| Home | Audio mag. | Stereo Review mag. | High Fidelity mag. | AE/AA mag. |
Noise (part 2) Last time, Mr. Ruck reviewed various forms of power-line-related noise and the effects it might have on electronic equipment. In Part 2, he will discuss electromagnetic pollution, or radio-frequency interference (RFI), which is carried in free space. Virtually every electronic device radiates some noise in the electromagnetic spectrum. In most cases, this is not enough power to be a problem, but when a substantial amount of power is radiated (generally on purpose), it can cause trouble. Most equipment is designed to operate in a relatively benign electromagnetic environment. Since only a small fraction of all electronic equipment designed for home use ever has a problem, manufacturers do not spend the time and money to make their equipment RFI-proof. Even much of the equipment sold to the broadcast and professional audio markets are not totally RFI-proof. When a piece of equipment displays interference, it is usually up to the owner-not the manufacturer or the source of the interference-to solve the problem. I often receive calls from people who experience these problems with their equipment. If the person is capable of “operating ” on the equipment, I can suggest simple fixes along the lines of those in this article. Unfortunately, when the caller has no ability in this area, I can only recommend finding a qualified technician. Where Does RFI Originate? If you have ever stuck a piece of wire into the input of a spectrum analyzer, you know how much RF is out in the ozone. But what are the most common RF sources? A huge amount of energy is radiated at 60Hz in the US (50Hz in the rest of the world). While this is not normally considered RF, it certainly is electromagnetic energy. The only difference is that each wave is very long. Above this are the very low frequency (VLF) stations, which broad cast in the 20 to 50kHz range. Most of these stations are operated by navies to communicate with submarines. The next band, called low frequency (LF) has navigation aids (LORAN) as well as several forms of communications. The first band that can cause problems for the average person is the medium wave (MW) band, otherwise known as AM radio. These frequencies range from 535 to 1,605kHz, and transmitters range in power from 1kW to 50kW, at least in the US. In other parts of the world, powers of up to 250kW are common. Since radio stations want listeners, they try to radiate as much of their transmitter power as possible. When you are close to the antenna, this can cause problems. There are many users of RF above the MW broadcast band, including radio amateurs, fixed point-to-point links and international short wave (SW) broadcasting. Only if you live very close (i.e., next door) to a transmitter will you have any problems with this band. If you live in the major lobe of the directional antenna from a 250kW SW transmitter plant, however, your house lights won't go off! At one time, almost all international telephone and telegraph ser vices were carried through fixed point-to-point SW links, but most of these have been moved to undersea cable and satellite links. A lot of unused spectrum is now in the SW bands. At about 30MHz, the band designation changes to low-band very high frequency (VHF). Until you get to Channel 2 (in the US) television, not much will cause a problem unless you are next door to a transmitter. Even the “ten four good buddies ” at 27MHz CB are not supposed to have more than 5W. From Channel 2 to Channel 6 (television) and from 88MHz to 108MHz (FM broadcasting), how ever, many sources of power are trying to reach their customers. In this band, the power is called “effective radiated power' (ERP). This is the FCC's way of solving broadcasting's version of the audio industry's difficulty in defining 'peak watts. ” ERP is the product of antenna gain and the power fed to the antenna. Thus, if you have an antenna with a power gain of five and feed 10kW in to the antenna, you can claim an ERP of 50kW. Maximum power can be as much as 100kW ERP (in the US). There are many low-power users above 108MHz, including aircraft and high-band VHF land-mobile two way radios. None of these users is a problem because they all use less than 100W of power. At 174dMHz, however, you find high-band TV, or Channels 7 through 13, which use substantial power in an effort to reach viewers. The ultra high frequency (UHF) band is from 216 to 470MHz. Again, most of these are low-power users, which include aircraft and land mobile two-way radios. From 470 to 890MHz, however, you find the UHF TV band in which the ERPs extend into megawatts. Like VHF users, they are trying to reach viewers, and at this frequency, the receivers have poor sensitivity. The broadcasters make up for this by pumping more power into the ozone. Above this is mostly point-to-point microwave using very low power, so not much can cause problems with your equipment. Another major