# Antennas and filters At the end you have an antenna outdoors for each band you want, and a chain whose loss you know. You also have a count that tells you if a change helped. Spurious Emissions · Rev A · checked 2026-10-10 · https://spurious-emissions.net/guides/antennas-and-filters/ Some links are affiliate links: if you buy through them, I may earn a commission at no extra cost to you. As an Amazon Associate I earn from qualifying purchases. ## 1. What this guide builds This guide builds the part of a receiving station that is not software. That part is the antenna, the cable, and the small parts in front of the radio. It gives you a plan for each band, a loss figure for each chain, and a way to measure a change. The station only receives. Nothing here transmits. | Signal | Frequency | Antenna | Radio | Decoder | Computer | |---|---|---|---|---|---| | Aircraft positions (ADS-B) | 1090 MHz | Narrow vertical for 1090 MHz | FlightAware Pro Stick Plus | readsb | `vhf` | | Ship positions (AIS) | 161.975 and 162.025 MHz | Narrow vertical for the marine band | Airspy Mini | AIS-catcher | `vhf` | | Aircraft text messages (ACARS, VDL Mode 2) | 129 to 137 MHz | Discone | Airspy R2 and RTL-SDR Blog V4L | acarsdec, dumpvdl2 | `vhf` | | Shortwave | Below 30 MHz | Small active loop | RX-888 Mk II | dumphfdl and others | `hf` | The radios, decoders and computers are those of my full-station guide, [Build the station](/guides/build-the-station/). This guide does not install software. You can use it with any radio and any decoder. ``` outdoors cable entry (indoors) home network -------- --------------------- ------------ 1090 MHz vertical -- cable ------- radio (amplifier and filter inside) --+ marine vertical --- cable -- [FM filter] -- radio ------------------------+-- computer discone ----------- cable -- FM filter -- splitter --+-- radio A ---------+ `-- radio B ---------+ loop (amplifier in its head) -- cable -- power injector -- 30 MHz low-pass filter -- limiter -- shortwave radio -- computer ``` One rule shapes the drawing. Each radio sits where its antenna cable enters the house. Antenna cable loses signal with every meter. Network cable carries the decoded data with no loss. ## 2. What you need Prices are US dollars unless the cell says otherwise. Each price is the named seller's price on 2026-10-07, from my full-station guide. Prices change. A part marked ◆ is one I use. A blank box is for your pencil. ### Antennas | ☐ | Part | Band | Connector | Price | Note | |---|---|---|---|---|---| | ☐ | ◆ [FlightAware 1090 MHz antenna](https://flightaware.store/products/antenna-1090mhz), 66 cm, 5.5 dBi | 1090 MHz | N type | $49.99, FlightAware store | A mast bracket is included. The page showed "sold out" on 2026-10-10 | | ☐ | ◆ [AIS-band vertical](https://www.amazon.com/dp/B09374SM2V?tag=spuriousemiss-20), tuned near 162 MHz | Marine band | See listing | See listing | A low-cost one is enough to start | | ☐ | ◆ [Diamond D-130J discone](https://www.amazon.com/dp/B00AR09A6I?tag=spuriousemiss-20), 25 to 1300 MHz receive, 1.7 m | Wide VHF and UHF | SO-239 (the D-130NJ has N type) | See listing, Amazon | One wide antenna for two radios | | ☐ | ◆ [MLA-30+ active loop](https://www.amazon.com/dp/B095K89WND?tag=spuriousemiss-20) with its power injector | Shortwave | SMA | See listing, Amazon | The amplifier is in the loop head | ### Filters, splitter and protection | ☐ | Part | Job | Connector | Price | Note | |---|---|---|---|---|---| | ☐ | ◆ [RTL-SDR Blog FM band-stop filter](https://www.amazon.com/dp/B01LE9LRPM?tag=spuriousemiss-20) | First part after the discone cable | SMA socket at each end | See listing, Amazon | Removes 88 to 108 MHz. Section 10 has its figures | | ☐ | ◆ [Nooelec Flamingo FM](https://www.amazon.com/dp/B079CNCR8V?tag=spuriousemiss-20) | On the marine vertical | SMA | See listing, Amazon | Optional. My count did not change with it | | ☐ | ◆ [Mini-Circuits ZX10-2-12-S+](https://www.minicircuits.com/WebStore/dashboard.html?model=ZX10-2-12-S%2B) 2-way splitter, 2 to 1200 MHz | Feeds two radios from one antenna | SMA | $42.20, Mini-Circuits | Mine came from a surplus dealer | | ☐ | ◆ [Turn Island Systems F3060](https://turnislandsystems.com/product/f3060-filter-shelf-30-60-mhz-lpf/) low-pass filter | In front of a direct-sampling shortwave radio | SMA | $80, Turn Island Systems | Section 13 says when you need it | | ☐ | ◆ [TAPR RF Limiter](https://tapr.org/product/rf-limiter/) | Last part before the shortwave radio | SMA socket in, SMA plug out | $20, TAPR | Protects the radio's input | Do not buy an amplifier at the start. Section 12 says why. ### Cable and small parts | ☐ | Part | Job | |---|---|---| | ☐ | Low-loss 50 ohm cable for 1090 MHz, N plug at the antenna end, SMA plug at the radio end | Aircraft antenna. Section 6 has the loss for each cable type | | ☐ | 50 ohm cable with a PL-259 plug at the antenna end and an SMA plug at the radio end | Discone | | ☐ | Cable to suit the marine vertical, SMA plug at the radio end | Ship antenna | | ☐ | RG-6 quad-shield cable, and two F to SMA adapters | Loop to the power injector | | ☐ | Short SMA plug-to-plug jumpers, 5 or more | Between filter, splitter and radios | | ☐ | Snap-on ferrites | Both ends of the loop cable | | ☐ | Grounding block for each cable where it enters the