AM station downgrade

I have been working on another formerly direction 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:

WSYB transmitter log, 1945

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 night time directional.

At some point in the early 1990’s, 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 6am 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, the ratings and income from the station do not justify the risk/cost.  Thus, non-directional night time operation was applied for and granted.  The station is now a Class D with 25 ass kickin’ night time watts.

WSYB had a two tower night time 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 night time array would become the AM antenna.  The night time ATU was built for less than 1,000 watts input power, so several components needed to be upgraded for 5,000 watt operation.

WSYB rebuilt ATU
WSYB rebuilt ATU

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 replace 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 in 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.

WSYB radiating element
WSYB radiating element

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.

Fixing another AM station’s antenna system

I have done several of these posts in the past, but it always seems to be of some interest, so it bears repeating.  AM antenna systems are not black magic.  They are actually pretty easy to understand if the fundamental knowledge is in place.  Medium Wave frequency wavelengths are fairly large compared to other broadcast frequencies.  Thus, the components are larger.

The three basic components of an AM antenna system are the tower, the ATU (antenna tuning unit) and the transmission line (AKA Coax).  The tower is the radiating element and they come in a variety of flavors; uniform cross section guyed, self supporting, series excited, shunt excited, etc.   A series excited tower has a base insulator and is fed directly from the ATU.  A shunt excited tower has a grounded base and uses a skirt or folded monopole design to transfer the RF to the main radiating element.  This design has an advantage as the tower can be used for other wireless and broadcast services.

The antenna work in question for this project is WINE, 940 KHz, Brookfield, CT.  The skirted tower is used for WRKI.  It also has two way and cellular clients.  The issue is instability of the WINE antenna system, which is likely due to improperly attached shorting wires between the skirt at the tower.  Over the years, the impedance of the skirt has gone way up.  The tower itself is 152.1 meters (499 feet) tall, or 170.3 electrical degrees.  The skirt length is about 82 electrical degrees and it is shorted at about 72 degrees.  There have been several papers written about folded monopoles for Medium Frequency (AKA AM or Standard) broadcast service.  The recommendations state that for best performance, the short to the tower should be between 62 and 90 electrical degrees.  Since the existing system falls in that range, there must be other problems with the antenna skirt and or shorting wire to the tower.

WINE skirted tower diagram
WINE skirted tower diagram

If one looks at this diagram, that configuration should look something like a gamma match, often used on dipole and yagi type antennas.  A gamma match can be thought of as a stub of transmission line which is bonded to the radiating element at some favorable wave length corresponding to the desired radiation resistance.  This is one of several configurations for folded monopole antennas and this type is most often seen on towers that support other wireless service antennas such as cellular and two way systems which are installed above the skirt.

There are a few interesting data points when looking at these type of antennas.  First is the ratio of the diameter of the skirt over the height of the tower, or D/H.  The larger this ratio is, the better the bandwidth characteristics of the antenna system are.  This makes sense, when you think about it. In this instance, the tower is 151 meters (495.4 feet) tall and the skirt is 3.3 meters (10.83 feet) wide, thus the ratio is 0.0218.

The licensed base impedance if 234 ohms with a good amount of inductive reactance. When Sprint and T-mobile changed their configuration on the tower, that impedance shifted dramatically.  The existing skirt is in fairly rough condition.  The bottom ring that connects to the ATU is made out of copper tubing.  It is attached to the skirt wires with steel saddle clamps, all are rusted and all of which are lose and can slide around.  At some point, the tubing filled up with water, then froze causing the tubing to split open.  At the top of the skirt, the jumper wire looks suspicious and the top ring does not go all the way around. The shorting stub to the tower looks like it is made out of battery jumper cable.  I purchased new cross wire clamps and found some spare copper weld skirt wire at another site.  Both the bottom ring and top ring were replaced as well as the shorting stub to the tower.

After the repair work was done, I had the tower crew reattach the short slightly below the last skirt to tower bonding point.  In that position, I found the impedance went way up.  Thus, going lower was going towards a resonance point.  I had them move the short up to the former shorting point and remeasured and found the impedance was 235 ohms, only 1 ohm off from the previously licensed values.

Initially, I thought it would be nice to find a better position for the shorting stub and get a lower base impedance.  This would make the whole antenna system work better (improve bandwidth, stability, etc).  However, there was a set of guy wires above the bonding point.  The tower crew would have had to disassemble the top ring to move above the guy wires.  We were running out of daylight and weather so I had them lock everything down where it was.  On a station running an all sports format that has no listeners and does not make any money, it does not make a lot of sense to spend gobs of money and time to rebuild the ATU for a new base impedance.  When I got the impedance back to within 0.11% of the licensed values, it was time to declare victory and go home.

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 simply 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 of 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 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 other composite device.

After studying the not always accurate schematic diagrams, the source of the problem seems to be 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 manufactures 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 mean time, 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

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 being 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, water proofed 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 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 set up.

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

AM antenna system sample loop isolation coil
AM antenna system sample loop isolation coil

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

Enter; the truck body tool box.  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 analyser, which came out to 200 μH.  Not to shabby.

Checking length of cable with TDR
Checking 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 some.

Tank circuit formula:

tank_circuit

Where:

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

Given that I have two left over 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 coil length in inches
n = is 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 set up, 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 tool box.  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:

Isocoil mounted on back of ATU
Isocoil mounted on back of ATU
Isocoil mounted on back of ATU
Isocoil mounted on back of ATU

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 where 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 tool box, 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 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… armoured Cat5e or Cat6 cable was used with water proof 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
Armoured category cable specifications

just thinking…

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