This transmitter is in service at WSBS, Great Barrington, MA as a standby. It was new in January of 1975.
Gates BC250GY transmitter, WSBS Great Barrington, MA
This was running into the dummy load for testing, which we try to do periodically.
Gates BC250GY AM transmitter audio section
The audio section is a pair of 8008′s 810′s running in parallel. This goes through a modulation transformer to the RF section.
Gates BC250GY RF section
The RF section consists of another pair of 8008′s 810′s running a parallel. The plate voltage for these tubes is 1,250 VDC which is fairly tame, all things considered. The transmitter is dirt simple 250 watt carrier power, 125% positive peak capable. It is not the most efficient unit under the sun but it can still be repaired with off the self parts.
Gates BC250GY Schematic
This is a somewhat faded schematic. The schematic shows a single 833A as the final, however, this particular transmitter has a pair of 810′s for the final, as shown in the above picture. Ham radio operators love these things as they are easy to convert to 160 or 80 meters for AM phone use. The bigger brother to this unit is the Gates BC1G, which is also pretty simple unit using 833A tubes in parallel with 3,500 VDC plate voltage.
I sometimes get the distinct impression that the corner office doesn’t understand what it takes to keep a radio station on the air and in good repair. It is most often the problems or “issues” that tend to get the most attention. The things that are working well tend to get ignored. After all, how often do you hear a news report about the airliner that landed safely.
Lightning strike, TV tower
When lightning strikes the tower and knocks the transmitter off the air causing major damage and expensive repairs, that is a problem. When lightning strikes the tower and nothing happens, no problem. What is the difference between those two situations?
Grounding strap, FM transmitter site
If the generator starts and runs during every power outage and has done so for the last five years straight, it is obviously a reliable unit, does it need all that maintenance?
Caterpillar 75 KW diesel GENSET
Money spent on preventing undesirable outcomes can be difficult to quantify as disasters and events that do not happen are ill defined. It is difficult to quantify the “amount saved” on something that didn’t or won’t occur. Using past situations is good start, but that only covers a fraction of possible outcomes. In order to invest money wisely, one has to look at the probabilities. If there is an unlimited budget, then the probability exercise should be minimal, however, there is very seldom an unlimited budget.
For example, how much does a back up STL system cost vs the risk of being off the air while the main STL system is being repaired? How often do failures occur, when are they likely to occur and for how long are all good questions. Is there an alternative to a full back up like an IP CODEC? Such a solution would cover all aspects of the STL system including antennas, transmission line, transmitters and receivers.
There are certain FM stations north of here that have neither RADOMES or antenna heaters. Once every two years or so, the antenna ices up and the transmitter folds back due to VSWR. How much of an impact to listeners notice when this happens? If it happened more often, say two to three times a year, would it be wise to invest in some type of deicing equipment?
What is the ownership and management opinion on off air conditions? I have often heard tell “Oh, its only the AM, we don’t mind if it goes off the air.” That is, until it actually goes off the air, then it is a big problem.
Based on my and others experiences, these are the things that will happen at an average transmitter site:
The electric will go off at least once per year for several hours.
The main transmitter will fail at least once every two years.
Lightning will strike the tower at least once per year.
The STL system will fail, at unknown intervals.
At studio sites, these things will occur:
The file server will crash depending on the operating system
The telephone lines and or T-1 service, internet service, ISDN etc will go out
The electric power will go out for several hours
The satellite dish will fail once every two to three years
If there is a tower, it will get struck by lightning
Other site specific things can occur like floods, blizzards, earthquakes, fire, etc.
Money spent on backup systems for those items is good insurance. Not only will the station stay on the air, the on call engineer’s phone will ring less often, which, if you are the on call engineer, should make you happy.
If a full backup is not available, a second transmitter for example, having a good stock of spare parts on hand can mean the difference between an early evening and an all nighter. Keeping good maintenance logs and well documented repair records can point out trends and give a good basis for ordering spare parts.
Repair trends are important. If the same part seems to be going bad over and over, it is time to dig deeper and find the cause of failure.
The old adage “An once of prevention is worth a pound of cure,” still holds true.