house, and bonding wire | Section 8 | | ☐ | 50 ohm SMA terminator | The noise test in section 15 | | ☐ | Self-amalgamating tape | Seals each outdoor connector against water | | ☐ | A mast or bracket, and its fixings | Section 9 | ### The lowest-cost path | ☐ | Part | Job | What it gives up | |---|---|---|---| | ☐ | One narrow vertical for the one band you want most | The antenna | The other bands | | ☐ | The shortest cable that reaches a place outdoors | Cable | Nothing. A short cable is the better cable | | ☐ | One RTL-SDR radio (the RTL-SDR Blog V4L is $37.95 at rtl-sdr.com) | Radio | Dynamic range. A strong signal overloads it sooner than a 12-bit radio | | ☐ | No splitter, no amplifier, no filter | | Nothing, until a count shows that you need one | For 1090 MHz, the lowest-cost radio that I can recommend is the FlightAware Pro Stick Plus ($45.99, FlightAware store). Its amplifier and its 1090 MHz filter are inside. A licensed amateur can also build a quarter-wave vertical with radials for one band. At 131 MHz a quarter wave is about 57 cm in free space. I did not build one for this station. The lowest-cost path gives up the wide discone and all of shortwave at once. It does not give up the most important step: the antenna goes outdoors. ## 3. Before you start **Skills.** You can fit a coax connector or you buy cables with the connectors fitted. You can add decibels. You can work safely at the height where the antenna goes, or you pay a person who can. **Time.** A plan takes one evening. One antenna on an existing mast takes a few hours. The counting in section 14 takes two days for each change, and most of that time is waiting. **Accounts.** None. **Values to choose first.** Write one line for each antenna. | Placeholder | Meaning | Example form | |---|---|---| | `` | The frequency you want from this antenna | `1090` | | `` | Cable length from the antenna to the radio, in meters | `12` | | `` | The cable type | `LMR-240` | | `` | The sum of every loss between antenna and radio (section 7) | `3.3` | | `` | Messages in one hour before any change (section 14) | `95` | **The law.** I am not a lawyer. This paragraph is for the United States. Reception is legal there. US law (47 U.S.C. section 605) says that you must not tell others the contents of a message that was not meant for the public. It makes exceptions for broadcasts, amateur radio and distress traffic. Your building, your landlord or your town may have rules for an outdoor antenna. Read them before you drill. > **Caution.** This chain is for reception only. The limiter, the splitter and the radios take milliwatts. Never connect a transmitter to any part of it. ## 4. Choose an antenna for each band Choose the antenna by the band, not by the price. A narrow antenna hears its own band well and rejects much of the rest. A wide antenna hears everything, and that includes the signals you do not want. | Band | Antenna | Why | What it costs you | |---|---|---|---| | 1090 MHz, aircraft | A narrow vertical made for 1090 MHz | Aircraft transmit with vertical polarization, from every direction | It hears only this band | | Marine band near 162 MHz, ships | A narrow vertical made for the marine band | The same two reasons | It hears only this band | | Many bands from 25 to 1300 MHz | A discone: a flat disc over a cone of metal rods | One antenna covers a very wide range | It has no gain, and it also collects the FM broadcast band | | Shortwave, below 30 MHz | A small active loop: about 1 m across, with an amplifier in its head | It is small, and you can put it far from the house | It needs power through its cable, and its amplifier is open to static | 1. List the signals you want and the frequency of each. 2. Give each signal a narrow vertical if it has a band of its own. 3. Use one discone for the signals that are left, if they are above 25 MHz. 4. Use one small active loop for everything below 30 MHz. Result: one line for each antenna, with its band. My station follows this list. Two narrow verticals serve aircraft and ships. The discone serves the aircraft text messages near 131 and 136 MHz. The loop serves shortwave. The discone's range of 25 to 1300 MHz is its maker's figure. I did not measure it. ## 5. Choose the place for each antenna This step matters more than any part on the list. ### Line-of-sight bands: height Signals at 130 MHz, 162 MHz and 1090 MHz travel in a nearly straight line. A wall, a roof, a hill or a line of trees between you and the transmitter removes signal. Height removes obstacles. 1. Put each vertical and the discone outdoors. 2. Put each one above the roof line if you can do that safely. 3. Keep the top of each antenna clear of the mast and of the other antennas. 4. Give the ship antenna the best view of the water. A ship's antenna is low, so your height counts most there. Result: each antenna has open sky in the directions the traffic comes from. Here is my own number. My ship antenna was a small whip indoors, next to the radio. It heard about 5 messages a minute. Then I put a low-cost vertical for the ship band on the roof, with the same radio and the same program. It heard about 30 messages a minute. That is 6 times more. Both rates are approximate readings from my log. The computer moved on the same day, so the cable changed too. My aircraft antenna made the same move, from a window to the roof. I did not measure that one. ### A small loop: distance from the house For a small loop, height matters little. Distance from the electronics in the house matters a lot. Power supplies, network equipment, lamps and chargers make noise on shortwave. The loop hears that noise more strongly than it hears a far station. 