Random Radio Thoughts
Cris Alexander, CPBE, AMD, DRB
It was great seeing so many of you at the annual Lookout Mountain Picnic, always a highlight of the year. It’s a great occasion to catch up and to visit with folks that live in the same area but, because of busy schedules, we just don’t get to see that often. I could say the same thing about our monthly meetings, which have been well attended (but never with the turnout that we get at Lookout!). If you haven’t made a meeting lately, you really should come.
As I was visiting with some folks up on the hill during the picnic, the subject turned to broadcast STLs and how they have changed over the years. In my television years (back in the 1970s and early ‘80s), we used analog terrestrial microwave links, often in the 2 GHz band, to get audio and video between studio and transmitter site.
When I made the transition to radio, over-the-air STL links were in the 950 MHz band. Mono links were often on discrete 125 kHz channels, and early stereo links were paired mono links occupying 250 kHz of total bandwidth. Later, composite links became all the rage, with 500 kHz channels being the norm. Of course, we didn’t have all that many frequencies to choose from in the 944-952 MHz band, and in the big cities this was a problem, especially where transmitter sites were clustered on antenna farms. 950 MHz signals tend to scatter quite a bit, and even the largest grid antennas had significant off-axis lobes, which often made channel sharing a challenge.
Then along came hybrid digital aural STLs with codec add-ons like the Moseley DSP6000 feeding analog STLs. These had the digital advantages of very low noise, but they were sample-rate limited and could at best provide four channels of audio throughput. Still, they were a big improvement over straight analog discrete or composite STLs. I had issues using them over long, over-water paths (like our path from Long Beach to Catalina back in the day), probably because of continuously changing reflections off the water, but they worked well over land paths.
Full-digital STLs followed, with the Moseley Starlink and others providing true digital point-to-point links in the Part 74 aural STL band. Those sounded great (and still do – many remain in use today, and they are still in production). Their limitation was low power output, one watt standard and five watts in the high-power version. We had to use linear power amplifiers in many installations to get the required D/U ratio at the receive antenna in urban settings or to get the necessary receive signal on longer paths.
All these links were unidirectional, and for decades this was fine. Broadcasters used a number of different means for remote control, telemetry and status. 110 kHz subcarriers on composite STLs were common for outbound remote control, and 450/455 MHz Part 74 telemetry return links (TRLs) were used for telemetry/status backhaul, but there were only eight such TRL frequencies available, so only a few could use all RF paths for remote control. Others used subcarriers on their FM signals for backhaul, which was fine if the transmitter was working and on the air but absolutely useless if there was a problem. The rest had to use leased phone circuits.
At some point in the 1990s, Moseley came out with their LANLink, which was unlicensed in the 902-928 ISM band and provided a minimal amount of IP bandwidth, but that was often sufficient for remote control, RDS and PSD. The LANLink was multiplexed into the station’s 950 MHz STL antenna system, which worked okay and eliminated the need for additional antennas, transmission lines and tower loading, but there was insertion loss on both ends. Still, it was a viable solution for many stations.
As HD radio became more widespread and transmitter equipment began to feature GUI and SNMP interfaces, the need for more bandwidth grew exponentially. Transmitter sites with available internet service were in good shape, as connections could be routed over the public internet to supply the site with the needed device-specific throughput. But many sites were remotely located and there were no public internet options, so broadcasters had to fend for themselves. It was in the early 2000s that bidirectional Part 101 links started coming into use by broadcasters.
Part 101 (Fixed Microwave Services) offered many bands from 932 MHz to 95 GHz, but the most practical were the 6, 11, 18 and 23 GHz bands. While channel bandwidths up to 80 MHz were available in some of those bands, 10, 20 and 30 MHz channels were the most frequently used (and easiest to coordinate, for obvious reasons). Employing 2048 QAM, throughput in a 30 MHz channel could exceed 100 mbps, which was more than adequate in most cases.
There was, however, one problem with broadcasters using Part 101 fixed microwave links as STLs. There was an FCC rule that prohibited the use of a Part 101 fixed microwave link as the final RF link to a broadcast transmitter site. On the surface, that would seem to prohibit the use of Part 101 links altogether for STL purposes, and clearly that was the intent of that particular rule. The FCC wanted broadcast licensees to stay in their own lane, so to speak, and utilize the Part 74 frequencies that were dedicated to broadcasters for that purpose.
