November 2016 Archives

AM Tower Heights by Frequency

This is a document apparently originating from ERI, Inc, being circulated around the internet on various blogs and sites. AM tower heights listed by frequency for 1/4-wave and 1/2-wave lengths, in feet and meters.

The All-Pass Audio Filter

The All-Pass Audio Filter is useful for a few reasons:

1) Symmetry of the audio. The human voice is often not symmetrical, and depending on the signaling circuits or transmitter design, it is desirable to limit any sort of inherent DC bias in the transmitted signal. This is an issue for AM and FM transmitters, as well as twisted-pair loops carrying voice communications.

2) Phase correction. The allpass network permits a correction of frequency-dependent phase rotation caused by preceding or following signal processing stages.

In digital signalling networks, these problems are addressed by line coding. The methods of AMI and B8ZS are used in telephony signalling to eliminate a DC bias which would compromise the wire loops and contribute to additional cross-talk coupling, as well as permit the signal to be AC coupled onto a line carrying DC power at -48 VDC or -130 VDC.

Similarly, AX.25 uses Non-return-to-zero inverted (NRZI) to mitigate a similar protocol-based DC bias which would otherwise result from the nature of the data packets. Ethernet relies on 8b/10b encoding in slower forms, and 64b/66b encoding in faster forms to neutralize DC bias as well.

A Compact N-Gauge Model Railroad Using Helical Storage

At the moment, this exists as a series of drawings on paper, so this will be a verbose "patent-style" description.

The design is a multi-level dogbone layout, flanked with helical double-track towers on either side which are integrated into the layout route. This permits the middle to support a bench or "desktop" scene in the vertical middle of the entire assembly, multiple flat single or double track "storage racks" connecting from helix to helix from floor level to the top of the entire layout. At the top, it may be desirable to include a multi-track storage yard without adornment. The scene in the middle has the possibility of being multi-level as well, with as many layers as the builder desires in the primary scene. The lower storage compartment should be protected with polycarbonate sheet to prevent accidental contact with locomotives and rolling stock, as well as to maintain a visual sight line with moving trains (consists).

Primary problems to be solved by the implementer are:

  • Power control and routing
  • "Parking" on the flat surfaces.
  • Most of the trackage in the layout design will be installed in the helix, which at a 2% grade, will require a solid mechanical stop to support the train in the helix.
  • Motive power requirements to lift trains through the layout and helixes. This translates directly into more locomotives, higher loads, and high quality locomotives that use both trucks for collecting DC power and delivering torque.

Advantages of this design are that the helixes can be interfaced on the front side or the back side, and the grade differential per level permits interstitial scenes such as a hump yard. One would need to frame or box the area below the upper scene, but this space can be used for mounting lighting to the scene below.

The helixes can be configured in a 1:1 fashion, or crossover at the bottom. They can be built as a left-hand helix and a right-hand helix, or both sides constructed of the same alignment helix. It may be desirable to avoid same-level rail crossings for unattended operation, or operation without intervention.

Minimum radius should be approximately 15", with the twin tracks centered at 1.5" apart. This permits the helix to occupy a space of approximately 36" x 36". It may be desirable to extend the helixes by using a 6" straight section of track or replacing that straight section with a turnout (switch) or using curved turnouts. Since there are two tracks, the inner helix would likely be uninterrupted and provide a path from the bottom to the top for a "long route" unless a block system or interlocking are provided for and a crossing installed in the outer helix track.

As an ordinary bench may be 36" deep, it would be desirable to make the entire structure as wide as possible, say ten or twelve feet, because three feet per side will be used by the helix for a total of six feet of non-layout space. This limits the center scene to four or six feet, which may be sufficient or excessively limiting to some modelers. 

Battery Power Systems Design

Battery-powered systems are necessarily complex for three reasons: charging, discharging, and heat management. In the world of lithium batteries, this is primarily managed by the BMS, or battery management system.

Panasonic gives excellent battery management advice for dealing with NiCd and NiMH types and the advice is applicable to SLA, VRRH, and other lead-acid battery configurations as well.