1. Put the loop outdoors. 2. Put it several meters or more from the house and from other electrical equipment. 3. Mount it with its middle about 2 m above the ground. More height brings little. 4. Fit a snap-on ferrite at each end of the loop cable. 5. Turn the loop slowly and listen. A small loop has two sharp nulls, and you can point one at a noise source. Result: the noise level on a quiet frequency is lower than it was indoors. The figures in steps 2 and 3 come from the designer of a well-known active loop (LZ1AQ, section 19). He writes that the antenna works even close to the ground. They are not my measurements. At the time of the episode my own loop was still indoors, next to the computer. That is the worst place for it. ### Where the radios go 1. Find the place where the cables can enter the house by the shortest path. 2. Put the radios and the computer there. 3. Run network cable from there to the rest of the house. Result: every antenna cable is as short as the building allows. ## 6. Choose the cable and the connectors ### Loss in the cable Cable loss grows with length and with frequency. The table gives the loss of 10 m of cable, in decibels. I computed it from the formula in the Times Microwave "Coax Selection Guide" for the maker's LMR cables.[^coax] | Cable | Outer diameter | 14 MHz | 131 MHz | 162 MHz | 1090 MHz | |---|---|---|---|---|---| | LMR-400 | 10.3 mm | 0.15 | 0.47 | 0.52 | 1.42 | | LMR-240 | 6.1 mm | 0.30 | 0.92 | 1.03 | 2.74 | | LMR-195 | 5.0 mm | 0.44 | 1.36 | 1.52 | 4.03 | | LMR-100A | 2.8 mm | 0.88 | 2.74 | 3.05 | 8.30 | The same guide lists generic RG-58 at about 2.0 dB for 10 m at 150 MHz, and 5.4 dB at 900 MHz. A cable from another maker with a similar name has its own figures. Read that maker's data sheet. How to read the table: - **At 1090 MHz the cable decides the result.** FlightAware's advice is a loss under 6 dB and a length under 15 m. With 10 m of LMR-100A you are already past 8 dB. - **At 131 and 162 MHz a medium cable is enough** for a run of 10 to 15 m. - **On shortwave the loss is small in any of these cables.** For the loop I use RG-6, a 75 ohm television cable. It is low in cost and well shielded. On shortwave, 10 to 20 m of it loses under 1 dB. On the bands above 100 MHz, each decibel that you lose before the radio is a decibel of the weakest signals. On shortwave the noise from outside is far above the radio's own noise. There a few decibels of loss change nothing that you can hear. Section 15 has the test for this. ### Connectors and adapters | Connector | Where you meet it | |---|---| | N type | The 1090 MHz antenna, and some discones | | SO-239 socket, PL-259 plug | The D-130J discone | | SMA | Radios, filters, the splitter, the limiter | | F type | RG-6 cable | 1. Buy or make each cable with the right plug at each end. 2. Use an adapter only where you cannot avoid one. 3. Use a short flexible jumper between a thick cable and a small radio. 4. Tighten SMA connectors with your fingers, then a little more. Do not turn the body of the cable. 5. Wrap each outdoor connector with self-amalgamating tape. Result: no adapter chains, no thick cable that hangs from a small socket, and no outdoor connector open to rain. I have no maker's figure for the loss in one adapter, so I give none. A good adapter loses little at these frequencies. The larger risks are a poor contact, water in an outdoor connector, and the weight of a stiff cable on a small socket. ## 7. Plan the chain and add up the loss ### The order of the parts Build each chain in this order, antenna first. ``` A. Aircraft 1090 MHz vertical -- cable -- radio with amplifier and filter inside (add nothing) B. Ships marine vertical -- cable -- [FM band-stop filter, optional] -- radio C. Messages discone -- cable -- FM band-stop filter -- 2-way splitter |-- leg A -- radio A `-- leg B -- radio B D. Shortwave loop -- RG-6 -- power injector -- 30 MHz low-pass filter -- limiter -- direct-sampling radio ``` The reasons for the order: - **Filter before splitter.** One filter then cleans both legs. - **Filter before amplifier.** An amplifier in front of the filter also amplifies the signal you want to remove. - **No amplifier on a short, direct run.** Add one only to pay for a loss that you know. Count messages before and after. - **Injector first on the loop cable.** Every part after it then sees no DC. The limiter blocks DC, so it must come after the injector. - **Low-pass filter before the limiter.** A limiter makes harmonics when it clamps. TAPR's manual says to put a filter ahead of it where possible. - **Bias tee off on every radio behind the splitter and on the shortwave radio.** A bias tee is a circuit in a radio that puts DC on the antenna cable, to power an amplifier. A radio that does this can feed DC into the other leg of the splitter. ### The loss budget 1. Write the cable loss: `` divided by 10, multiplied by the figure in section 6. 