source of RF power is radar. Depending on how the radar is used, it will range in frequency from about 100MHz to more than 20GHz. Most radars operate in the microwave band with relatively low power and cause very little interference. At KFOG, where I work, we have a problem with Navy ships operating their long-distance search radars in San Francisco Bay. These radars operate in the VHF spectrum with megawatts of power into a high-gain antenna. Every time the radar beam is pointed at the studio, we hear a hum in everything. Every radio, TV and electronic device along the bay must also go nuts. Most long-range air-surveillance radars operate in the same frequency band, although their transmissions are blanked over populated areas. How RFI Sounds The first step in solving an interference problem is to figure out the source of the RF. If the transmitter is using amplitude modulation (AM), it will generally be rectified, and you can hear exactly what is going on. If you listen long enough, you can get the call sign, and with a little research (the FCC is most helpful), you can determine the source of the problem. If the transmitter is using single side band (SSB), the problem be comes a little more difficult. When rectified, SSB is so distorted that identifying the transmission is impossible. SSB has a characteristic sound that is impossible to describe, but once you hear it, you will never forget it. Most amateur radio operators use SSB in the HF and SW spectrums, and they are normally cooperative in helping solve your problem. Most people think you cannot determine FM interference by listening. Not so. If the envelope detection is linear, all you will hear is a little hum from the transmitter power supply. If a device is nonlinear enough to have an RFI problem, how ever, it is probably not a linear detector. In this case, you can use slope detection, which will recover audio from FM. This type of detection works on the principle that the detector's sensitivity is different above and below the carrier frequency. Thus, the envelope changes, and you will hear the modulation (al though somewhat distorted). Again, if you get the call sign, you have found the source of the problem. Television has a characteristic sound, too. Since TV uses AM for the visual carrier, it is easily detected and sounds like a “buzz ” at 60Hz. Actually, it is a combination buzz at 59.95Hz (vertical sync) and 15,575Hz (horizontal sync). Also, the timbre of the buzz will change with picture changes, as the video level changes between sync pulses. If you are not sure which station is causing the problem, monitor your TV set and try to correspond buzz changes with picture changes. If you stay up long enough, maybe you will catch the station going off the air at the same time the buzz stops. Normally, the TV sound carrier is at a level far below the visual carrier, and because it is FM, it is hard to hear below the buzz. One easy method of verifying that the buzz in your equipment is television interference rather than a power-supply ground loop is to look at it on a 'scope and trigger the horizontal from the power line. If it is power line related, the waveform will be stationary, but if it comes from rectified TV, the waveform will slowly drift across the 'scope tube. Radar is a periodic hum in your equipment. The frequency of the hum is the radar's pulse repetition frequency (PRF). Long-distance search radars have low PRFs and move slowly, so there is a relatively long time between “hums. ” Short distance search radars (such as typical ship radars) have high PRFs and turn a lot faster. Other types of radar do not rotate, but either track an object or swing up and down (height finders). I doubt that you will have a problem with these unless you live very close to an airport or near a military test range. Don't assume that you will actually hear cases of RFI. Depending on the specific kind of RFI and the specific equipment involved, it is possible to have a problem that is not obviously audible. For example, you might detect an FM station well enough so that there is no slope detection. In this case, the DC level from the detection might shift the operating parameters of the circuit sufficiently so that the circuit no longer operates as designed. Another possibility is that the circuit is linear enough so that it does not rectify an AM radio station, but passes it along to the output. In this case, you might have an amplifier that is operating in slew-rate limiting in an effort to amplify a signal at about 1MHz. This results in distortion in the audio. In the latter two examples, you would probably look for the problem in all the wrong places. The symptoms might also be intermittent because changing the connection schemes changes the RF pickup. I once spent a lot of time chasing what looked like oscillation, but was really AM pickup. What To Do Most cases of RFI involve a situation in which the transmitter is operating within the rules, so aside