AM – Page 9 – Engineering Radio

WKIP

This was the radio station that I listened to (or rather, my parents listened to) when I was a very young kid. From this source, things like school closings, weather, lunar landings, news, sports and traffic could be heard. At one point, there was a guy called the “Traffic Hawk,” (real name Don Foster) who flew in a Cessna 172 east and west over main street in Poughkeepsie advising drivers of any slow downs in the area. That’s right, Poughkeepsie, New York, population 30,000, had it’s own eye in the sky, broadcasting live from the aircraft overhead. Actually, I think he also flew up and down South Road (US Route 9) in the vicinity of the IBM plant, which employed quite a few people in those days.

There was also a guy who tried to break the Guinness Book of World Records by staying awake the longest, this happened several times.

For me, it was the school closings. I hated school with an absolute passion. Everyday, I would ride the school bus and say a little prayer; “…please God, make it today. Make the boiler stop working, or the electricity go out. Make the kitchen catch on fire or the roof cave in. You are a great and mighty God and I don’t ask for much. Please destroy my school today.” Alas, God did not seem interested in this.

Anyway, back to the topic at hand.

WKIP first signed on in 1940 with the studios and transmitter located at The Nelson House, 42 Market Street, Poughkeepsie. That building is long gone and the location appears to be the parking lot for the Dutchess County Office building. Being neighbors with some influential guy from Hyde Park made for a nice dedication speech:

It signed on with a power of 250 watts on 1,420 KC on June 6th, 1940. Soon thereafter, it changed frequency to 1,450 KC as a part of the AM band shift brought about by NARBA.

Over the years, the station went through several ownership changes. The first major technical change came in 1961, when the station transmitter site moved to its current location, then called Van Wagoner Road, now Tucker Drive. The station increased power to 1,000 Watts and installed a directional antenna for daytime use. It is one of those rare nighttime non-directional, daytime directional stations.

The directional antenna consists of two towers; tower one is 180 degrees tall (103.4 Meters or 340 feet) with 35 degrees of top loading. That is used for both the day and nighttime array. Tower two is 85 degrees tall (48.8 Meters or 160 feet) and is used only for the daytime array. This pushes the major lobe of radiation towards the north. I don’t know the reasoning behind that, but somebody spends a good amount of money to make it so.

Here is an air check from the early 1980s. Weather on that day was “Sunny, cloudy, whatever… take your pick.”

Good old Steve Diner.

Today, the station looks like this:

The 1961 WKIP transmitter building with tower

When I was growing up, my cousins lived within walking distance of this. We used to come over and throw rocks at the tower when the station was unmanned on Saturdays and Sundays. At least, I think it was unmanned because no one ever came out and yelled at us.

WKIP backup transmitter, phasor and main transmitter

Mid-1980s MW-1A still runs. The BE AM1A is the main transmitter. The phasor is the Original 1960s Gates Phasor.

This video shows how the studios used to look before they were rebuilt by Clear Channel Circa 2002 or so. At about the 2:02 mark, you will see the room pictured above as it looked in 1990.

The space between the video above and the picture below looked bad with nothing in it. It looks better now.

That clock is a collector’s item and belongs in a museum.

AM station downgrade

I have been working on another formerly directional class B AM station, this one is in Rutland, VT. WSYB has been on the air since 1931 with the same call letters serving the east-central part of Vermont. In 1931, it was operating on 1500 kc with 100 watts of power. In March 1941 it moved to 1490 kc with 250 watts before settling, a few months later, on 1380 with 1,000 watts, directional night time protecting CKPC in Brantford, Ontario, Canada.

The transmitter site was first located at 80 West Street (now known as BUS US 4), in Rutland. It was moved to its current Dorr Drive (Formerly Creek Road) location in 1938, when the station was requesting a power upgrade to 250 watts. Whilst cleaning out the old transmitter building, a copy of an operating log, dated December 7, 1945 was discovered in the attic above the transmitter room:

Back from the time when readings were required every 30 minutes.

In 1956, WSYB was allowed 5,000 watts daytime non-directional with 1,000 watts nighttime directional.

At some point in the early 1990s, the original towers were replaced with solid-leg Pirod towers, each 195 feet tall.

After that, things went the way things do; AM steadily declined in favor of FM, local programming was mostly replaced by syndicated satellite stuff, there were several transfers of ownership, etc.

A translator on 100.1 MHz was added in 2016; the two-bay Shively antenna was installed at the top of the South West tower. There is local programming on the station from 6 am to noon on weekdays. There may also be some gardening shows and other such programming on weekends.