They say the first thirty years are the hardest, perhaps it is true. This Harris MW1A transmitter turns 31 this year:
Harris MW1A AM transmitter, WINE, Brookfield, CT
It is on the air as the main transmitter for WINE-AM in Brookfield, CT. These are not necessarily bad transmitters, although they do seem to require regular infusions of MOSFETS to stay at full power. This is Harris’s first solid state AM transmitter design, based on the work of Himmler Swanson. This is not a PDM transmitter, rather, each module has RF MOSFETS and audio transistors. The output of all twelve modules are combined for a carrier output of 1,ooo watts with +125% modulation. Thus, I would call it low level AM modulation.
This is also the only transmitter that I know of where blown fuses can cause damage to the RF devices.
The RF output transistors and audio transistors are still available from Harris. Non-OEM parts for this include the 2N5038 for the RF MOSFETS and the MJ15011 for the audio transistor. Inside the front of the transmitter is a row if incandescent light bulbs that glow increasingly as the various MOSFETS go bad. At 1,000 watts carrier power, the ratio of PA volts to PA amps is 52.5/22.5 respectively. If that ratio is off by any measure, there is a problem.
The other thing with this transmitter is it is very sensitive to any kind of VSWR. Any change in the output impedance will quickly make itself apparent. My Harrisburg MW1A had two ATU settings, one for winter and one for summer. It was a slightly tall tower on 1230 KHz, thus any change in the ground system (e.g. snow cover) would upset the tower base impedance.
The other thing that goes bad is the large Rotron fan in the bottom of the cabinet. They go bad about every 10-15 years or so.
The owner has spent some money on this particular unit, rebuilding and replacing several modules with new MOSFETS etc. Will it last another thirty years? Depends on if the RF and audio devices remain in production.
Man, this is taking longer than I though it would. We moved the Harris FM25K last week, all went well. The only hangup, as you can see, is the harmonic filter and the height of the racks next to the transmitter. The transmitter had to go on a 4×4 to get the filter up over the racks. The output from the transmitter to the harmonic filter cannot be changed in any way, shape or form (e.g. adding a little bit of line section to the top of the transmitter), else the transmitter will not run. So, up on 4×4′s it is.
WRKI WINE transmitter room
There we were, all ready to turn the transmitter on. Press the high voltage on button, lots of volts but no current and no power output. Seems something is wrong with the outboard IPA driver (over in the bottom of the rack, that thing pulled out with the manual on it).
The IPA is a Silicon Valley Power Amplifier 500 watt unit, which replaced the internal IPA driver about ten years ago. The tube in the Harris FM25K needs at least 390 watts to drive the transmitter to full power. Unfortunately, this particular amplifier was not in the best environments prior to the recent move. It was sitting in an unconditioned building on top of the backup transmitter in high heat and humidity. According to the manufacture, such abuse is bound to take it’s toll sooner or later. The later being, of course, the night we want to turn the thing back on and go home.
Time to drop back and punt. I found an old RVR 250 watt amp at a sister station nearby, which was also in pretty bad shape but repairable. That unit was pressed into service temporarily and with 200 watts drive, the old 25K put out about 11 KW. We need to affect permanent repairs to the RVR power amp before we place into temporary service. I don’t want any 2 am phone calls. The Silicon Valley Power Amp needs to have the amplifier module sent back to the manufacturer and rebuilt. They will refurbish the entire thing for something like $900.00 plus shipping. Considering what it does, that is worth it.
This is a little short cellphone video of the turn on at half power. This is a very loud transmitter, as such, I think the audio is a little distorted.
When this beast gets up to full power, I will update this, again.
Every good transmitter, tube transmitters in particular, require harmonic filtering. The last thing any good engineer or broadcaster wants is to cause interference, especially out of band interference to public safety or aviation frequencies. All modern transmitters are required to have spurious emissions attenuated by 80 dB or greater >75 Khz from carrier frequency. In reality, 80 dB is still quite high these days, especially in the VHF/UHF band, where receivers are much more sensitive than they used to be. A good receiver noise floor can be -110 dB depending on local conditions.
The principle behind a low pass filter is pretty easy to understand. The desired frequency is passed to the antenna, while anything above the cut off frequency is restricted and shunted to ground via a capacitor.