from sympathy, you will not get much from the local FCC. Threatening the transmitter operator with the FCC will not get much action either, unless your problem stems from a 'ten four good buddy’ operating with an illegal RF power amplifier. Determining the source of the problem is important, however, be cause the cures for a MW RFI problem might be different from those for a VHF or UHF RFI problem. For in stance, grounding equipment to a cold-water pipe might help a MW RFI problem, but not a VHF or UHF RFI problem. The reason is that at VHF and above, the conductor length be comes critical. Ten feet of ground wire between your equipment and “earth ” is enough to have substantial impedance, and if you happen to pick up a resonant length, you can make your RFI problem worse in stead of better. RF vs. Audio Generally, the best form of power and ground distribution is the “star ” system. Figure 1 shows a typical amplifier (greatly simplified) with a “star ” ground system. Note that the input and output connector grounds are not connected to the chassis, but are insulated and brought to the single-point common-ground connection. This is great for audio because the wire is extremely short compared to an electrical wavelength. At RF, how ever, even a few inches of wire can become a substantial part of a wave length (especially at UHF TV). Thus, the shield can carry RF into the device, and before it reaches the common-ground point, it will re-radiate RF into the input of the amplifier's first stage. This is shown in Fig. 2. As you can see, even when you follow good audio practice and care fully shield the interconnect wire, you can still induce RF into the amplifier input. This can happen on the amplifier output as well. RF can follow the ground lead, which will re radiate RF back into an otherwise shielded box. This problem is not limited to home audio equipment. My Nagra IV-S had a ground wire that went from pin 1 of the microphone input jack through the microphone pre amps to the chassis. I had an identical problem with a “professional ” turntable preamp. One easy way to diagnose this problem is to hold the outside of the wires with your hands. If the problem changes when you move your hands up and down the wires, you probably have a “hot shield ” and might not need any major modifications. The solution is simple. Bypass the input and output ground directly to the chassis ground as close to the connectors as possible. Connecting ground to ground might sound strange, but you must remember that ground at audio is (generally) not ground at RF. Figure 3 shows a properly bypassed amplifier. If you use a relatively small (0.01u4F) capacitor, it is essentially an open circuit at audio, but will bypass the RF on the shield directly to the chassis. It is important to use a capacitor with low internal series inductance, so I recommend (horrors!) ceramic disk capacitors be cause they have far lower series impedance at RF than most other types. Since you are essentially connecting the capacitor from “ground ” to “ground, ” it cannot contaminate your audio signal at audio frequencies. Because no audio or DC voltage is developed across this bypass capacitor, you can use as low a voltage capacitor as you can find, which is also good for RF because it is physi cally small. I do not recommend extremely tiny, multilayered ceramic capacitors unless you are really pressed for space because they do have some internal series inductance. Try to install the bypass capacitors with as little lead length as possible and be sure that you make a good connection to the chassis. I suggested this cure first because it re quires no modifications to your equipment's signal path. Help for Bad Cases If you have a situation where the RFI is in everything, try shielding your entire setup. This is a lot easier than it sounds. All you have to do is build a large box to house your equipment and line the inside with aluminum foil. (The heavy-duty type is easier to handle.) Make sure you overlap the seams to provide good conduction between the pieces of foil. In many cases, this fix will eliminate your RFI problems. If you still have a problem after trying the shielded box and bypassing all your inputs, you need some “black magic. ” Once again, test the wires for 'hot spots. ” If you find a particularly hot place, try moving the wires to relocate that spot in a “null ” of the RF. Sometimes changing the length of your interconnection wires will also cause the unit to become resonant at another, noncritical frequency. Check the power line. Perhaps it is bringing in RF and reradiating it within your shielded box. Bypassing the power line as it enters your box
FIGURE 1: A typical amplifier, simplified, with a “star ” ground system. This is generally accepted as the best power and ground distribution setup. FIGURE 2: A shield carries RF into the device, then reradiates it into the input of the amplifiers first stage before the RF reaches the common-ground point.
