I have been tasked with installing one of these systems for a sixteen channel bi-directional STL. This system was first mentioned here: The 16 channel bi-directional STL system. As some of you pointed out, the unlicensed 5.8 GHz IP WLAN extension was the weak link in this system. It was not an interference issue, however, which was creating the problems. The problem was with layer two transparency in the TCP/IP stack. Something about those Cambium PTP-250s that the Wheatstone Blade hardware did not like and that created all sorts of noise issues in the audio. We installed the Wheatstone Edge Routers, which took care of the noise issue at the cost of latency. It was decided to go ahead and install a licensed link instead of the license free stuff as a permanent solution.
Thus, a Cambium PTP-820S point-to-point microwave system was purchased and licensed. The coordination and licensing took about three months to complete. We also had to make several changes to our network architecture to accommodate the new system. The PTP-820 series has a mast mounted radio head, which is the same as the PTP-250 gear. However, for the new system, we used three different ports on the radio to interface with our other equipment instead of the single port PTP-250 system. The first is the power port, which takes 48 VDC via a separate power cable instead of POE. Then there is the traffic port, which which uses Multi-Mode fiber. Finally, there is the management port, which is 1GB Ethernet and the only way to get into the web interface. The traffic port creates a completely transparent Ethernet bridge, thus eliminating all of the layer two problems previously encountered. We needed to install fiber tranceivers in the Cisco 2900 series switches and get those turned up by the IT wizards in the corporate IT department.
Andrew VHLP-2-11W 11 GHz microwave antenna
The radios mount directly to the back of the 24 inch 11 GHz Andrew antenna (VHLP2-11) with a UBR100 interface. The wave guide from the radios is a little bit deceptive looking, but I tried not to over think this too much. I was careful to use the O ring grease and conductive paste exactly where and when specified. In the end, it all seemed to be right.
Cambium PTP-820S mounted on Andrew antenna
Not wanting to waste time and money, I decided to do a back to back test in the conference room to make sure everything worked right and I had adequately familiarized myself with the ins and outs of the web interface on the Cambium PTP-820 radios. Once that was done, it was time to call the tower company.
Cambium PTP-820S on studio roof
One side of these are mounted on the studio building roof, which is a leased space. I posted RF warning signs around the antennas because the system ERP is 57.7 dBm, which translates to 590 watts at 11 GHz. I don’t want to fry anybody’s insides, that would be bad. The roof top installation involved pulling the MM fiber and power cable through a 1 1/4 inch EMT conduit to the roof. Some running back and forth, but not terrible work. I used the existing Ethernet cable for the management port. This will be left disconnected from the switch most of the time.
Cambium PTP-280S 11 GHz licensed microwave mounted on a skirted AM tower
The other side is mounted at about 85 feet AGL on a hot AM tower. I like the use of fiber here, even though the tower is skirted, the AM station runs 5,000 watts during the daytime. We made sure the power cables and Ethernet cables had lighting protectors at the top of the run near the dish and at the bottom of the tower as well as in the transmitter room rack. I know this tower gets struck by lightning often as it is the highest point around for miles.
PTP-820S RSL during aiming process
Aligning the two dishes was a degree of difficulty greater than the 5.8 GHz units. The path tolerances are very tight, so the dishes on each end needed to be adjusted in small increments until the best signal level was achieved. The tower crew was experienced with this and they started by panning the dish to the side until the first side lobe was found. This ensured that the dish was on the main lobe and we were not chasing our tails. In the end we achieved a -38 dBm RSL, the path predicted RSL was -36 dBm so close enough. This means the system has a 25 dB fade margin, which should be more than adequate. While were were aligning the transmitter site dish, a brief snow squall blew through causing a white out and the signal to drop by about 2 dB. It was kind of cool seeing this happen in real time, however, strangely enough, the tower crew was not impressed by this at all. Odd fellows, those are.
Currently brushing up on FCC part 101 rules, part C and H. It is always good to know the regulatory requirements of any system I am responsible for. As AOIP equipment becomes more main stream, I see many of these type installations happening for various clients.
This is the second Gates Air FAX-10 that I have installed. This one is in the shipping container transmitter site from the previous post of the same name. In this case, we dispensed with the equipment rack that came with the transmitter and installed it in a standard Middle Atlantic rack. The Harris rack configuration wastes a lot of space and since space is at a premium, we decided to do it our own way.