Low pass RC filter
In this case, the resistor is actually an inductor with high reactance above the cut off frequency. Often, these filters are lumped together to give better performance. This is a picture of an RVR three stage low pass filter:
RVR three stage low pass filter
RVR is an Italian transmitter maker that sells many transmitters and exciters in this country under names like Bext, Armstrong, etc. The inductors are obvious, the capacitors consist of a copper strip sandwiched between teflon insulators held down by the dividers in between the inductors.
Schematically, it looks like this:
Low pass filter schematic diagram
For the FM broadcast band, a good design cutoff frequency would be about 160 MHz. This will give the filter a steep skirt at the first possible harmonic frequency of 176 MHz (88.1 x 2 = 176.2).
Values for components:
Capacitors
Value
Inductors
Value
C1
20 pf
L1
74.7 nf
C2
54 pf
L2
75.1 nf
C3
54 pf
L3
73.9 nf
C4
20 pf
The inductors are wire, or in this case copper strap, with an air core. It is important to keep the transmitter power output in mind when designing and building these things. Higher carrier powers require greater spacing between coil windings and larger coil diameters. This particular filter is rated for 1 KW at 100 MHz.
I have been spending my days in Brookfield, Connecticut, dragging transmitters around and reconnecting them in various ways. The WRKI-FM WINE-AM transmitter site is finally moving into the “new” transmitter building at the base of the tower. Today, we moved WINE.
WINE was first signed on in 1963 on 940 KHz from a 170 degree non-directional tower on top of a pretty high hill. That same tower serves as the antenna support for WRKI, which signed on in 1957. The station runs 680 watts daytime, however since it is non-directional, it has some pretty serious power reductions at night. The post sun set power drops in two steps, 450 watts for the first hour, then 189 watts for the second hour, followed by 4 watts night time.
The 4 watt night time signal goes about 2-4 miles before it becomes unlistenable. The Post Sun Set Authority (PSSA) allows the station to stay on the air with at least some coverage up to about 6:46 pm in the winter time and 10 pm in the summer, which is better than nothing.
The problem is, the Harris MW-1A transmitter goes down to 250 watts and no lower. In order to make the night time power, the station switches to a dissipation network to burn off 246 watts of RF, at 50% percent AC-RF efficiency, which just ends up being a waste of power. Further, the station engineers have been ignoring the PSSA because there are too many steps and it was easier to just switch to night power at sunset.
What we decided to do instead, was install a small low power night time transmitter, a Radio Systems TR-6000. The MW1A can then be set to use the low power level for the first step of the PSSA, then switch the dissipation network in for the second step of the PSSA, finally switching in the night transmitter at the proper time.
Harris MW1A AM transmitter, WINE 940 KHz, Brookfield, Ct
This is the Harris transmitter, new Circa 1981, was cleaned up and moved into the new transmitter building.
WINE Parallel dissipation network and dummy load
The dissipation network. This will have to be reconfigured for the proper power levels, once the night transmitter is installed. The dissipation network is on the right, a dummy load is on the left. The two large RF contactors switch the dissipation network in and out, or select which transmitter is feeding the antenna/dummy load. This is the really, really old school way of doing it. Most transmitters manufactured after 1990 or so can run at any power level, making a dissipation network unnecessary.
Before re-installing the dissipation network/dummy load, we lined the enclosure with copper mesh. I don’t want that thing interfering with any of the other equipment nearby, which would be the STL receivers, satellite receivers or Town of Brookfield police dispatch radios.
Schematically, it looks like this:
WINE 940 KHz Brookfield, CT night time dissipation network
This is the picture behind the transmitters, shows the coaxial cable feed through ports and the dissipation network on the wall.
WINE WRKI transmitter room, behind the transmitters
How effective are they at filling in or expanding coverage for FM stations? The answer is, it depends. Most have heard of the quadcast around New York City on 107.1 MHz formed in 1996-98. It was well documented in Radio World and several other publications as a cleaver way to overcome the suburban rim shot problem. Four signals on 107.1 were synchronized using GPS timing data, then fed the same program material. They were WYNY, Braircliff Manor, NY; WWXY, Hampton Bays (Long Island), NY; WWYZ, Long Branch, NJ; and WWYY Belvidere, NJ. These being four separate Class A FM stations, the 60 dBu contours did not overlap. There was some mutual interference in some areas, but there were few if any reception negative zones where the signal strength is equal between stations.