2. Add the loss of each filter from its maker's figures. 3. Add the loss of the splitter, if there is one. 4. Write the sum as ``. Result: one number for each chain. Two worked examples, with a generic house and cable lengths I chose for the arithmetic: | Chain | Cable | Filter | Splitter | `` | |---|---|---|---|---| | 1090 MHz vertical, 12 m of LMR-240 | 1.2 × 2.74 = 3.3 dB | none | none | 3.3 dB | | Discone at 131 MHz, 15 m of LMR-400 | 1.5 × 0.47 = 0.7 dB | about 0 dB | 3.3 dB | 4.0 dB | The first chain is inside FlightAware's limit of 6 dB. In the second chain the splitter costs far more than the cable. That is the normal case for a discone with two radios. ## 8. Lightning, static and grounding > **Warning.** An outdoor antenna and its cable can bring lightning energy and static charge into the house. That can start a fire, destroy equipment, and injure or kill a person. Read this section before you put any antenna outdoors. Have a qualified person do the bonding, or have one inspect it. Three facts first. - Nothing protects equipment from a direct lightning strike. Correct bonding and surge protection can reduce the damage from a strike nearby. - Wind alone puts a static charge on an antenna and its cable. That charge can jump through the first amplifier or radio it finds. - A mast that is not bonded correctly is a hazard to people, not only to radios. A published summary of the US code warns of flashover and possible electrocution.[^nec] What I can tell you, and what I cannot: 1. Find the rule that applies where you live. In the United States it is the National Electrical Code, Article 810, as your state adopted it. In another country, use the national electrical code of that country. 2. Read it, or read a summary by a qualified author. 3. Plan a path to ground for the mast and for every antenna cable, at the place where the cable enters the building. 4. Bond that path to the building's own grounding system. Never use a ground rod that stands alone: a surge then goes through your equipment. 5. Keep the bonding wire short and as straight as you can. 6. Ask a licensed electrician to do this work or to inspect it, if you are not qualified. Result: every cable passes a grounding block before it reaches a radio, and the block is bonded to the building's grounding system. I read a published summary of Article 810.21 of the 2020 edition, not the code text.[^nec] It gives a minimum size for the bonding wire. It says that you must bond a ground rod you add to the building's grounding system. It gives a minimum size for that wire too. I do not repeat the sizes here. Take them from the code that applies to you, in the edition that applies to you. Lightning protection for a whole station is a larger subject than this section. The ARRL keeps a list of articles on it (section 19). The parts in this guide help with static only at the edges. The limiter in section 13 clamps a signal that is too strong for the shortwave radio. It is not a lightning protector. The amplifier in the head of an active loop is the part most open to static. I keep a second loop as a spare. ## 9. Install the antennas and the cable > **Warning.** Contact between a mast, an antenna or a cable and an overhead power line can kill you. Before you lift anything, look up. Keep the mast and the antenna far from any power line. The distance must be much more than their full length. If you cannot do that, choose another place. > **Warning.** A fall from a roof or a ladder can kill you. Do not work alone. Do not work in wind, rain or ice. Pay a person with the right equipment if you do not have it. 1. Assemble the antenna on the ground. Follow the maker's sheet. 2. Connect the cable to the antenna on the ground. 3. Wrap the connector with self-amalgamating tape. 4. Fix the mast or bracket to the building. Use fixings that suit the wall or roof. 5. Mount the antenna vertically. A vertical that leans loses signal toward the horizon. 6. Fix the cable to the mast with a little slack at the antenna. 7. Make a drip loop: let the cable hang lower than the entry hole just before the hole. Water then runs off. 8. Connect the cable to its grounding block (section 8). 9. Bring the cable to the radio by the shortest path. Do not bend it sharply. 10. Connect the chain in the order of section 7. Result: the decoder for that band shows its first signals. Near salt water, the first ship messages arrive within minutes. On my first day with an indoor whip, they came within two minutes. For the loop, do the steps in section 5 and section 13. Mount it where it is quiet, not where it is high. ## 10. The FM broadcast filter, and how to tell if you need one FM broadcast stations use 88 to 108 MHz. They are far stronger than a message from an aircraft. A wide antenna collects them with everything else. A strong signal that you do not want can overload a radio, and then the weak signals are lost. A band-stop filter removes one band and lets the rest pass. | Filter | Removes | By how much | Loss outside that band | Passes DC | |---|---|---|---|---| | ◆ RTL-SDR Blog FM band-stop | 88 to 108 MHz | About 50 dB or more | Practically zero below 500 MHz. Under 0.5 dB to 1 GHz | Yes, 80 mA tested | | ◆ Nooelec Flamingo FM | The FM broadcast band | Over 40 dB typical | "Minimal" (no figure given) | Yes, 250 mA minimum | These are each maker's figures. One detail matters for aircraft voice. The RTL-SDR Blog filter is 3 dB down at 76 MHz and at 122 MHz. The Flamingo is 3 dB down at 80 MHz and at 115 MHz. So the lower part of the aircraft voice band, 118 to 122 MHz, loses some signal with the first one. ### How to tell if you need one Use the tests in this order. Stop when one gives a clear answer. 