Gates Air FAX-10 in Middle Atlantic rack
The bottom of the rack has the transmission line dehydrator. The top of the rack has the Dielectric A60000 series 1 5/8 inch coax switch, a Tunwall TRC-1 switch controller and the Burk ARC-16 remote control. I cut the rack panel top to accommodate the coax switch. The racks were removed from an old studio site several years ago and were in storage since that time.
Gates Air FAX-10
The Gates Air FAX-10 transmitter on the air, running a sports-talk format.
Dummy load and Broadcast Electronics FM10B transmitter
View from the other side showing the test load and BE FM10-B transmitter. This transmitter had a problem that I have run into before with BE FM transmitters. The jumper between the exciter and IPA had the wrong phase rotation causing reflected power. I added a foot to it’s length and that problem disappeared.
Shipping container transmitter site from the early 1990’s.
I do not particularly like these. I know, they are relatively inexpensive, easy to come by, easy to install, etc. However, a shipping container was not designed to house a transmitter, they have certain drawbacks. These are, in no particular order:
- Air conditioning. Using a traditional Bard type equipment shelter HVAC unit requires cutting through a lot of fairly heavy gauge steel. What’s more, the steel walls are uneven, requiring filler.
- They are by necessity, fairly narrow. Arranging racks and transmitters along the length of the unit restricts access to either the front or the back of the equipment. Meeting NEC clearance requirements for electrical panels, transfer switches and disconnects can pose problems.
- They are not very tall. Mounting overhead equipment can be problematic as one does not want to drill through the top of the container. Crosswise unistrut is one solution, but it lowers the overhead considerably.
- Electrical work is slightly more dangerous. Doing any kind of electrical work, trouble shooting, repairs, etc is a little more nerve-racking when everywhere around you is a metal surface at ground potential.
- They are difficult to insulate against cold and heat.
- The door latching mechanisms bind, wear out or otherwise fail over time.
All of those things being said, I am now rebuilding a transmitter site in one of these shipping containers.
Inside view of shipping container transmitter site
Fortunately, the original electrical work was not bad. The transmitter is a twenty year old BE FM10B, which will be retained as a backup. The new transmitter is a Gates Air FAX-10. We have installed several of these Gates Air transmitters in the last two years or so and they seem to be pretty solid units. This is the second 10KW unit I have installed.
Gatesair FAX-10 transmitter in Middle Atlantic Rack
We decided to install the FAX-10 in a Middle Atlantic rack, since we did not have a whole bunch of extra room for a separate transmitter rack. The 1 5/8 inch coax switch is installed in the top of the transmitter rack along with a Tunwall TRC-1 switch control unit. The other rack will have the STL and all other ancillary gear. My idea is to have nothing in between the door and the FM10B so it can be easily removed when that day comes. Something, something about planning ahead since it will be likely myself removing the FM10B.
Westwood One, Premiere, Skyview Networks, et al. will be changing their satellite from AMC-8 at 139° W to AMC-18/SES-11 at 105° W longitude. More from AMC8transition.com. There are several considerations for this move:
- Dish design and two degree compliance
- Obstacle clearance
- Transponder frequencies
Two degree compliance is going to be an issue for many stations. Those old 2.4 and 2.8 meter mesh dishes are going to have issues with 105º West because that is a very crowed part of the sky. From New York, it looks something like this:
|TELSTAR 12 (ORION 2)
|TELSTAR 12 (ORION 2)
Generally speaking, dishes need to be 3.7 meters (12.14 feet) or larger to meet the two degree compliance specification. For many, this means replacing the current dish. This is especially true for those old 10 foot aluminium mesh dishes that were very popular in the 90’s because of the TVRO satellite craze.
If the existing dish is acceptable, then the next issue may be obstacle clearance. Generally speaking the 105 degree west slot (south of Denver) will be easier to see that the 139 degree west slot (south of Honolulu) for much of the United States. Still, there may be trees, buildings, hills, etc in the way. Site surveys can be made using online tools (dishpointer.com) or smart phone apps (dishalign (iOS) or dishaligner (Android)). I have found that I need to stand in front of the dish to get the best idea of any obstacles. While you are there, spray all the dish holding hardware with a penetrating oil like WD-40, Rostoff or something similar. Most of these dishes have not moved since they were installed, many years or decades ago.