In my company, we began using Part 101 links in 2010, and we got around the “final RF link” rule by employing an 802.11 5.8 GHz unlicensed link between the Part 101 radio at the tower and the network at the transmitter building. In doing so (and we specifically told the FCC what we were doing), we used the 802.11 devices as the “final RF link” and were in full compliance with the rule. That 802.11 link also gave us a way to cross the base insulator of the “hot” AM tower supporting the Part 101 antenna, so it served a dual purpose.
Thankfully, the “final RF link” prohibition was deleted in the early 2010s, and broadcasters are now free to use Part 101 links as STLs without limitation. And while we continue to use 802.11 links in many of our AM and FM facilities, we have made the move to fiber in many others, providing a direct but RF and DC isolated path between the on-tower microwave radio and the network at the transmitter site.
There are some caveats with Part 101 links. First, there is a specific procedure, outlined in §101.103, for frequency coordination. This process is called PCN, for Prior Coordination Notification, and it requires “…prior coordinat[ion] with existing licensees, permittees and applicants in the area, and other applicants with previously filed applications, whose facilities could affect or be affected by the new proposal in terms of frequency interference on active channels, applied-for channels, or channels coordinated for future growth.” This is a labor-intensive process for individual applicants, who would need database search and interference analysis tools to determine what frequencies/bandwidths are available for a particular path (in both directions) and who those “licensees, permittees and applicants” are. Thankfully, there are commercial frequency coordinators out there, such as Micronet Communications and Comsearch, that will do all this for you for a very reasonable fee.
Another “gotcha” is that in areas where there are DOD facilities, the 18 GHz band may well be off limits, and you may not find this out until your link is already up and running. I know this because in Denver, where we have Buckley Space Force Base, I prior coordinated, filed for and was granted an 18 GHz path and placed the equipmentrder. As I was awaiting delivery, I got a notice from the FCC that the DOD had late responded to the PCN notice and objected to any 18 GHz use in the area. The grant was rescinded. Thankfully I was able to change to an 11 GHz frequency and convert the equipment order before any antennas or radios shipped. So while 18 is an attractive option for typical broadcast STL path lengths, beware!
We now use Part 101 links all over our company as primary STL links; we have all but abandoned all our Part 74 950 MHz links. We need the throughput and bidirectional connectivity at most all our transmitter sites, FM and AM.
Here in Denver, we use mostly Cambium PTP820S 11 GHz links, but we do have one Trango 11 GHz link to Lookout Mountain. We also have one Cambium PTP450i link to our Englewood transmitter site. This arrangement puts all our transmitter sites on the studio technical network and eliminates the need to route anything, except at Lookout where we employ a dual-WAN router to permit seamless connectivity either over the Trango link or the internet. We use public internet backups on all our microwave links.
In the Los Angeles area, we have a mountaintop tower site that has no utilities other than power, which we brought in ourselves at great expense. We use an 11 GHz Cambium link there to provide everything – STL, remote control, SNMP, equipment monitoring/control, security, video surveillance and telephone. We use multimode fiber to connect the on-tower Cambium radio to our network.
There are many equipment options for Part 101 links, including Moseley, Trango, Cambium, Dragonwave and others. There are IDU/ODU packages with both indoor and outdoor (antenna-mounted) units, all IDU systems that employ waveguide runs to the antenna, and all ODU systems where the radio mounts directly to the antenna and is either powered directly with usually -48VDC or PoE. Most offer hot-standby options where two radios are mounted to a single antenna, although this has a 3 dB loss.
I personally prefer the shelf-standby option where a fully configured, ready-to-hang radio is on the shelf and available for each band in use. While lightning seldom causes damage to tower-mounted radios, it can happen, and if you have a hot-standby radio on the tower, chances are that it, too, will be blasted along with the primary. Shelf standbys are immune to this, but you will need a tower climber to swap out the radio. Still, that’s better than waiting months for delivery of a replacement radio or repair.
STLs have come a long way in my 47-year broadcast engineering career. We’re in a good place now with lots of options. Keep those options in mind, and don’t think you’re limited to the few Part 74 frequencies.