In early 2003, I was a part of the disassembly of the quadcast. In the end, it is difficult to point to any one thing that lead to the breakup. The station’s owners, Big City Radio, had filed for bankruptcy. I am not sure if the company ever had the correct formula for marketing and sales, given the strong suburban, but weak and lacking building penetration in Manhattan signal. The station initially had a country format, something that arm chair quarterbacks said would not work in New York City. After a few years, Big City had changed the format to Rumba, a Spanish/Caribbean music format, which did worse than Country. The fact is, that it never lived up to expectations and the station were worth more separately than together. Given the right circumstances, it could have worked.
The other synchronized FM broadcasts are those where boosters are employed. These are a good deal more difficult to configure because the booster signal is within the main stations 60 dBu contour. Often cases, where there is severe terrain shadowing or other limitations, a well positioned booster that is in a population center can greatly improve the signal in those areas. This was formerly the duty of an FM translator, however, those stations seem to be taking on a life of their own, without regard for the intent of the current FCC rules. Boosters can also be called a single frequency repeater or single frequency network (SFN).
The disadvantages of a SFN are the aforementioned negative reception areas. To the receiver, this will create a multipath or picket fencing situation, which is objectionable to most listeners. The advantages are, of course, better coverage in key areas, spectrum efficiency, and the ability to create a network of common frequency systems. Think of how easy it would be if all NPR stations were all on the same frequency, for example.
The key to making a booster work is to synchronize several aspects of the RF and Audio signals:
RF carrier frequency
Stereo pilot frequency and phase
Audio amplitude and phase
This is best done using GPS receivers to synchronize the exciters and an AES/EBU audio path from the studio to both transmitters fed by one processor. Once this is accomplished, a certain amount of delay can be added to the audio content on either the main or booster transmitter to move the interference zones away from heavily populated or trafficked areas.
WDBY, Patterson, NY 60 dBu contour
This is the situation with WDBY in Patterson, NY. The main transmitter site is located on a hill in Patterson and has a power level of 900 Watts at 610 feet (186 meters) HAAT. The main population area is Danbury, CT, to the south east, about 12 miles away. Between the two, there are several imposing hills, which create reception issues in Danbury. Therefore, WDBY FM1 was placed in service on the Danbury Medical Center. The booster has a power output of 1,200 Watts, at 0 feet (0 meters) HAAT (49 meters AGL).
WDBY FM-1 signal, Danbury, CT 60 dBu contour
Therefore, the southern area of the 60 dBu contour is filled in by the booster. The interference zone between the two transmitters is determined by the amount of delay in the audio between the two units. If both are time the same, the interference will occur at precisely 1/2 the distance between the transmitter sites, which in this case is 10.18 KM from booster. Looking at the population maps, it might be better to move that more toward the north, away from Danbury.
The formula for computing audio delay time is:
A-B=C where A is the distance between the transmitters and B is the distance to the interference zone from any given transmitter. The product of that is multiplied by a constant of 3.34 to obtain the time delay in micro seconds. Therefore, if the interference zone is desired to be further outside of Danbury, say 15 KM away, then the equation looks like this:
20.358 kM -15.0 kM = 5.358 KM
5.358 KM x 3.34 = 17.89 μS delay from the main transmitter site will put the interference zone out in the middle of nowhere, away from Danbury. This is total delay between the two stations, therefore any difference in STL paths needs to be included in this figure.
Nautel equipment has most of these features built into it, therefore, the implementation of a SFN using Nautel exciters and transmitters should be relatively straight forward.
It may be surprising to some, but number of wires allowed in any given conduit is not “as many as can be jammed in there.” The National Electrical Code, AKA NEC or NFPA 70 gives specific guidance on the numbers of current carrying conductors allowed in any specific size and type of conduit.
This is due to the fact that current carrying conductors generate heat. Cables enclosed in a conduit need to dissipate that heat so that the insulation on the cable doesn’t melt, which would be a bad outcome.