Transponder frequencies will not be the same, so when the dish is aligned to the new satellite, those frequencies will need to be changed. The network satellite provider will furnish this information when it becomes available. This generally requires navigating around various menu trees in the satellite receiver. Most are fairly intuitive, but it never hurts to be prepared.
The window of opportunity is from February 1, 2017 (first day of AMC-18) until June 30, 2017 (last day of AMC-8). Of course, in the northern parts of the country, it may not be possible to install a new dish in the middle of winter. It may also be very difficult to align an existing dish depending on how bad the winter is. Therefore, the planning process should begin now. A quick site evaluation should include the following:
Network Satellite Receive Location Evaluation
Dish is 2°compliant? (Y/N)
Distance to receiver location:
Dish Azimuth (T):
Dish Azimuth (M)
Dish Height AGL:
(permanent or removable? Owned or not owned?)
A .pdf version is available here. Based on that information, a decision can be made on whether or not to keep the old dish or install a new one. We service about 25 studio locations and I am already aware of three in need of dish replacement and two that have obstructive trees which will need to be cut. This work cannot start too soon.
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
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.
In my spare time (lol!) I have been fooling around with one of those RTL 2832U dongles and a bit of software. For those that don’t know, the RTL 2832U is a COFDM demodulator chip designed to work with a USB dongle. When coupled with an R 820T tuner a broadband RF receiver is created There are many very inexpensive versions of these devices available on Amazon, eBay and other such places. The beauty of these things is that for around $12-30 and a bit of free software, one can have a very versatile 10 KHz to 1.7 GHz receiver. There are several good software packages for Windoze, Linux and OSX.
The one I recommend for beginners is called SDR-Sharp or SDR#. It has a very easy learning curve and there is lots of documentation available on line. There are also several worth while plugins for scanning, trunking, decoding, etc. At a minimum, the SDR software should have a spectrum analyzer, water fall display and ability to record audio and baseband PCM from the IF stage of the radio.
Some fun things to do; look at the output of my reverse registering smart (electric) meter (or my neighbor’s meter), ACARS data for the various aircraft flying overhead, a few trips through the EZPass toll lanes, some poking around on the VHF hi-band, etc. I also began to think of Broadcast Engineering applications and a surprising number of things came to mind:
- Using the scanner to look for open 950 MHz STL frequencies
- Inexpensive portable FM receiver with RDS output for radio stations
- Inexpensive Radio Direction Finder with a directional antenna
- Inexpensive Satellite Aiming tool
Using SDR sharp and a NooElec NESDR Mini+ dongle, I made several scans of the 945-952 STL band in a few of our markets. Using the scanner and frequency search plugin, the SDR software very quickly identified all of the in use frequencies. One can also look at the frequency span in the spectrum analyzer, but this takes a lot of processing power. The scanner plugin makes this easier and can be automated.
Analog and digital 950 MHz STL frequencies, Albany, NY
I also listened to the analog STLs in FM Wideband mode. Several stations are injecting their RDS data at the studio. There is one that appears to be -1500 Hz off frequency. I’ll let them know.
Next, I have found it beneficial just to keep the dongle and a small antenna in my laptop bag. Setting up a new RDS subcarrier; with the dongle and SDR# one can quickly and easily check for errors. Tracking down one of those nasty pirates; a laptop with a directional antenna will make quick work.
Something that I found interesting is the water fall display for the PPM encoded stations:
WPDH using RTL 2832U and SDR Sharp
Not only can you see the water marking on the main channel, you can also see the HD Radio carriers +/- 200 KHz from the carrier frequency. That is pretty much twice the bandwidth allotment for an FM station.
WDPA using RTL 2831U and SDR Sharp
Those two stations are simulcasting. WPDA is not using Nielson PPM nor HD Radio technology. There is all sorts of interesting information that can be gleaned from one of these units.
