tech stuff – Page 41 – Engineering Radio

Tower lighting transformers, isolation chokes, etc

Series excited AM towers require some way to get standard AC across the base insulator to the tower lights if tower lights are required. While many new AM towers do not have base insulators, through the use of a folded unipole, it is still a very popular design and has several technical advantages.

There are two methods for getting 60 Hz AC from zero RF potential to an excited tower:

Tower lighting choke
Austin Ring transformer

LBA Group TC-300 tower lighting choke courtesy LBA Group, Inc

When to use it depends on the tower and the RF potential on the base of the tower. For towers that are under 140 electrical degrees (RF) and carrier power levels up to 100 KW, a lighting choke works well. They are simple and less expensive than an isolation transformer. They can be installed inside the ATU cabinet or placed in their own weatherproof enclosure as required. Tower lighting chokes will add series impedance to the base of the tower and needs to be compensated for by adding capacitance to the circuit. This will become more pronounced at the lower end of the band, where, if one is not careful, RF from the tower can be coupled to the transmitter building’s AC mains, which is very undesirable.

Tower lighting chokes generally consist of three separate windings, one for the beacon, one for the side lights, and one for neutral. Their inductance is typically in the 800-1000 µH at 1 MHz region. They can be stacked to increase their peak voltage handling capacity:

LBA Group tower strobe light choke courtesy LBA Group, Inc

Peak voltage is determined by the base impedance and carrier power + modulation. On any AM station these days, a 150% peak modulation figure should be used (125% modulation allowed by FCC rules plus a 25% safety factor). For example, station B has a base impedance of 50Ω (typical 90° guyed tower) and a carrier power of 50 KW. The peak modulation power will be 600 KW. Thus, the peak voltage will be E_peak = √P_peak x R, or E_peak = √600,000 watts x 50 ohms or 5,477 volts. With higher base impedances, the base current goes down but the base voltage goes up. A typical 140° tower will have a base impedance of 760Ω. Thus the peak base voltage for a 50KW carrier power modulating at 150% will be 21,354 volts. This is the worst case scenario, as few installations are designed that way and every tower impedance is different than the theoretical self impedances given.

For towers over 140 electrical degrees, it is better to use an isolation transformer because of the RF peak voltage/peak current conditions at the base of towers that are electrically tall. The ring transformer design minimizes stray inductance or capacitance at the base of such towers. Austin Insulators (previously Austin Decca) makes a variety of tower base ring isolation transformers. These have varying input and output voltages.

Diagram of typical Austin Ring transformer courtesy Austin Insulators, Inc

I have seen these at many locations over the years. They are rugged and add only a small bit of capacitive reactance to the base of a typical tower. They also completely isolate the building AC mains from the tower. For very high-power installations, Austin has the A-9600, which was designed for the Navy VLF transmitter towers where base peak RF voltages can run 200,000 volts or more:

Austin A-9600 oil filled isolation transformer — Austin A-9600 oil-filled isolation transformer courtesy Austin Insulator, Inc

Voltage drop is another consideration in tower lighting design. Long runs from the transmitter building to the tower should be on heavy gauge wire and at 230 volts if possible. FAA Circular AC 150/5345-43F “Specification for obstruction lighting equipment” advises that the input voltages for incandescent lighting systems vary by not more than ±3%. Additional tower lighting and painting information can be found in FAA Circular AC 70-7460-1K.

Modulation Dependent Carrier Level

Coming to an AM transmitter near you. The FCC announced that starting immediately, stations can employ Modulation Dependent Carrier Level or MDCL technology on AM transmitters. According to Public Notice DA 11-1535 (.pdf):

Use of MDCL technologies requires a waiver of Section 73.1560(a) of the Commission’s Rules, which sets upper and lower limits for an AM station’s operating power. We hereby establish procedures for AM broadcasters to seek a rule waiver in order to use energy-saving MDCL technologies.