The current owner has decided like they have done in other markets, that AM directional antenna systems are a maintenance nightmare, the risk of FCC sanctions are high for an out-of-tolerance antenna array, and the ratings and income from the station do not justify the risk/cost. Thus, non-directional nighttime operation was applied for and granted. The station is now a Class D with 25 ass-kickin’ nighttime watts.

WSYB had a two-tower nighttime antenna system. The tower closest to the building (SW) was also the daytime, non-directional tower and it now holds the FM translator antenna and STL antenna. Thus, it was decided to ground that tower and keep those antennas in service. The far tower (NE), which was the second tower of the nighttime array would become the AM antenna. The nighttime ATU was built for less than 1,000 watts of input power, so several components needed to be upgraded for 5,000-watt operation.

I had available these nice vacuum capacitors that came out of another decommissioned antenna system. The vacuum capacitors are great because the voltage/current ratings are much higher than the mica capacitors that were in the circuit before. You can see black goop where one of the Sangamo mica capacitors on the input leg failed several years ago. These vacuum capacitors are rated at 15 KV and the current rating at 1.38 MHz is probably in the 70-80 amp range. I had to move the base current meter from the former daytime (SW) tower out to the NE tower. The day-night switch was taken out of the circuit. The transmission line to the far tower was replaced with 7/8 inch foam dielectric cable. A slight touch-up of the coil on the input leg of the T network was all that was required to bring it into tune.

The electric lines to the tower have been temporarily disconnected. As soon as they are reconnected, I will vacuum out all the mouse crap and other debris. The ATU building also needs some work sealing it up against the elements.

The tower base impedance is 75 ohms, +j95 making the base current 8.6 amps daytime and 0.58 amps night time.

For me, the magic of radio exists at that boundary between the real objects (towers and antennas) and the ether. The transference of electrical voltages and currents into the magnetosphere is something that still fascinates me to this day. Coupling a 5,000-watt medium wave transmitter to a tower and watching it work is something that I will never grow tired of.

The Energy Onix Pulsar transmitter

Engineering Radio: The Oh Dear God Edition.

I have been tasked with fixing one of these glorious contraptions. Aside from the usual Energy Onix quirks; design changes not reflected in the schematic diagram and a company that no longer exists, it seems to fairly simple machine. Unfortunately, it has spent its life in less-than-ideal operating conditions.

Energy Onix Pulsar 1000 in the wild. Excuse the potato quality photo — Energy Onix Pulsar 1000 in the wild. Excuse the potato-quality photo

Upon arrival, it was dead in the water. Found copious mouse droppings, dirt, and other detritus within and without the transmitter. Repaired the broken start/stop switches, fixed the RF drive detector, replaced the power supply capacitors, and now at least the unit runs. The problem now is the power control is unstable. The unit comes up at full power when it is first switched on, then it drops back to 40 watts, then after it warms up more goes to about 400 watts and the audio sounds distorted. This all points towards some type of thermal issue with one of the power control op-amps or another composite device.

After studying the not-always-accurate schematic diagrams, the source of the problem seems to be the carrier-level control circuit. This is based around a Fairchild RC4200AN (U10 on the Audio/PDM driver board) which is an analog multiplier chip. That chip sets the level of the PDM audio output which is fed into the PDM integrator circuit. Of course, that chip is no longer manufactured. I can order one from China on eBay and perhaps that will work out okay. This all brings to mind the life cycle of solid-state components. One problem with the new technology; most solid-state components have a short production life, especially things like multiplier chips. Transmitters are generally expected to last 15-20 years in primary service. Thus, transmitter manufacturers need to use chips that will not become obsolete (good luck with that), or purchase and maintain a large stock of spare parts.

In the meantime, the chip is on its way from China. Truth be told, this fellow would be better off with a new transmitter.

The isocoupler and the SX2.5

The second post in the series, “things to do with a truck body tool box.”

We have this client who, several years ago, moved their translator to their AM tower. All is well for a few months, then the much beloved Harris SX2.5 transmitter begins burping. The SX2.5 transmitter was of an age when, apparently, VSWR fold-back circuits were just a gleam in Hilmer Swanson’s eye. The correct description of the sound made over the air during this event would be “motor boating,” because that is what it sounds like. Obviously, very undesirable.

Thus, the isocoupler was removed from the tower, dried out, waterproofed, and replaced. That lasted about six months.