Aiming a satellite dish at AMC-8 can be a bit challenging. That part of the sky is pretty crowed, as it turns out. Dish pointer is a good general reference (www.dishpointer.com) and the Dish Align app for iOS works well. But for peaking a dish, the RTL 2832 dongle makes it easy to find the correct satellite and optimize the transponder polarization. Each satellite has Horizontal and Vertical beacons. These vary slightly in frequency, thus, but tuning to the correct beacon frequency, you can be assured that you are on the right satellite. All of the radio network programming on AMC-8 is on vertically polarized transponders, therefore, the vertical beacons are of interest. Here are the vertical beacons for satellites in that part of the sky:
||C band Vertical beacon (MHz)
||L band (LNB) Vertcial beacon (MHz)
For those in the continental United States, there is not much else past 139W, so AMC-8 will be the western most satellite your dish can see. Of course, this can be used in other parts of the world as well, with the correct information. Bringing a laptop or Windows tablet to the satellite dish might be easier than trying to drag a XDS satellite receiver out.
AMC8 vertical beacon output from LNB
In order to use the RTL-2832U, simply split the output of a powered LNB, install a 20-30 dB pad in between the splitter and the dongle. Using the vertical beacon on 949.25 MHz, adjust for maximum signal.
Some other uses; look for the nearest and best NOAA Weather radio station. Several times the local NOAA weather station has been off the air for an extended period of time. Sometimes, another station can be found in the same forecast area. Heck, couple these things to a Raspberry Pi or Beaglebone black and a really nifty EAS receiver is created for NOAA and broadcast FM. One that perhaps, can issue an alarm if the RSL drops below a certain threshold.
I am sure there are plenty of other uses that I am not thinking of right now…
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
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
The coax is Cablewave FCC38-50J which has a velocity factor of .81 and the TDR shows it to be 100 feet also.
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:
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)
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
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:
||Deviation from Carrier (KHz)
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.
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
Anyway, it would not be hard to make coils and install capacitors for the right frequency
We are extending LANs out to transmitter sites for many reasons; backup audio, control and monitoring, security systems, VOIP phones, etc.
I am casually (very casually) toying around with creating my own Linux based remote control system. The ongoing Windows 10 upgrade debacle continues to not end, I can’t help but think that there are many potential clients who could use a reliable transmitter/studio remote control and monitoring system based on a stable operating system. Hmm, sounds like a sales pitch 😉
Anyway, I have run across several Ethernet board manufactures that offer a variety of boards with 8-12 contact closures and a variety of analog and digital inputs. Most new transmitters have some sort of web GUI which are great for transmitter control and monitoring. As we all know, there is more than just a transmitter at any given transmitter site. In addition to the transmitter, I would like to control and monitor things like tower lights, interface and control of coax switches, temperature monitoring, generator status, the old non-web interface backup transmitters, STL signal strength for those old 950 MHz links, etc.
Since Google is my friend (when they are not storing my search data), I came up with this: Internet-ethernet-12-channel-relay-board
That particular PC board is made in Bulgaria, which is home to this: Mount Buzludzha
What I like about these particular boards is the DRM software (DRM has, apparently, many different meanings) which will run on Linux or Windows. There are also iOS and Andriod applications that can be used as well. It appears that the GUI can be customized for various uses. This seems like it is written in Java, so perhaps I could have some Java expert customize it for radio use. It looks like up to 32 boards can be controlled by a single instance of the DRM software. Alarm reporting would be via SNMP trap and email.
I don’t know, there is one particular cluster of stations that needs new remote control gear at almost every transmitter site. Perhaps a little alpha testing is in order? It could be fun…
Anyway, just a thought…
Alternate title: Building and ATU in a truck body tool box.
Alternate title II: I should get paid extra for this shit.
There is an AM radio station that is near death but the owners do not want it to go away. Nor to they want to spend very much money to keep it around, thus the dilemma. At the transmitter site, there are a multitude of problems; leaking roof, very old rusty ATU, rotting support posts and transmission line bridge, equipment racks rusting out, nothing is grounded properly, the building is full of junk, snakes and mice have moved in. To further complicate things, the tower and transmitter building serve as an STL relay point for two of the market’s FM stations. There is also two translators with antennas on the tower. The ATU and tower light choke box are rusting through, which is causing arcing and broadband RF noise that is interfering with the FM station’s STL receiver. There was a home made isocoupler for one of the translators that was allowing AM RF back into the building which was creating havoc with everything. Because of this, the AM station is currently silent. In short, it is a mess.