Several transmitter manufacturers offer some version of MDCL in their newer models with the ability to update some older models. Harris Corporation offers something called “Amplitude Modulation Companding” (AMC) and “Adaptive Carrier Control” (ACC). While Nautel includes an option called “Dynamic Carrier Control” (DCC) on all NX series transmitters with the ability to upgrade some older transmitters. Continental offers Controlled Carrier Level Modulation (CCM) on later-model shortwave transmitters by installing SSM modulator, which can be retrofitted.

Nautel’s Dynamic Carrier Control (.ppt) (.pdf available here) reduces the carrier level during moderate modulation periods. The effect of this is to increase the perceived loudness at the receiver. During higher modulation periods, the carrier is increased to prevent distortion. The net effect is between 3 – 6 dB carrier reduction. During quite periods, the carrier is returned to full power to reduce noise.

Nautel AM Dynamic Carrier Control wave forms — Nautel Presentation on Dynamic AM carrier control

The potential savings are from 20-40%, which for a 10 or 50 KW station, would represent a significant reduction in the power bill. For a 50 kW station running an older transmitter, the savings would fall in the $37,000 to $56,000 per year range. In some cases, smaller stations may be able to get rid of a demand meter, which would also represent significant savings. The threshold for demand meters around here is 5,000 KWh per month.

The FCC further notes that:

The reduction in AM signal power at certain modulation levels inevitably exacts some penalty upon audio quality. Depending on the content of the audio program, MDCL algorithms may introduce some audio distortion or may decrease the signal-to-noise ratio in the receiver. In addition, MDCL algorithms may erode coverage slightly at the fringes of the AM station’s protected service area. Both the long experience of transmitter manufacturers and broadcasters abroad, and the initial reports from experimental operations in Alaska however, indicate that such adverse effects are generally imperceptible.

This would be especially true for higher-powered stations that stand to save the most money.

The malfunctioning STL antenna

Right after Tropical Storm Irene, it was noted that the STL signal strength at the WHUD transmitter site was low. Normally it was 300+ µV, but now reading around 100 µV, which is a problem. Upon further investigation, it was revealed that the STL transmitter on the intermediate hop had higher than normal reflected power.

Time to call the tower crew.

The STL transmit antenna for WHUD’s STL (WPOU464) hop is a Scala Paraflector (PR-950), mounted at the 280-foot level on this tower:

The fact that it happened after a major storm and the transmitter was showing higher than normal reflected power indicates a problem with either the antenna or the jumper between the 7/8″ Cablewave coax and the N connector on the antenna. A measurement with a spectrum analyzer shows very high return loss:

This shows the distance to fault 413 feet, with a return loss of -7.4 dB. That distance is either near or at the antenna and -7.4 dB indicates a lot of reflected power. We had the tower climber take apart the jumper connections and terminate the jumper with a known good 50-ohm load. The return loss did not change. We then had him swap out jumpers and reconnect to the antenna. That did the trick:

Much better, most of the power is now being radiated by the antenna, the VSWR is 1.02:1. The impedance bump at 51 feet is a sharp bend in the coax where it is attached to an ice bridge. Reconnecting the transmission line to the transmitter and turning it on confirms that all is normal again. The problem with the jumper was found in one of the connectors, it was full of water.

Water contaminated Andrew flexwell connector — Water-contaminated Andrew flexwell connector

I cut away the boot, water had entered the connector from the back because waterproofing and tape was not applied all the way to the coax. This was installed in 1998 when the station moved from Peekskill to its current location in the town of Fishkill. The fact that it happened now in the nice weather when Mt. Beacon is still accessible and not in the middle of winter means the radio gods are smiling on us.

Things that make you go Hmmmm? Part IV

The case of the blown fuse, or rather the blown up fuse:

Blown 10 amp fuse on Harris SX5 PA board

F32 is blown into small bits and had to be vacuumed out of the bottom of the transmitter. The reason why is the pair of MOSFETs connected to that circuit were shorted. Of course, the reason for the shorting of MOSFETs needed to be investigated. What I found was on the underside of the PA board where the brass stand-off attacked the toroid combiner board, the nuts attaching the stand-off to the combiner board were loose and there was a big arc mark.

I tightened everything up and replaced the MOSFETS, marking them with a pen in case they short again, in which case the drive section needs to be closely examined.