1. **Look at the antenna.** A wide antenna, such as a discone, needs the filter. Fit it and continue with section 11. 2. **Look at the radio.** A radio with a filter for its own band inside does not need a second filter. The Pro Stick Plus is one. 3. **Look at the spectrum.** Open a receiver program that shows a spectrum. Tune to your band. Raise the gain. The maker of the filter says you need one if you see copies of FM broadcast stations there. A copy is about 200 kHz wide and carries music or speech. 4. **Count.** Use the method in section 14: one hour without the filter and the same hour with it. Result: you know if the filter stays, from a number and not from a feeling. My results, so you know what to expect. On the discone the filter is the first part after the cable. I saw no loss from it, so it stays. I did not measure what it gains me. On the narrow ship antenna I added a Flamingo FM. The message rate did not change. I cannot say that it helps there. I left it in. ## 11. The splitter: two radios on one antenna A 2-way splitter has one connector in and two out. It is not free. Each radio gets a little less than half of the signal. | Figure | Value | Source | |---|---|---| | Loss of a perfect 2-way split | 3.0 dB | Half the power to each leg | | Total loss of the ZX10-2-12-S+ at 140 MHz | 3.15 and 3.28 dB on its two legs, typical | The maker's data sheet | | Isolation between the two legs at 140 MHz | 30 dB typical | The same | | Maximum power in | 0.5 W | The same | So each radio gets about 47 % of the signal. The rest becomes heat in the splitter. The episode says 45 %: there I used an expected loss of 3.5 dB. 1. Connect the FM band-stop filter to the cable from the antenna. 2. Connect the filter to the common connector of the splitter (marked S). 3. Connect one radio to each of the two other connectors. Use short jumpers. 4. Turn the bias tee off on both radios. 5. Do not leave a leg open. If you remove one radio, fit the 50 ohm terminator in its place. Result: both decoders show messages. Use a real power splitter with the frequency range you need. A television splitter is made for 75 ohm cable. A plain T adapter is not a splitter: it matches neither radio. I have not measured what the split cost me in messages. If it cost messages, an amplifier in front of the splitter may earn its place there. That is the next section. ## 12. The amplifier: when it helps and when it hurts An amplifier for a receiving chain is called an LNA, a low-noise amplifier. A 20 dB amplifier makes every signal a hundred times stronger in power. It does the same to the noise that arrives with the signal. It also adds a little noise of its own. A radio has a floor of noise and a ceiling. A signal under the floor is lost. A signal over the ceiling distorts, and the distortion covers weak signals. An amplifier lifts everything toward the ceiling. ### When it hurts It hurts when the radio already hears the antenna's noise. That is the case on a short, direct run to a good radio. I measured this once. I put a generic 20 dB amplifier between the FM filter and the Airspy R2, on a direct run from the roof discone. | | Messages in one hour | |---|---| | Before the amplifier | 95 | | With the amplifier | 42 | The strong messages got about 12 dB louder. The faint ones were gone. I took the amplifier out after ninety minutes, and the rate came back within a quarter of an hour. That radio did not need an amplifier. This is one test on one day, made before I fitted the splitter. The floor-and-ceiling explanation is my best reading of the result. I did not measure the cause. ### When it may help It may help when there is a loss after it that it can pay for. | Situation | Amplifier? | |---|---| | A short direct run to a 12-bit radio | No. This is my measured case | | A radio with an amplifier inside (the Pro Stick Plus has 19 dB) | No. Add nothing | | An active loop | No. The amplifier is in the loop head | | A splitter, or a long cable, with a loss of several dB | Maybe. Measure first | ### Where it belongs 1. Put the amplifier before the loss that it pays for. That means before the splitter, or at the antenna end of a long cable. 2. Put a filter before the amplifier if a strong signal is near. On a discone that is the FM band-stop filter. 3. Lower the gain setting of the radio after you add the amplifier. Start 10 dB lower. 4. Count before and after (section 14). 5. Remove the amplifier if the count does not rise. Result: a number that says the amplifier stays or goes. > **Caution.** An amplifier at the antenna end takes its power through the cable. That DC then passes through every part between the power source and the amplifier. The limiter blocks DC. The data sheet of my splitter does not say if it passes DC. Check each part before you turn a bias tee on. I have not run an amplifier at the antenna end or in front of the splitter. Both are plans, not results. ## 13. The shortwave chain: injector, low-pass filter, limiter ``` loop -- RG-6 -- power injector -- low-pass filter -- limiter -- radio (HF input) ``` ### The power injector The amplifier in the loop head takes its power through the signal cable. The injector puts DC on the cable toward the loop and passes the signal toward the radio. 