and then using shielded cable for power-supply cords inside the box will minimize this problem. For safety’s sake, use capacitors rated for at least 1kV and bring a good ground to the capacitors' common. Also tie your aluminum foil to that ground. This will not necessarily help reduce the RF, but it will keep your beneficiaries from collecting on your life insurance policy. RFI Filters Most radio/ TV stores stock “off-the shelf' filters. Depending on which frequency is causing a problem, they might work. Unfortunately, an un shielded power-line cord between a filter and your equipment can act as enough antenna, especially at VHF frequencies, to cancel any relief from RFI. They are certainly easy to try, but don't get your hopes up. Many manufacturers build RFI filters into their equipment designs. Generally, the filters are designed to keep the digital circuitry's RFI from “leaking ” out of the enclosure or to keep RF out of the enclosure. Most of them have relatively good attenuation from about 1 to 100MHz, but many do not provide enough suppression to keep UHF TV from creeping up the line cord into your equipment. For example, a Panel Components Corporation power-line filter (Number 8843.N1.20.60), as shown in Fig. 4, has 72dB of insertion loss at 4eMHz but only 47dB at 100MHz. The loss above 100MHz is not even specified. Once again, knowing the frequency of your RFI problem will help you prescribe the most effective cure. Good audio construction practice dictates that power-line-related wiring should be isolated from audio wiring. If you build the power supply FIGURE 3: Bypassing the input and output ground directly to the chassis ground can eliminate a “hot shield. ” FIGURE 4: A Panel Components power-line filter helps eliminate RFI at some frequencies. FIGURE 5
in its own “box ” and bypass the DC outputs, there is little chance that RF can get into your circuit through the power line. Use a 0.01 uF ceramic disk capacitor or even better, a 'feedthrough-bypass’ type that also provides a convenient method of get ting DC out of the power-supply enclosure while keeping RF in the enclosure. You may also combine an RF filter with other amateur designs. For in stance, you could make an electrical distribution system that includes transient as well as RF protection. An example of this is shown in Fig. 5. Remember, though, that RFI in your equipment at VHF and UHF frequencies might not be attenuated by such a device, as line cords can pick up sufficient RF at these frequencies to negate any improvements. Examine Your Options If the RFI is so bad that the simple cures outlined above have not helped you, I would consider moving rather than rebuilding your equipment! You must determine where in the circuit the RF is causing the problem and then shunt the RF to ground before that stage or reduce that stage’s susceptibility to RF. You must also be careful not to degrade the audio in any way. One of the better references for RFI suppression is the Radio Amateur's Handbook, published by the American Radio Relay League (ARRL). Many hams are quite knowledgeable and have practical experience in this area. The ARRL also publishes a special booklet titled Radio Frequency Interference-How to Identify It and Cure It. While it is written for the radio amateur, it covers solutions to audio problems. Other sources of information are publications from the Consumer Electronics Group of the Electronic Industries Association. They have a booklet titled Consumer Electronics Service Technician Interference Handbook-Audio Rectification, which goes into some detail about solving this problem. They have another similar booklet about television interference. Although that booklet refers to TV problems, if you have RFI in your stereo, you probably have TV problems as well. The book lets are free from the Consumer Electronics Group (see Sources for the ad dress). Before you consider bypassing the audio circuitry, check the power-sup ply wiring. In many cases, designers of audio grounding and DC power distribution systems do not consider the length of the wires. Bypassing the power supply (both plus and minus if bipolar) and ground leads to the chassis at appropriate points might solve the problem. Because you are bypassing either DC or ground to the chassis ground, you can use ceramic disk capacitors. I recommend 0.01 uF as a starting value. Bypassing audio circuits to ground is not easy. Capacitors that are acceptable to audio, such as polystyrene or polypropylene, use a rolled construction, which has sufficient self inductance to make them ineffective at RF, particularly at VHF and above. On the other hand, I do not recommend ceramic capacitors for this application because you are placing them across an audio signal path. It is difficult to suggest an exact the capacitor is another solution to RFI capacitance value. In short, use as large a value as you can live with. Normally, you can select a value that does not affect the audio high frequency response, but has sufficient reactance at RF to reduce that stage ’s susceptibility to rectification. Remember, you do not have to eliminate the RF, just reduce it so that it does not force the stage to rectify it. You must also insert something in series with the audio path to increase the impedance at RF sufficiently. In some cases, a small (10 0-ohm to 1 k-ohm) resistor is sufficient, but you probably do not want to put that large a resistor in series with a moving-coil cartridge. Another alternative is a small RF choke. This adds inductance without resistance, but is sensitive to AC fields. You might solve RFI, but add a 60Hz hum problem. A third alternative is a ferrite bead.