Once again, the isocoupler was removed from the tower, a capacitor was remounted, drain holes and a small vent were added to the top of the unit and it was replaced. That lasted about a year.

I am getting a little tired of this and so is the client. Time to rethink the entire setup.

We had several leftover parts from various AM decommissioning over the last few years which included these nifty sample loop isolation coils:

AM antenna system sample loop isolation coil

Why not repurpose one of these to make an isocoupler for the translator?

Enter; the truck body toolbox. This one is slightly smaller than the last one, measuring 23.5 x 18 x 16 inches (60 x 45 x 40.5 cm).

The isolation coil consists of 35 turns of 3/8 coax on an 11.5 inch diameter form. The coil length is 15 inches. I calculate the length of the coax on the coil to be out to be right around 100 feet using the π x D x (turns) formula. I measured the inductance with my analyzer, which came out to 200 μH. Not to shabby.

Checking length of cable with TDR — Checking the length of cable with TDR

The coax is Cablewave FCC38-50J which has a velocity factor of .81 and the TDR shows it to be 100 feet also.

Coil impedence and reactance — Simple coil impedance and reactance

At 860 KHz, the isolation coil presents 1,200 impedance. I don’t think that will be good enough for that cranky old SX2.5. I decided to make a parallel LC circuit (AKA a tank circuit) to bring up the impedance.

Tank circuit formula:

Where:

F_R = Resonance frequency in Hertz
L = Inductance in Henrys
C = Capacitance in Farads

Given that I have two leftover capacitors, one is a .001 μF and the other is a .0012 μF, those values determine where the coil needs to be tapped. I also wanted to have a good bit of coil in the circuit on the tower side before the capacitor tap to dampen any lightning strikes on the tower. Thus the inductance needs to be about 28 μH.

Using Wheeler’s coil inductance formula:

L= (d² x n²)/(18d+40l)

where:

L = inductance in micro Henrys
d = coil diameter in inches
l = is the coil length in inches
n = is the number of turns

I removed a small portion of the outer jacket on the coil at approximately the 28 μH point (12 turns) then installed a .0012 μF capacitor. I used a small variable capacitor to tune for resonance on the carrier frequency. With this setup, at 860 KHz, there is >47,500 impedance. That goes down to about 16,000 ohms +/- 10 KHz.

That should make things better.

Then I mounted the coil and capacitor in the truck body toolbox. There is a fair amount of stray capacitance from the box itself, which raised the resonant frequency by 5 KHz.

Device Under Test, initial testing of isocoil after fabrication — Device Under Test; initial testing of isocoil after fabrication

Resonance is slightly above the carrier frequency with the permanent fixed .0012 μF capacitor. I think this will change once the unit is connected to the station ground plane. The network analyzer indicated there is too much capacitance in the circuit. Unfortunately, this may be as good as it gets, however, the analyzer shows the impedances are still pretty high:

Frequency (KHz)	Impedance (Ohms)	Deviation from Carrier (KHz)
850	9,950	– 10
855	14,720	– 5
860	28,590	0
865	59,580	+ 5
870	24,780	+ 10

The base impedance of this tower is 34 ohms on the carrier frequency, so the isocoupler should be invisible to the transmitter across the 20 KHz occupied bandwidth of the station.

The FCC38-50J cable has a loss of 1.04 dB per 100 feet at 100 MHz, which is the figure I will use to calculate the insertion loss on the FM translator antenna system.

The old isocoupler is made with RG-214, but likely a somewhat shorter length. RG-214 cable has a loss of 1.9 dB per 100 feet at 100 MHz.

Installation:

Before and after measurements with the network analyzer show a very slight change in the reactance at the tower base. Nothing major and easy enough to tune out with the series output inductor of the ATU.

If I were to do this again, I would simply tap the coil at ten turns from the bottom, measure the inductance and install the proper value capacitor. Since this had to be constructed with the parts on hand, less the truck body toolbox, it because a bit cumbersome to get close to the resonant frequency.

All this got me thinking; there are other possible uses for such a design. Crossing a base insulator with an Ethernet cable always presents some unique problems. I know the WISP forum that I read, they are always talking about how difficult it is to mount an antenna on an AM tower. What if… armored Cat5e or Cat6 cable was used with waterproof RJ-45 jacks? Something like that could carry Ethernet data and DC voltage past the base insulator to a three or four around sectorized access point and an edge switch or router mounted on the tower.

Armoured category cable specifications — Armored category cable specifications

just thinking…

Anyway, it would not be hard to make coils and install capacitors for the right frequency