1. Connect the loop cable to the antenna side of the injector. 2. Connect the radio side of the injector to the next part. 3. Power the injector as its sheet says. Result: the noise level at the radio rises when you power the injector. With no power, the loop is nearly deaf. ### The low-pass filter A direct-sampling radio measures the antenna voltage many million times each second, with no tuner in front. The RX-888 Mk II is one. At 64.8 million samples each second it cannot tell a signal above 32.4 MHz from one below it. A signal from 32 to 65 MHz folds back onto shortwave. A low-pass filter removes it first. Figures for the F3060 with its 30 MHz filter in and its shelf filter bypassed, from its manual: | Frequency | Loss | |---|---| | 1 MHz | 0.5 dB | | 14 MHz | 1.2 dB | | 28 MHz | 3.7 dB | | 32.5 MHz | 36.3 dB | | 51 MHz | 63.7 dB | | Above 100 MHz | Over 60 dB | 1. Open the F3060 and find the two pairs of jumpers. 2. Set the pair marked for the 30 MHz filter to the 30 MHz position. 3. Set the pair marked for the shelf filter to BYPASS. 4. Connect the injector to IN and the limiter to OUT. Result: signals above 30 MHz do not reach the radio. The shelf filter cuts 28.5 dB at 1 MHz and 20 dB at 3.5 MHz. Leave it bypassed. Turn it on only if strong AM broadcast stations overload the radio at night. A radio with a tuner and its own filters may not need this part. Read its manual. With a direct-sampling radio at this rate the filter is not optional. ### The limiter The limiter protects the radio's input from a signal that is too strong. Mine is there for static and for my own transmitter. I transmit on shortwave from the same property, at 5 to 20 watts. Figures from TAPR's manual and product page: | Figure | Value | |---|---| | Starts to limit at | About -6 dBm in | | Output with 0 dBm in | Under -1 dBm | | Loss for a weak signal | Under 1 dB up to 54 MHz | | Made for | 54 MHz and below | | Maximum continuous power | 200 mW | | DC | Blocked, at the input and at the output | 1. Connect the output of the low-pass filter to the INPUT of the limiter. 2. Connect the OUTPUT of the limiter to the HF input of the radio. Result: no change that you can hear. A limiter that works is invisible on weak signals. > **Caution.** The limiter is not a transmit switch. Above about 1 W in, a small resistor inside is made to burn out and cut the signal. TAPR does not guarantee that this protects the radio. Keep the loop away from your transmitting antenna. If you never transmit, the limiter is still low-cost insurance for an expensive radio. ## 14. The method: change one thing, count before and after This is the method the episode teaches. It needs no test equipment. The decoder is the instrument, and the count of good messages is the reading. ### Where to read the count | Feed | Where | What to read | |---|---|---| | Aircraft positions | `http://vhf:8080/graphs1090/` | The message rate graph | | Aircraft positions | The command below | The total number of messages since the decoder started | | Ships | The Statistics tab of the AIS-catcher page at `http://vhf:90/` | The message count | | Ships | `docker logs shipfeeder 2>&1 \| tail -n 5` | The summary that the container writes every 60 seconds | | Aircraft messages | The statistics charts of ACARS Hub at `http://vhf:8081/` | Messages for each decoder, for the last hour and longer | The addresses are those of my full-station guide. For aircraft positions, read the total at the start and at the end of the hour, and subtract. ```bash curl -s http://localhost:8080/data/aircraft.json | python3 -c "import json,sys; print(json.load(sys.stdin)['messages'])" ``` Result: one whole number. The decoder's documents call it the total number of Mode S messages since it started. ### Way 1: the same hours on two days Traffic changes with the hour and with the day of the week. So compare the same hours. 1. Choose a window of one or two hours when traffic is steady. A weekday mid-morning is a good start. 2. On day 1, change nothing. Write the count for the window. This is ``. 3. Make one change. Change nothing else: not the gain, not the frequency list, not the cable. 4. On day 2, write the count for the same window. 5. On day 3, undo the change. Write the count for the same window. 6. Compare. The change is real if day 2 differs from both day 1 and day 3 in the same direction. Result: three numbers. The third one protects you from a quiet day. Use the same weekday for all three if the traffic near you has a weekly pattern. ### Way 2: two radios at the same time This way removes the weather, the hour and the traffic from the comparison. It needs a splitter, two radios of the same model and two copies of the decoder. 1. Connect both radios to the two legs of one splitter. 