When slipped onto a wire, this will cause the wire to have impedance,
FIGURE 6: Using a ferrite bead before the capacitor is another solution to RFI problems.
but not so that it is resonant or susceptible to magnetic interference. Beads are made from slightly different materials and have different frequency characteristics. For example, beads made from manganese zinc ferrite (“B ” material) work at MW frequencies, while beads made from magnesium-manganese-zinc ferrite ( “F ” material) work better at VHF frequencies. Bargain beads do not indicate their composition. Since bead kits are relatively inexpensive, I suggest purchasing those that contain beads of known composition. Ferronics Inc. -------------- The Spike Master As part of the research for my article, I uncovered a product from Disc washer that is very similar to my Fig. 5. The major difference is that it includes two additional MOVs be fore the series chokes. The entire circuit is neatly packaged in a plastic box and includes four grounded out lets, as well as a circuit-breaker, an on/off switch and a pilot light. I have no doubt that the Spike Master is very effective in clipping any transients on the power line be fore it gets to the equipment plugged into its outlets. The combination of all five MOVs would eliminate most transients that might damage your equipment. You must, however, temper Disc washer's claims about RFI suppression with the fact that the line cord for any equipment plugged into the Spike-Master will act as an antenna, bypassing the device's RF attenuation. This is true of any “plug-in ” RFI filter. In addition, the values Discwasher is using, as well as the unit's plastic construction, make it relatively ineffective at VHF and UHF frequencies. I tried using the SpikeMaster on my home audio system and noticed no difference. I also tried it in place of the high-isolation regulating transformer I normally use for my computer and had a couple of sys tem crashes, which might have been caused by sags or surges against which the SpikeMaster is not designed to protect. Within these limitations, the SpikeMaster is well designed and nicely packaged. It certainly could not hurt to have one in series with your audio system or computer. ---------------- has kits for about $15 that provide a good selection of beads. The kits also include information on bead applications. A typically modified stage is shown in Fig. 6. Note that the capacitor follows the resistor/inductor/bead. Be careful where the “ground ” side of the capacitor goes. If you do not pick a good “ground ” spot in the circuit, you can make things worse. I have found instances where different parts of the circuit board worked for otherwise identical channels. Do not assume anything, and if you start to get cocky, I guarantee that a particularly quirky problem will bring you down to earth. 'Cut and try’ is the rule with stubborn RFI problems. The booklet from the Consumer Electronics Group suggests a “pi ” network in series with the transistor base consisting of a series RF coil and two shunt capacitors. It suggests about a 5 to 10mH coil and 250 to 500pF capacitors. (Common values are 220 to 470pF.) The booklet also suggests an L-type filter in the collector power-supply circuit, with a series coil and shunt capacitor near the transistor collector, using values similar to those in the pi circuit. I have never needed this much attenuation, and the values are most appropriate for MW RFI (i.e., an AM radio station). One other possibility is a small capacitor (100pF or so) between the base and collector. Depending on the situation, this might make the problem worse by forcing the emitter base junction to rectify the RF, or it might make it better by reducing the high-frequency gain of the transistor. Finally, the capacitor might not be necessary at all. Sometimes you just have to raise the series impedance at RF with a series resistor, choke or ferrite bead so that the distributed capacitance in the circuit bypasses the RF. Once again, “cut and try. ” Wishful Thinking Someday, when I have no financial restraints, I will build myself a “perfect ” listening room. In addition to careful acoustic design, I will include copper-screen shielding on all six sides, soldered together and “earthed. ” All power entering the room will go through a high-isolation transformer. Until then, however, 1 will have to deal with the realities of electromagnetic pollution. hope my suggestions will help you deal with these problems, too. SOURCES 1. Ferronics Inc., 66 N. Main St., Fair port, NY 14450. 2. Consumer Electronics Systems Technician Interference Handbook-Audio Rectification and Consumer Electronics Systems Technician Interference Handbook- Television Interference, Consumer Elec tronics Group, Electronic Industries Association, 2001 Eye Street NW, Washington, DC 20006. 3. Panel Components Corporation, PO Box 6626, Santa Rosa, CA 95406. 4. The Radio Amateur's Handbook and Radio Frequency Interference-How to Identify It and Cure It, American Radio Relay League, Newington, CT 06111. ------------------- Also see: Noise in hi-fi audio gear (part 1)
|
Prev. | Next |