites for those stations are mountaintop locations which are very difficult to get to in the wintertime.  Having a backup site available takes some of the pressure off during storms or other emergencies.

Shively 2640 -04/2 filter for W250CC and W245DA

The NICOM FM antenna was mounted on the tower when W250CC went on the air in October of 2016.  When it was installed, the base impedances for both AM stations were measured.  For some reason, WENU 1410 KHz seems to be more sensitive to any changes on the tower, thus the WENU ATU needed a slight touch-up.  When working on diplexed AM systems, it is also important to make sure that both trap filters, which are parallel resonant LC circuits, are tuned for maximum rejection of the other signal.  During this particular installation, nothing was added to the tower and no change in the base impedance for either station was noted.

Shively Filter, connected to transmitters and antenna
Shively Filter, connected to transmitters and antenna

As a condition of the construction permit, measurement of spurious emissions of all stations sharing the common antenna needed to be completed to ensure compliance with FCC 73.317(b) and 73.317(d).  I made careful measurements of the potential intermod products between the two translator frequencies.  This measurement was completed with my TTI PSA6005 spectrum analyzer.

The primary concern here is mixing products between the two transmitters. Both transmitters are BW TXT-600 with low pass filters before the output connector. There are three frequencies of interest;

  1. (F1 – F2) + F1 or (97.9 MHz – 96.9 MHz ) + 97.9 MHz = 98.9 MHz
  2. F2 – (F1 – F2) or 96.9 MHz – (97.9 MHz – 96.9 MHz) = 95.9 MHz
  3. F2 + F1 or 97.9 MHz + 96.9 MHz = 194.8 MHz

That, plus harmonic measurements out to seven or eight harmonics of the fundamental frequency should be enough to demonstrate compliance with FCC out-of-band emissions standards. Being that this site has LTE carriers, it is very important to measure the harmonics in those bands. Mobil data systems often use receiver pre-amps, which can amplify harmonics from the FM band and make them look out of compliance. Having a base set of reading to fall back on is always the best course in case the “out of tolerance” condition gets reported to the FCC.

Measurements on these frequencies must meet the emissions standards outlined in FCC 73.317 (d), which states:

Any emission appearing on a frequency removed from the carrier by more than 600 kHz must be attenuated at least 43 + 10 Log10 (Power, in watts) dB below the level of the unmodulated carrier, or 80 dB, whichever is the lesser attenuation.

Harmonic frequencies to be measured:

Harmonics for 96.9 MHz fundamentalHarmonics for 97.9 MHz fundamentalComments
193.8195.8 
290.7293.7 
387.6391.6 
484.5489.5 
581.4587.4 
678.3*685.3*US LTE Band 71
775.2*783.2*US LTE Band 5
872.1*881.1*US LTE Band 5
969.0979.0 

*Frequencies that fall within the mobile data LTE bands. Traces were recorded and saved for these frequencies.

All of that information, once compiled is attached to the FCC form 350-FM, which, once filed grants Program Test Authority.

BW TXT-600 V2 translator transmitters
BW TXT-600 V2 translator transmitters under test and measurement
Print Friendly, PDF & Email

7 thoughts on “A tale of five signals”

  1. Even though I am not a professional engineer (I’m more of an amateur) I enjoy reading articles like this. Also, nice to see those Optimod-FMs still at work!

  2. I’m betting the tower makes more in rent on a monthly basis than the radio stations combined do. 😉

  3. Love your articles. How does a skirt fed tower perform vs an insulated tower base of the same length?

  4. Unless something has changed, the diplexer schematic is slightly in error in that there were a couple of caps right up at the wall feed-thru.
    I think you and I have chatted before about how components on the tower side of the last J-plug seems not quite right.
    Kurt Gorman likes to do whatever is necessary to lower the voltages before the combining circuits, and that makes sense. But I always find myself looking at it when he does it and kinda wondering.

  5. Matt, Shhhh, we don’t want the competition to find out about our air chain processing 😉

    Scott, those AM’s are doing okay, revenue wise. I am pretty sure that combined, they are making more than the tower lease.

    Steve, Good question; insulated towers (also know as series excited towers) are generally broader bandwidth than skirted towers. That makes them easier to match. Other factors such as take off angle, skywave, etc are the same for both types.

    Bill, you are correct. I suspect the capacitors down stream of the base current meters are add-ons to cancel out greater than expected inductive reactances before the notch filters. I agree with you on the placement, too.

  6. Nicely done (as usual). Looks like a Unity 2000i in the rack adjacent to the Optimod…interesting processor.

    What CAD/drawing software do you use for your tower drawings?

  7. Allen, I have Visio which I like as it is not overly complicated. Over the years, I have made or downloaded various tower stencils, etc.

Leave a Reply

Your email address will not be published. Required fields are marked *