Battery heater:

This is a battery heater, which is a necessity in Alaska. The catch with battery management is always keeping them cool when it is warm out, and keeping them warm when it is cool out. The fact that they are primarily lead and have a large amount of thermal mass does not always help. Obviously, passive methods of cooling are good but have to be custom-designed for the application. In the case of NiMH batteries, high temperature at the end of charging is a normal by-product of charging, but should be managed to prevent damage to the cells.

Charging should be structured to state of charge, particularly making sure that each individual battery is charged to it's proper voltage. Strings of batteries mitigate the effects of a shorted cell by minimizing the resulting overcharge voltage. However, a battery going into an open-cell or high-resistance state causes all batteries to be effectively disconnected from the charger or load, rendering a larger capacity of batteries unusable. I recommend that a three-bus approach be used, where battery packs are attached to a load bus using ZVS or other style of buck-boost converter or switch, a charger connected to that load bus, and the battery under charge isolated from the load bus and connected to the charging bus when under charge. This permits a rotating "once a week" charge of each pack, even if the power is out, with a terminal behavior of removing packs from the discharge set as packs are charged for the last time. This way, as few cells as possible are "killed" by being left in an uncharged state unless the load dictates such. Entropy: you can't avoid it.

Discharge should be figured to the chemistry of battery as well. In the case of NiCd and NiMH chemistries, a terminal voltage of 1.0 or 1.1 V per cell may be the lowest that should be permitted. In the case of 12V VRRH or SLA batteries, a terminal voltage of 10.5 or 10.8 volts should be used depending on the current being drawn and the resistance of the wires between the batteries and the load. Once a battery is discharged, it should be recharged as soon as feasible or possible to prevent effective battery death.

Temperature management has two extremes: cold and heat. Even those are relative to the environment of the battery depending on the location of the system in the world. Death Valley regularly has extreme temperatures exceeding 120 degrees F, and the American South may see 90-100 degree F temperatures during a typical summer. Likewise, low temperature extremes are more likely in certain areas than others. These factors shape the design and implementation of a battery thermal management system.

Heat pipes and solid/liquid metal (gallium, indium) cooling systems may be used as pack temperatures approach 90 to 100 degrees F. However, these systems have to be designed not to operate if temperatures are cooler than 70 degrees F as battery capacity starts dropping off below 65 degrees F. Paradoxically, many battery chemistries tend to generate heat as they are discharged which is a benefit to low external temperatures but a caveat when external temperatures are warmer. However, heat-pipe systems may be made of any number of materials, and have correspondingly different behaviors depending on the substances involved. For instance, propane boils at -43.6 degrees F (-42 degrees C), which means it can act as a vapor-phase cooling system if one end of the system is compressed to a liquid state. Likewise, the temperature of the system can be held to a constant as long as a liquid is boiling into a gas state, such as a pot on a stove full of water. This is how a double-boiler manages effective temperature control. In the case of using water, the temperature of the water must not be allowed to drop below 33 degrees F or ice crystals can form at extremes.

Additionally, geothermal contact may be involved in temperature management, using loops of electronegatively inactive, grounded, non-oxidizing metals in well casing, or direct contact with rock outcrops or burial below the soil surface. Depending on the materials involved, these systems can be efficient and/or maintenance free or require periodic maintenance to renew materials or sacrificial anodes. 

In the case of an outside temperature below 55-degrees F, the outside air temperature is lower than the temperature of the ground, which makes contact with ground a good thing. In the case of an outside temperature above 55-degrees, contact with the ground is not desirable because battery capacity decreases rapidly below 65-degrees F.

The other layer of complexity to battery management is that of corrosion. Chemicals in batteries can corrode various metals to different degrees, and it is desirable to prevent some metals from reacting with some electrolytes. Further, some batteries require venting, which may be implemented simply using fans or by using aquarium tubing in others. In general, one should assume that batteries will eventually fail, and in the case of SLA, VRRH, and lead-acid batteries, that either hydrogen and oxygen gas will be vented, or sulfuric acid will be vented or drain into the enclosure. Any nearby electronics will be damaged by the gasses from the batteries.

Yet another area of concern would be the degree to which small rodents of the area find the battery box to be habitable in any and all states of weather.