2. Give both decoders the same frequencies and the same gain. 3. Count for one hour. The two counts must be close. Write the ratio of B to A. 4. Put the part under test in front of radio B only. 5. Count for one more hour. Write the new ratio of B to A. 6. Compare the two ratios. Result: if the ratio moved, the part did it. Both radios heard the same sky. I have not run this way myself. It is the standard way to compare two receiving chains. ### How big a change is real A count has a random spread. For a count of N messages, a rule of thumb from counting statistics is a spread of about the square root of N. Messages also come in bursts, so the true spread is larger. | Count | Square root | Treat as "no change" | |---|---|---| | 25 | 5 | 15 to 35 | | 100 | 10 | 80 to 120 | | 400 | 20 | 360 to 440 | The third column is two square roots each way. It is my own rule, not a standard. A longer window gives a larger count and a smaller spread in proportion. My amplifier test as a worked example: 95 before, 42 with it. The square root of 95 is about 10. The fall of 53 is more than five times that. One hour was enough to see it. Then I removed the amplifier and the rate came back, which is step 5 of way 1. The hour after the removal was an evening peak with 144 messages. That number says "it works again". It does not say "better than before", because the hour was a different one. ### Write it down | Date | Window | Chain as built | Count | Note | |---|---|---|---|---| | | | | | | Keep this log with the station. A count with no note of the chain is worth nothing a month later. ## 15. Check that it works 1. **Every feed counts.** Each page in section 14 shows new messages. Wait 10 minutes for ships near salt water and for VDL Mode 2 near a large airport. ACARS comes in bursts: wait an hour. 2. **The loss budget is written.** Each chain has its ``. The 1090 MHz chain is under 6 dB. 3. **The antenna sets the noise on shortwave.** Note the noise level on a quiet frequency. Replace the antenna cable with the 50 ohm terminator at the radio's end of the chain. The noise must fall by 6 dB or more. On my station it fell by 30 to 40 dB. If it does not fall on some band, an amplifier may help on that band, and a filter will not. 4. **The level into the shortwave radio is right.** The ka9q-radio project's rule for the RX-888 is an average of -20 to -25 dBFS at the converter. dBFS is the level compared with the largest value the converter can measure. A count of overranges that keeps rising means too much gain. 5. **No DC where it does not belong.** The bias tee is off on every radio behind the splitter and on the shortwave radio. The loop takes its power only from its injector. 6. **Every outdoor connector is wrapped,** and every cable passes its grounding block. 7. **The baseline exists.** You have the count of each feed for one normal day. Compare every later change with it. ## 16. If it does not work | Symptom | Likely cause | What to do | |---|---|---| | Few messages, antenna indoors | Walls, and the noise of the house | Put the antenna outdoors (section 5). Mine heard 6 times more | | Fewer messages after you add an amplifier | The strong signals now overload the radio, and the weak ones are lost | Remove it. On my station a 20 dB amplifier cut ACARS from 95 to 42 messages an hour | | Copies of FM broadcast stations in a band where none belong | The radio is overloaded by the FM band | Fit an FM band-stop filter as the first part after the cable. Lower the gain | | Aircraft voice from 118 to 122 MHz is weaker with the FM filter | The filter's edge: it is 3 dB down at 122 MHz | Accept it, or use another antenna for voice | | Both radios behind the splitter are weak | The split costs each leg more than 3 dB | This is normal. Measure it. Then test an amplifier in front of the splitter | | A radio behind the splitter runs hot, or the loop has no signal | DC is wrong: a bias tee is on, or the limiter blocks the loop's power | Bias tee off everywhere. Injector first on the loop cable | | Shortwave is full of signals that should not be there | No 30 MHz low-pass filter, so signals above 32.4 MHz fold back | Fit the filter (section 13) | | NAVTEX and the low bands are weak | The F3060 shelf filter is on | Set its jumpers to BYPASS | | Shortwave is noisy on every band, loop indoors | The loop hears the house | Move the loop outdoors and away from the house. Turn it to put a null on the noise | | No signal at all on shortwave after you transmitted nearby | The limiter's resistor may have burned out | Read TAPR's manual: the board has two spare resistors that a solder bridge selects | | A feed was good and got worse after rain | Water may be in an outdoor connector | Open it, dry or replace it, and wrap it again | | The count changed, but you also changed the hour | Traffic, not the part | Compare the same hours (section 14) | | The 1090 MHz range is short with a long thin cable | Cable loss | Add up the loss (section 7). Use a thicker cable or a shorter run | ## 17. What I did wrong 1. **I started with the antenna indoors.** A small whip next to the radio heard about 5 ship messages a minute. A low-cost vertical on the roof heard about 30. Put the antenna outdoors before you buy anything else. 2. **I added an amplifier that the radio did not need.** I expected more messages. I got fewer than half: 42 an hour in place of 95. Filter first. Amplify only to make up for a loss that you know about. 