Conduit fill tables are found in Chapter 9 of the NEC. There are several tables that give the number of conductors for each size and type of conduit. Then there is the general rule of thumb that more than two cables, the maximum conduit fill is 40%. This comes in handy when several different size conductors are being run in the same conduit.
An example of this is when several circuits are going across the room to the same general location, in this case, a row of transmitters and racks. Instead of running individual conduits for all those units, one or two conduits from the electrical panel are run to a square wireway, then the individual circuits are broken out and wired from wireway to the individual loads. In this case, the following equipment is being connected:
Harris FM25K: 100 amp 3 phase high voltage power supply (#2 THHN), 30 amp 3 phase transmitter cabinet (#10 THHN)
Harris FM3.5K: 70 amp split phase (#6 THHN)
Harris MW1A: 30 amp split phase (#10 THHN)
Two equipment racks: 20 amp single phase (#12 THHN)
Coax switch: 15 amp single phase (#14 THHN)
Dummy Load: 15 amp single phase (#14 THHN)
Antenna switch/dissipation network for AM station: 15 amp split phase (#14 THHN)
Convenience outlets for back wall: 20 amp single phase (#12 THHN)
Excluding grounding conductors, which will be addressed below, the total current carrying conductor count is thus:
#2 THHN: 3 each
#6 THHN: 3 each
#10 THHN: 7 each
#12 THHN: 6 each
#14 THHN: 6 each
Ampacities based on NEC table 310.16, THHN insulation in dry locations, maximum temperature rating is 90° C (194° F) based on ambient temperature of 30° C (86° F)
Grounding conductors for each of those circuits, based on NEC Table 250.122 (all conductors are copper):
100 amp circuit: #8
70 amp circuit: #8
30 amp circuit: #10
20 amp circuit: #12
15 amp circuit: #14
The final conductor count is:
#2 THHN: 3 each
#6 THHN: 3 each
#8 THHN: 2 each
#10 THHN: 9 each
#12 THHN: 9 each
#14 THHN: 9 each
The plan is to use two 1 and 1/2 inch EMT conduits between the electrical service panel and the 4 x 4 square wireway. According to NEC Chapter 9, Table 4, the 40% cross sectional size of this conduit is 526 mm2. It is easier to simply use metric measurements for this. The cross sectional wire areas are found in Chapter 9, Table 5. Chart of various conductor sizes and areas:
Conductor
Area (mm2)
Total conductor
Total area (mm2)
#2 THHN
74.71
3
224.13
#6 THHN
32.71
3
98.13
#8 THHN
23.61
2
47.22
#10 THHN
13.61
9
122.49
#12 THHN
8.581
9
77.229
#14 THHN
6.258
9
56.322
Thus, in order to break this up into two 1 and 1/2 inch conduits, the #2, #6 and #8 (main transmitter HV power supply, backup transmitter and grounds) are run in one conduit, the remaining circuits in the other. The idea is that the main transmitter and backup transmitter will not be running simultaneously for long periods of time. Those cable areas total 369.48 mm2, well within the 40% limit of 526 mm2 for 1 and 1/2 inch EMT. The rest of the circuit’s cable areas total 256.041 mm2. That leaves room for additional circuits in the second conduit if future needs dictate. The extra conduit area will make pulling the wires through easy.
From the square wireway to the HV power supply, 1 and 1/4 inch conduit will carry the three #2 and one #8 ground. 1 and 1/4 inch EMT has a cross sectional area of 387 mm2, the conductors contained within will be 271 mm2. Less room here, but still well withing the 40% limit.
Blogging has been light due to work load being heavy, at the moment. We are engaged in moving transmitters out of this old house:
WINE 940 WRKI 95.1 former studio and transmitter site
Into this new transmitter building:
WINE WRKI transmitter building at base of tower
The former building was the original studio for WRKI, 95.1 MHz, which signed on in 1957. The co-located AM station, WINE 940 KHz, did not sign on until 1963. As such, the building is a little worn around the edges, so to speak. The FM transmitter has an auxiliary cooling device, for those hot summer days as the building itself is un airconditioned:
WRKI Harris FM25K transmitter, circa 1986
The rest of the building is in similar condition. Ceiling tiles are falling off the ceiling and getting ground into the floor, junk is pile up in almost every corner, rodent feces, and the basement, don’t even get me started on the basement.