The red box on the bottom is the ATU, the plywood box on the top with the peeling yellow paint is the home made isocoupler, the tower light choke box is behind the isocoupler.
Crumbling old ATU output capacitor in series with tower
This was the capacitor that was feeding the antenna, .0041uf, 10KV 8 amps.
We started remediation on this last February, which is not optimum time for replacing rotting wooden posts. However, we were able to clean out the building. The leaking roof has been repaired. I was able to find a few old racks from a Schafer Automation system to replace the rusted out original racks. I began the process of grounding the equipment racks, the incoming transmission lines for the STL, etc.
Cool morning, Garter Snakes warming themselves on top of a Moseley DSP-6000
We will have to find out how they are getting in, the plug up those holes.
Then there was the ATU and tower light choke enclosures. Original to the 1952 sign on, they were past their serviceable days. Since this is all being done on a budget and nobody wants to spend money on an AM station that has little or no listeners and even less revenue, we had a problem.
Then somebody suggested building an ATU in a truck body tool box. Well… This isn’t the Meadowlands, so if there are no other alternatives then okay, I guess. Off to Amazon to order a tool box. This particular unit seems fine, my only comment is on the gauge aluminum (or aluminium if you prefer), which is slightly thin for holding up all those parts.
Fabrication shop, ATU built in a truck body tool box
Still, the box itself is nice enough and certainly better than the old one. I was able to reuse the inductor and the Delta current meter but the old Sangamo capacitors crumbled in my hands when I removed them. I also saved the feed through bowls, J-plugs and other parts. I used some copper strap to run a good RF ground from the input to the ground connection. Overall, I am pretty pleased with the finished product. It is a little bit tight in there, but this station only runs 1 KW, so it should be fine.
Replacement ATU mounted
So, new pressure treated posts installed, the box was mounted and the transmission line connected.
Replacement ATU under power.
Reused Schafer Automation racks, much better than the 1950’s Gates racks
The reused racks are old, but serviceable and a big improvement over the old, rusting out racks. I was able to bond each rack to the ground strap that used to connect to the RCA BTA-1 transmitter. There is one more rack to install to the right of these two. That should give us more than enough rack space for this site.
The station is back on at full power and not interfering with the FM STLs or the translators. You can actually touch the rack and not get an RF burn!
We are also working on an air conditioner.
Other work at this site; cleaning out the building, replacing the tower light photocell, installing a ground buss bar, some STL lightning protectors, dress the transmission lines, etc. It is a work in progress.
This is a topic I have covered before, but it is worth doing it again for future reference. The previous post covered downgrading an AM transmission facilities for WGHQ, Kingston, NY.
This is part II of that process.
WGHQ transmitter site, towers 1 and 2 removed
The old towers have been cut up and put in a scrap metal dumpster. They are off to China to be melted down and made into a submarine or a missile or a tank or something useful like that.
The directional array had a three towers in a straight line with a common point impedance of 60 Ohms. Dropping two towers greatly changed the electrical characteristics of the remaining tower, therefore the existing ATU needed a bit of reworking to match the 50 Ohm transmitter output.
First step, correct a few deficiencies left over from the old array.
Vise grip tower feed
This vise grip RF connection has to go. The problem is where the tower erectors attempted to solder the copper tubing. That tower base plate is pretty big and I would wager they didn’t use enough heat to make the solder connection. They were probably working in the winter time, thus the “temporary” fix. This tower was put up in 1993, so that temporary fix lasted 23 years.
I removed the offending tool and soldered the connection to another part of the tower with silver solder. The smaller cross bar made a good connection point.
RF feed correctly connected to the tower
After soldering, I cleaned up and sprayed some grey primer on it to prevent rust forming where I scraped the paint off.
Next, I made an impedance measurement:
WGHQ 920 KHz tower base impedance measurement
That junk on the upper part of the graph is coming from WHVW on 950 KHz. The tower itself looks pretty good, 77.6 Ohms resistance with 130 Ohms inductive reactance. Since this is not a part of a directional antenna system, the ATU design is pretty straight forward. Given that WHVW on 950 KHz is located 10.41 miles away, a low pass filter design is optimum. A basic low pass filter T network has inductive input and output legs with a capacitive shunt leg to ground.