3. **My shortwave loop was still indoors when I made the episode,** next to the computer. For a small loop, distance from the house electronics matters most. I knew that, and I had not fixed it. 4. **I did not measure every change.** I have no number for the aircraft antenna's move to the roof. I have no number for what the splitter cost. I have no number for what the FM filter gains on the discone. A change with no count is a guess. ## 18. Not run and not confirmed - No step in this guide was run on a new, empty station as written. The chain is the one on my station. The figures are each maker's. - **The cable table.** I computed it from the formula in a Times Microwave selection guide that a dealer hosts. The maker's own site refused my requests on 2026-10-10. The guide has no date that I could find. I did not measure any cable. - **RG-6 loss on shortwave** ("under 1 dB for 10 to 20 m"). It comes from my full-station guide. I did not read a maker's data sheet for it today. - **The loss of an adapter.** I found no maker's figure and give none. - **The splitter's cost in messages.** Not measured. The 3.15 and 3.28 dB are the maker's typical figures at 140 MHz. - **That the splitter passes or blocks DC.** Its data sheet does not say. - **What the FM filter gains on the discone.** Not measured. I only saw no loss. - **An amplifier at the antenna end, or in front of the splitter.** Not run. Section 12 gives the reasoning, not a result. - **The cause of the amplifier result.** Overload is my best reading. I measured the count, not the cause. It was one test on one day. - **The tool that gave my hourly counts.** The episode's fact sheet records the numbers, not how I read them. - **Way 2 of section 14,** two radios at the same time. Not run. - **The square-root rule and my "two square roots" limit.** A rule of thumb. Messages come in bursts, so it understates the spread. - **The command that reads `messages` from `aircraft.json`.** The address and the field name are from the projects' documents. I did not run this line. It needs `python3` on the computer. - **The statistics pages.** The AIS-catcher Statistics tab, the one-minute log summary and the ACARS Hub charts are from each project's documents, read on 2026-10-10. I did not open the pages today, so I do not give menu names. - **The spectrum test for FM** (section 10, step 3). The sign to look for is the filter maker's. I name no program and give no steps for one. - **Loop placement figures.** "Several meters" and "about 2 m" are from LZ1AQ's page. I did not test them. My loop was indoors. - **Whether my loop is outdoors now.** It was indoors when I wrote the episode. - **The two warnings in section 9.** They state well-known hazards. I give no distance from a power line: ask your electricity supplier for its figure. - **The quarter-wave vertical** in section 2. I did not build one. - **Grounding.** I read a summary of NEC 810.21 (2020 edition), not the code. I did not read the code of any other country. I did not read the ARRL lightning articles in full. - **Radios other than mine.** I did not check which shortwave radios need a low-pass filter. I checked only the direct-sampling case. - **Prices.** They are the prices of 2026-10-07 in my full-station guide. On 2026-10-10 the RTL-SDR Blog's older product page for the FM filter showed $14.95, and the FlightAware store showed its antenna as sold out. - **"FM stations are far stronger than a message from an aircraft."** General knowledge. Not measured at my station. ## 19. Sources Read on 2026-10-10 unless the line says otherwise. **The episode.** The fact sheet, the links file and the script of episode 11, "Antennas and filters". **The full-station guide.** [Build the station](/guides/build-the-station/), revision A, checked 2026-10-07: the parts, the prices, the order of each chain, FlightAware's cable advice, the terminator test, the level rule for the RX-888, and the traps. **Cable.** Times Microwave Systems, "Coax Selection Guide" (attenuation table and formula), as hosted by a dealer: **Filters and splitter.** RTL-SDR Blog FM band-stop filter: · Nooelec Flamingo FM: · Mini-Circuits ZX10-2-12-S+ data sheet: · Turn Island Systems F3060 manual (May 4, 2025): · TAPR RF Limiter page and manual (revision B): and **Antennas.** FlightAware 1090 MHz antenna: · LZ1AQ, wideband active small loop: · Background, from the episode's links, not read again today: , , , **Grounding and lightning.** Mike Holt's summary of NEC 810.21, 2020 edition: · ARRL, grounding: · ARRL, lightning protection, with its list of articles: **Counting.** readsb, `README-json.md` (the `messages` field of `aircraft.json`): · graphs1090: · docker-adsb-ultrafeeder: · AIS-catcher (the Statistics tab): · docker-shipfeeder (`VERBOSE_LOGGING`, a summary every 60 seconds): · docker-acarshub (README and its feature list): **Law.** 47 U.S.C. 605: (read for the full-station guide on 2026-10-07). [^coax]: Loss in dB for 100 feet = k1 × √f + k2 × f, with f in MHz. LMR-400: k1 0.12229, k2 0.00026. LMR-240: 0.24208 and 0.00033. LMR-195: 0.35686 and 0.00047. LMR-100A: 0.70914 and 0.00174. I divided by 3.048 to get the loss for 10 m. The guide's own table gives 1.5 and 5.1 dB for 100 feet of LMR-400 at 150 and 1500 MHz, which agrees. [^nec]: Mike Holt Enterprises, "Understanding 2020 NEC Requirements for Bonding and Grounding", Article 810: .