Engineering Radio – Page 54 – One Hundred percent Human Generated Content for your Reading Enjoyment

The efficacy of the computer generated voice

I was just listening to the latest broadcast of severe thunderstorm and tornado warnings rolling in across WXL-37 for upstate NY:

It looks a little bit hairy to the north. There is a lot of rumbling around to the west of us and we are prepared to head for the basement in event of a tornado in this area.

At some point in time, somebody decided that computer-generated voices were exactly right for emergency communications. Never mind some of the quirks that can be encountered. These are mostly pronunciation errors for places like Saugerties, normally spoken as Saw-ger-tees but the NOAA computer voice says S-ouw-jer-tees. That is understood well enough, but frankly, there are other place names that go by so fast that I cannot make sense of what the computer is saying.

Another good example of this is the Coast Guard’s computer voice confusion around the word “November.” In the military (NATO) phonetic alphabet, November is the word used to express the letter N. For some reason, the word itself seems to be a bit of a mystery to the computer, which sometimes renders the word November as “NOVEMBER OSCAR VICTOR ECHO MIKE BRAVO ECHO ROMEO.” For those of us who have been in the military, this makes perfect sense. Why just say “November” when you can say much more, waste time, and confuse the un-aware? This particular computer voice is nick-named “Iron Mike.”

Computer-generated voices can be hit or miss.

Then there is the computer voice from Shannon VOLMET:

Even on HF Single Side Band, that voice is clearly more understandable than the NOAA voices in use today. The issue is, many broadcast stations now use the NOAA computer voice to broadcast weather alerts to their listeners. If I were driving in my car with lots of background noise, I likely would not get most of the information being relayed by the broadcast station via EAS. I suppose gone are the days of a professional broadcaster’s voice clearly imparting information and comforting the listeners during times of calamity. Sigh.

The rotary phase maker

I alluded to this in an earlier post: Open Delta three phase service. Some transmitter sites are fairly remote and three-phase power is not available. Occasionally, with lower-powered radio stations, this is acceptable because those transmitters can be configured to run on single-phase power. However, almost any transmitter above five kilowatts or so will require three-phase power. This is the case at the WQBJ transmitter site in Palatine Bridge, NY. The site is located in the middle of farmland and only has single-phase service. The nearest three-phase service is several miles away and the utility company wants several hundred thousand dollars to upgrade the line.

WQBJ transmitter site electrical service

The station is a class B FM with a six-bay full wave-spaced antenna. Even so, the TPO is 17 KW, which makes some type of three-phase service a requirement.

WQBJ six bay Shively 6810 antenna — WQBJ six-bay Shively 6810 antenna

The main transmitter is a Broadcast Electronics FM30B, which is now 25 years old.

WQBJ main transmitter, Broadcast Electronics FM30B

The backup transmitter is a CSI FM20T, which is almost forty years old.

Rather than do an open delta service, which is not desirable for several reasons, both transmitters have their own rotary phase makers. From a reliability and redundancy standpoint, this is the right way to equip this site. The rotary phase makers are essentially a motor generator combination which takes the split phase power and generates a third phase.

WQBJ phasemaster, backup three phase converter — WQBJ Phasemaster type T, backup three-phase converter

The phasemaster is is a 40 KVA unit and is connected to the backup CSI transmitter

WQBJ Roto Phase, main three phase rotary converter — WQBJ ARCO Roto Phase, main three-phase rotary converter

The Roto Phase unit for the main transmitter is actually two 40 KVA units connected in parallel through dry core isolation transformers. Incidentally, the Roto Phase units need to have their bearings changed every ten years or so. This requires the units be disconnected, placed up on their end. To get the old bearing out, the housing has to be cooled with liquid CO2. Both units are due for new bearings soon, which should be a pleasant job indeed.

Repairing the Nautel VS2.5 transmitter

The newish Nautel VS2.5 transmitter installed at WJJR had an RF module failure. This particular model transmitter does not have slide-in RF modules as other Nautel transmitters do. To fix this transmitter, it has to be pulled out of the rack, flipped over, and opened from the bottom. The module replacement is very straightforward, there are five solder pads that connect to wires carrying the input, output, power supply, and bias voltages.