Another complexity is that batteries should not be placed in parallel unfused lest a battery develop a shorted cell, or several cell failures following a shorted cell. When a 90 Ah battery is rated for 2,000 amps into a dead short, one can be certain that a battery failure will result in the venting of charging gasses and electrolyte fumes while the good battery forces power into the dying battery.

Additionally, serviceable battery types should be serviced on a periodic basis, with distilled water added as appropriate and electrolyte when necessary. Serviceable batteries should be vented and humidity levels managed. Non-serviceable batteries should have discharge and charge rates limited to prevent cell over-temperature or over-pressure.

Marti Audio Performance

Marti makes wideband FM remote broadcast equipment. These are some specifications gleaned from product manuals or sales literature:

sub-audible tone encoder (27Hz, other equipment uses tones at 25, 35, 50, and/or 75 Hz and sometimes in combinations). 
FM compressor/limiter (usually a necessary part of preventing splatter, over-deviation, and excessive occupied bandwidth. Marti may have implemented it as a high-speed limiter or brickwall low-pass filter).
2:1 compander option (These are useful if you have highly dynamic content, but it is required at both ends to be effective. Telephone audio is often compandered; it is the difference between A-law and mu-law signalling. )
Receive Bandwidth: 20-50 KHz, depending on selected frequencies
Deviation: 1.5, 5.0, 7.5 or 10.0 kHz depending upon frequency and filter

Frequency Response:
50 Hz to 3 kHz: ±1.5 dB, 10 kHz Channel BW, 1.5 kHz Dev, 20 kHz filter
50 Hz to 7.5 kHz: ± 1.5 dB, 25 kHz Channel BW, 5.0 kHz Dev, 25 kHz filter
50 Hz to 10.5 kHz: ± 1.5 dB, 36 kHz Channel BW, 7.5 kHz Dev, 36 kHz filter
50 Hz to 10.5 kHz: ± 1.5 dB, 41 kHz Channel BW, 10.0 kHz Dev, 50 kHz filter
S/N Ratio @ 100 µV Input:
10 kHz BW @ 1.5 kHz Dev: 44 dB
25 kHz BW @ 5.0 kHz Dev: 53 dB
36 kHz BW @ 7.5 kHz Dev: 57 dB
50 kHz BW @ 10.0 kHz Dev: 57 dB
THD + Noise:
10 kHz BW @ 1.5 kHz Dev: 2% or less, 50 Hz to 3 kHz
25 kHz BW @ 5.0 kHz Dev: 2% or less, 50 Hz to 7.5 kHz
36 kHz BW @ 7.5 kHz Dev: 2% or less, 50 Hz to 10.5 kHz
50 kHz BW @ 10.0 kHz Dev: 2% or less, 50 Hz to 10.5 kHz
IMD (For 20 dB Signal-to-Noise): 75 dB
Image Rejection: 100 dB

Notably, the Marti doubles the standard deviation of +/- 5 KHz FM radio to 10 KHz. This is because broadcast needs more "headroom"; some NYC STL remotes were used at 200 KHz deviation for 20 KHz audio simply to spread the signal over more spectrum to achieve a greater signal-to-noise ratio. This is the same as process gain in spread-spectrum communications, because "line spectra" -- that is, unmodulated carriers present in the received spectrum -- have less of an individual effect in the recovered audio. A better way to avoid such effects is to switch to a digital format with an acceptable CRC, FEC, or other Error Correcting Code (ECC) to mitigate the errors while using a spectrally dense mode such as QAM256 (8-bits per symbol or baud).  However, the problem with QAM is that one must be able to tell where the bottom of the S/N ratio is, and adjust the range of power dynamically, as well as the upper limits of power. Transmitter power tends to follow a cube-square law in terms of cost per watt, in dB. QAM also requires linear amplifiers, which increases power inefficiency and dissipation.

Durable Transmitter Plant Design

In a modern transmitter plant, whether broadcast AM, FM, TV, paging, or repeater, a circulator is a valuable part of the installation and often omitted to save costs. The circulator, in combination with SWR bridges, enables a site manager to keep the transmitter on the air during difficulty, as well as prevent minor issues from becoming large ones.