The basic floor plan for the new building is simple:
WRKI WINE transmitter room floor plan
Right now, the preliminaries are being done, mounting the coax switch, running conduit, pulling wires, etc.
A few design notes:
This building is much closer to the tower, which is sited on a high hill (715 feet, 218 Meters) and sticks up 500 feet (152.1 Meters) above that. Basically it is the area lightning rod, thus special attention will need to be paid to grounding and bonding. I decided to isolate the electrical ground in favor of the RF ground for lightning protection. This involves putting toroids on the electrical ground conductors.
The building itself is shielded with continuous steel plating, but that has been cut in a few areas to install air conditioners. Those areas will have to be repaired and the AC units bonded to the steel plate.
Back up cooling will be in the form of a large exhaust fan and intake louver.
The tower itself is AM radiator for WINE. It is 170 degrees tall, which means high RF fields at the base, therefore good RF bypassing is needed.
The transmitter room itself is fairly small for what needs to go in there. careful design and placement is required.
Here are some in progress pictures:
WRKI backup transmitter, Harris FM3.5K, coax switch in the background
The first order of business was retuning a Harris FM3.5K transmitter to function as the backup. The current backup transmitter is an RCA FM20E, which no longer runs. After the move is completed, that transmitter will likely be scrapped.
I attached super strut to the ceiling at four foot intervals. I used this strut to support the 4 port coax switch. All coax in the transmitter room is 3 1/8 inch hardline, which has a power rating of 40 KW. Since the transmitter power output is 20 KW, this leaves a lot of head room for problems. When working with 3 1/8 inch coax, it is important to remember to cut the inner conductor 1 1/2 to 1 3/4 inches sorter than the outer conductor, otherwise the stuff doesn’t go together right.
The 30 KW air cooled dummy load was moved up from the other building and connected to the coax switch. This allowed the backup transmitter to be tested.
WRKI backup transmitter and dummy load
Three inch ground strap connects all the transmitters, racks, and dummy load to the station ground.
WRKI ground strap, new transmitter building
Electrical requirements are being met by a 400 Amp service backed up by a 120 KW generator. Once the conduit work is finished and all the wires pulled, the coax to the old building can be cut and brought into the new building, then the station can go on the air with the “new” backup transmitter.
Since I posted about Andoid phone apps for engineers last February, I noticed several others have picked up the thread and published articles as well. Good for you! I am glad that we can be of assistance here at Engineering Radio.
Radio World had a good article back in July on this subject. I’m not going to link to it, just because. I figured I’d add a few others that didn’t make the grade last go around, either because they didn’t exist, or I didn’t know about them. I am limiting my choices to free apps for Android phones.
Router passwords. Often, very often in fact, the default password on any given router does not get changed when it is installed. I found this app to be accurate and useful for speeding up various router tasks required in day to day radio engineering. Things like opening ports for VNC, routing outside IP addresses to internal ones, etc.
Navaile Electrical Calculator. Great for National Electrical Code questions, wire sizes, breaker sizes, box fill, conduit fill, voltage drop, etc.
RF & Microwave tool box. Has handy calculators for filters, mismatch, return loss, etc, just in case those things need to be done by hand.
GPS test. Shows available GPS signals, gives time, location and accuracy in WGS84 datum.
Shortwave Schedules. Data base of shortwave schedules searchable by station, time, and frequency.
Note pad. Just what it says.
There are several other good ones, but these are the ones that I tend to use most often.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.
~1st amendment to the United States Constitution
Any society that would give up a little liberty to gain a little security will deserve neither and lose both.
~Benjamin Franklin
...radio was discovered, and not invented, and that these frequencies and principles were always in existence long before man was aware of them. Therefore, no one owns them. They are there as free as sunlight, which is a higher frequency form of the same energy.
~Alan Weiner
Everyone has the right to freedom of opinion and expression; this right includes the freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers
~Universal Declaration Of Human Rights, Article 19
Open Mic