Nautel VS2.5 transmitter RF modules and combiner

The troubleshooting guide gives good instructions on how to check the PA MOSFETS with a DVM. I found that 1/2 of the device in PA1 was bad:

All in all, not a very hard repair. This was under warranty, so a replacement RF pallet was sent to the station without charge. The problem is more about where the transmitter is located:

Killington Peak is the second tallest mountain in Vermont, topping out at 4,235 feet (1,291 meters). In the winter, one can take the chair lift to the top. In the summer, the road is drivable with a four-wheel drive. In those in between months, access to the top can be very tricky at best. We had a pretty wet spring this year, so the roads up the mountain are just now becoming passable for vehicles.

Even after reaching the parking lot, there is still a 10 minute walk to the peak, another 200 or so feet up a steep, rocky trail.

Further complicating things, this transmitter is wedged into this little shack, which holds; a BE FM3.5A transmitter (defunct WJJR), a Harris HT3 transmitter (WZRT), an ERI combiner, two racks of equipment (STLs, Exciters, remote controls, etc) a backup QEI transmitter, an Onan generator transfer switch:

Killington Peak fire tower, WJJR WZRT transmitter building

Both stations run into this ERI half-wave spaced antenna:

It is very tight in this transmitter room. There is a new tower on Killington Peak, which is still under construction. At some point, the plan is to move into the larger building next to the new tower.

On a clear day, the view from the top is spectacular. On this day, the peak was in the clouds, so not so much:

It is a great site, the HAAT is 2590 feet (790 meters) and the stations carry forever on relatively low power outputs.

Fifth Generation WLAN

Like all data-carrying technology, WLAN, or WiFi, continues to evolve into a better, faster, and more robust platform. The IEEE wireless ethernet specification 802.11ac combines all of the past developments, plus some added features, into one specification. Here are some of the highlights:

Operation on 5 GHz only. Many more available channels in this spectrum than in 2.4 GHz
Increased channel bonding makes wider channels carry more data. In the 5 GHz spectrum channels are 20 MHz wide and do not overlap. 802.11ac allows for 40, 60, 80 or even 160 MHz channels. This is great for short distances, longer distances will be prone to greater interference over wider channels
Modulation schemes that allow up to 256 QAM. A 256 QAM constellation is going to look pretty crowded unless it is on a wide channel. Again, this would be good for short distances.
Increased MIMO. Up to 8×8 MIMO (Multi In Multi Out) which can greatly improve throughput. MIMO means multiple transmitters and antennas in the same unit. The first number is the transmitter count the second number is the antenna count. Thus an 8X8 system will have eight transmitters and eight antennas. This allowed beam forming by use of phased antenna arrays, which can greatly reduce multi-path
MU-MIMO (Multi-User MIMO). Basically, the access point sends the data frame only to the desired host, thus instead of acting like an ethernet hub sending the frame to every connected host, the AP is acting more like an ethernet switch.

The goal of all of these modifications is to get gigabit transfer rates over WLAN.

What does all of this have to do with radio broadcast, one might ask? That is a good question.

There are several applications that have to do with remote broadcasting. Many sports areas, nightclubs, or other likely places to be broadcasting from have WIFI installed. Using a laptop with an AoIP client installed not only can connect to the studio for audio delivery, but the same laptop can also use RDP or VNC to control the station’s automation computer as well. This means easier integration of the remote into voice-tracked or syndicated programming.

Secondly, wireless LAN bridges between the studio and the transmitter site can act as an STL, a backup STL, a remote control return link, a bridge for a network-connected transmitter, a VoIP phone link, IP security camera backhaul, or almost anything else that can send ethernet data. I have found it useful to simply have a computer available at the transmitter site, even if it is only to download manuals and whatnot. We have taken several old Windows XP machines and reloaded them with a Linux variant and installed them at various transmitter sites. It saves the trouble of having to download a manual on the smartphone and then page back and forth across a really small screen to read it. As for using unlicensed WiFi to link to a transmitter site; the link between the WICC studio and transmitter site runs 78 Mbps most days. This is a two-mile link over mostly water. I will say, when there is fog, the link rate drops to 32 Mbps, which is still pretty good, all things considered.

Of course, office network applications; laptops, tablets, smartphones and other personal devices.

Finally, Broadcast Engineers really need to keep abreast of networking technology. There are many, many applications for WiFi units in the broadcast industry.