A circulator is valuable for two important reasons. First, a circulator assures that the transmitter will always see a 50-ohm match, within a few percent of that impedance. Second, the circulator prevents mixing products from being generated as a result of rectification and amplification in the last stage of the amplifier. Any amplifier designed to operate as a broadband device is subject to this noise generating effect. The circulator acts to insure the investment in transmitter plant, at the cost of the feedline and antenna in the event of fault.

In the case of lightning and other associated arc-over events in the coax, the SWR bridge provides useful feedback to the transmitter plant controller to temporarily cease transmitting, and/or lower and ramp up transmitter output. For instance, combining these functions with flow and pressure monitoring of the dry nitrogen or dry air being fed into the feedline can immediately identify the cause of failure, and shift the source of pressurization into a higher-flow mode or to a more plentiful source such as an air dryer rather than bottled gas. If dry air is used as the source, or excessive gas flow is noted, the transmitter power can be backed down since the air-dielectric insulation will no longer be as strong or effective as it was when it was dry nitrogen under a few pounds of pressure.

A self-diagnosing transmitter needs four SWR bridges. One SWR bridge should be installed between the transmitter and the circulator, another bridge between the circulator and the feedline, yet another bridge located between the feedline and antenna (equipped with high-voltage clamping diodes), and the fourth bridge sampling the RF to the circulator dummy load.

Useful temperature monitoring points are outside air temperature, temperature at the feedline-antenna junction, the middle of the feedline, the bottom of the feedline, the circulator temperature, the circulator dummy load temperature, the transmitter output temperature, the transmitter input temperature, generator input temperature, generator output temperature, generator body temperature, battery temperature, day-tank and storage tank temperature

If the RF is getting to the antenna and you have a bad match, then the antenna is the problem. If the transmitter has a bad match at the ground, then the feedline and/or antenna are compromised. If transmitter isn't making enough RF into a good match, then the transmitter is experiencing internal problems. The circulator lets you deal with bad stuff happening after it, at the cost of continuing whatever arc welding is presently happening in the feedline without extensive monitoring. The feedline is a great heatsink, so if it gets warmer than the outside air, that is an indicator that something is wrong. A fault at the antenna or in the feedline that is noticed can be mitigated by dropping the RF power level, which keeps the station on the air but may not cover as much contour or footprint on the ground. According to the FCC Parts, the engineer would have to file an STA in the case of a broadcast transmitter but it is not turning $100K of feedline into expensive, oxidized scrap.

There is a caveat to this concept: the failure of a single antenna bay in a four-element or larger array may result in a noticeable rise in SWR which, depending on the transmitter power level, may not constitute an alarm. For instance, a 1.5:1 match may be typical of a cheap antenna for a 100 to 200 watt transmitter. In a four-bay "ring dipole", this SWR may be indicative of a single bay failure. In the northern parts of the world, ice accumulation may result in antenna detuning by adding additional capacitance to the antenna. The effect compares to slug tuning an antenna with a block of PTFE. It is therefore important to keep a record of the system SWR response. Regard variations from initial install with careful suspicion, and balance the cost of the tower climber for testing with the cost of replacing the antenna. Additional factors are possible loss in revenue for replacement or inability to transmit, plus filing fees for a "silent" STA.

Other noteable points of protection inside the shelter and around it:
1. A water shield should be placed over the transmitter to prevent water ingress in the event of roof or other failure. This is largely dependent on electrical codes, but can be the difference between rebuilding a transmitter and resetting one that tripped off.

2. The air conditioner drain line should be fitted with a float switch. For example:

In the event of the switch being tripped, an engineer should be dispatched to the site. To resolve possible false trips, three switches can be placed in physical parallel with each other on the same level, and the combined or individual switches fed into the site controller. If two or three switches trip, one can safely assume that there is an irregularity in the air conditioning drain system.

Depending on the climate and likelihood of entry, the air conditioner drain line should be fitting with a screen, or a series of increasingly smaller screens. Insects and rodents may enter the drain line and nest in the drain line, causing a blockage for liquids. Additionally, molds and fungi may form in the drain liquid and introduce turbulence in the drain line as well.

About this Archive

This page is an archive of entries from November 2016 listed from newest to oldest.

October 2016 is the previous archive.

December 2016 is the next archive.

Find recent content on the main index or look in the archives to find all content.