T network diagram
Each leg is used to match the 50 Ohm transmission line impedance (R1) to the 77.6 Ohm tower impedance (R2) and cancel out the 130 Ohms of inductive reactance. This is a vector impedance problem, much like a vector force problem in physics. Some basic arithmetic is required (always include the units):
X1, X2, X3 = √(Zin x Zout)
X1, X2, X3 = √(50Ω x 77.6Ω) or X = 62.28Ω
The value of inductance or capacitance for each leg is calculated using the basic inductance or capacitance formulas:
L (μH) = XL / 2πf(MHz)
C (μF) = 1 / 2πf(MHz) XC
Thus the input leg, or X1 = 62.28Ω / (6.28 x 0.92 MHz) or 10.78 μH
The Shunt leg, or X2 = 1 / (6.28 x 0.92 MHz x 62.28Ω) or .0028 μF
The output leg is a little different. The tower has 130 Ohms of inductive reactance that needs to be cancelled out with a capacitor. Rather than cancel out all of the inductive reactance, then add an inductive output leg, the tower reactance can be used as part of the tuning circuit. The design calls for 62.28 Ohms inductive reactance, so 130Ω – 62.28Ω = 67.27Ω, which is the value needed to be cancelled by a capacitor:
Output leg, or X3 = 1 / (6.28 x 0.92 MHz x 67.27Ω) or .0025 μF
A little Ohm’s law is used to calculate the base current for both the day and night time operations.
Ohm’s law pie chart calculator
Thus the daytime base current is I = √(P/R) or I = √(1000 W/77.6Ω) or 3.58 Amps.
Night time base current is I = √(38 W/77.6Ω) or 0.70 Amps
Current handling requirements:
Base current is calculated to be 3.6 Amps at 1,000 Watts carrier power. Allowing for 125% peak positive modulation makes it 5.7 Amps. Having safety factor of two or 11.4 Amps output leg and 14 Amps input leg.
Voltages: 353 maximum input voltage, 439 output.
Thus, 20 amp, 10 KV parts should work well.
The designed schematic for the ATU:
WGHQ ATU Schematic diagram
Putting it all together.
Since the tower looks fairly broad at 920 KHz, we are going to attempt a nice broadband ATU to match it. This station is currently programmed with a classic country format, and I have to tell you; those old Conway Twitty, Merle Haggard, Patsy Cline, et al., songs sound pretty good on the old AM radio. The Subaru stock radio has HD, which also has a nice broad IF section, thus allowing all those lovely mid-high range frequencies through.
This is the existing ATU, which I believe was built by Collins in 1960:
Existing WGHQ T network ATU
The ATU building is a little rough, but the ATU itself is in remarkable shape for being 56 years old. The input leg inductor is in the center and will be reused as is. The large Jennings vacuum capacitor at the bottom is a part of the shut leg. Its value is 2000 pF at 15 KV. The top vacuum capacitor is series output cap, its value is 1000 pF at 15 KV. The basic plan is to move the upper cap down in parallel with the bottom cap. The shut leg inductor will be kept in place to tune out any access capacity. For the output leg, I have a 2500 pF mica cap and a 10-100 pF variable cap connected in parallel. The inductor on the output leg will be removed.
After some re-work on the ATU components, I tuned everything up. The easiest way to do this is to disconnect the legs, measure them individually and adjust them for the desired reactance, which in this case is 62.28 ohms or thereabouts. The output leg was measured with the tower connected since the tower reactance is a part of the tuning circuit. The input leg was right about 10 μH. The shunt leg turned out to be about 0.002 μF. This is often the case, theoretical values are slightly different than field values due to stray capacitance and inductance in the connecting straps, etc.
This is the load, as measured at the output terminals on the transmitter:
WGHQ tower load as measured at the transmitter output terminals
Slightly asymmetric on 910 KHz, but overall pretty good. There is a fair amount of phase rotation in the transmission line due to the length from transmitter to tower (855 feet, 260.6 meter), which works out to be 0.93 wave length allowing for the 86% velocity factor of the transmission line.
Time to pack up and go home.