Last Modified: November 14, 2013
Contents: Basics; Measuring Beads; What Size; Where to Put Them; Where Not to Put Them; Stereos; Odds & Ends;
Ferrites are mixtures of iron oxide and one or more metals typically manganese, nickel, and zinc. Occasionally rare earths such as yttrium and scandium are also added (they are not always or predominately iron oxide, and may contain “soft iron” meaning magnetically soft not physically soft). They provide high magnetic permeability and high resistivity, although some formulations (known as mixes) are conductive. Combined with a variety of stabilizers and binders they can be molded to just about any desired shape, with toroids, bars and beads the most common ones encountered by amateurs. By selecting the right mixture of metals, initial permeabilities (ui) from 10 to as high as 5000 or more are easily obtained. To a lessor extent the temperature coefficient can also be adjusted to meet a specific use. There are literally thousands of uses for ferrites, and modern amateur transceivers are loaded with them. But it isn't their use as RF transformers or baluns or ununs we're concerned with in this article. It is their use as RFI suppressors, and primarily in split bead configuration.
When the right material is used, they can cure a myriad of RFI problems. If not, you're wasting your time and energy. Although this article explains how to determine which material will work, it is by no means an absolute. If you don't have a way to measure them, you're much better off buying new beads with known parameters. Remember too, some mixes work better for HF, while other work better for VHF. A few salient points need to be made however.
Just because a specific ferrite mix has a higher initial permeability (ui), does not necessarily mean it will have superior performance over one with a lessor ui. Again, it depends on the specific use, frequency of operation, and a few dozen other parameters.
The reactance of any choke, is a factor of the square of the number of turns. That is to say, 4 turns through 1 bead, is the same as 1 turn through 16 beads! Further, turns should always be as evenly spaced as possible, and not overlapped. Doing otherwise will reduce the effectiveness (reactance) of the choke.
A single turn choke, no matter the mix, is worthless for RFI suppression at HF frequencies.
Installing beads on DC power cables does absolutely nothing towards reducing RFI. If RFI issue is induced on a DC power cable (a very rare occurrence), the choke should be installed only in the positive lead. Installing one on the negative lead has the opposite affect, as it RF isolates the ground connection.
Surplus ferrite beads from unknown sources shouldn't be used. Most, like those from Radio Shack, and All Electronics, make lousy chokes at any frequency below ≈150 MHz.
Specific mixes required for baluns, are not necessarily the ones required for use as chokes. Jim Brown, W9YC, has a series of on-line articles all about chokes, baluns, common mode current problems, and other data. While some of it is esoteric in nature, the articles are well written, and easy to follow.
Split beads, our main interest here, are tubular chunks of ferrite split along their length and typically mounted in a plastic enclosure as shown. They're available in a variety of mixes, inside and outside diameters, and with or without enclosures. Since our focus here is high frequency RFI suppression, we'll concern ourselves primarily with just one, mix 31. Just for the record, mix 43 is still a good choice for frequencies over 15 MHz or so, and is readily available from a variety of sources.
A mix 31 split bead has an initial permeability (expressed as ui) of 1,500 and a nominal operating range of 1 to 500 MHz. When placed over a wire where there is RF energy flowing (between one and several hundred megahertz) it is equivalent to placing an inductor and resistor in series with the wire for RF currents. Depending on the frequency of the RF energy the equivalent impedance can be as high as 5 k ohms or more yet DC or audio frequencies pass through unrestricted. The resistance to the flow of RF current is closely tied to the linear inch of material parallel to and surrounding the wire, and it goes up by the square of the number of turns. It is actually a complex impedance, with X decreasing and R increasing with increasing frequency. At some frequency X=R (loss tangent equals effective permeability) and the Q becomes unity [Q=1]. It is this property which makes their use in RFI suppression so ideal, and the best part is you don't have to cut the wire! There are Smith Charts, and loss plots on Mixes 31, 43, and 44 in the Photo Gallery.
Good, high quality split beads are not inexpensive. From $3 to $5 each, multiplied by 20 or more, and it can easily run into a healthy sum. This fact has pushed many amateurs to seek relief by purchasing surplus units from after-market sources. The problem is, you don't know what you're getting, as some of these surplus split beads are virtually worthless for HF RFI suppression. Knowing what you have is therefore important.
In order to be effective, the impedance of the choke must be mostly resistive within the frequency range we're trying to cover. If they're mostly inductive, applying them may actually makes matters worse, not better! This is also the reason single turn chokes, even multiple ones in series, are nearly worthless at HF frequencies. Steve Hunt, G3TXQ, compiled a chart showing the difference between mix 43, and mix 31 in several configurations, and it is located here. The black line in the charts indicate when the reactance of the choke is mostly resistive. In you're specific application, if the black line does not fall within your operating frequency range, the choke might increase the RFI, rather than suppress it.
In some cases, RFI from electronic devices, including fuel pumps, can be best cured by capacitively bypassing the leads to ground, and across one another. However, caution should be exercised! This technique should never be used to bypass speaker systems as noted below. Good wiring practices must be followed, including the use of heat shrink tubing, and the avoidance of any type of tape material, not matter what it is! Remember! Any capacitor, ceramic ones in this case, can fail short. If the circuitry is not properly fused, the results could be catastrophic! Use bypass caps as a last resort, and only if you're technically competent to install them correctly.
The section explains how to measure the reactance of a split bead using an inexpensive antenna analyzer. While effective, it isn't a panacea, and here's one reason. The reactance of most ferrite mixes are rated +Ø, -20%, thus two closely related mixes might measure the same reactance. To specifically identify a ferrite material requires a VNA (vector network analyzer) costing tens of thousands of dollars. This said, the antenna analyzer is close enough to weed out worthless surplus beads, and that's one of the things we're trying to do.
Nowadays, the MFJ 259B Antenna Analyzer has become almost ubiquitous in the modern ham shack. You can use one to check those surplus units to make sure they'll do the job. You'll need enough hookup wire, size 22 is ideal, to make 3 passes through the bead plus enough to connect the ends to the analyzer. Set the frequency to 2 MHz, and measure the reactance. If it is mix 31, the reactive value (X) will be approximately 400 to 500 ohms. Pushing the mode button three times will bring up the inductance menu which should show 40 uh or so.
Putting mix 43 under same test will require you to increase the turns to 5. The readings will then be approximately the same as mix 31. By the way, the 259B doesn't have enough range to check either mix at much more than 2.5 MHz unless you reduce the number of turns. This is because the 259B has a maximum reactive range of 650 ohms. Either test will exceed this value at approximately 2.5 MHz.
As pointed out previously, at some frequency X=R and the Q=1. For either mix this occurs near 40 MHz for a one turn core, and approximately 20 MHz for a two turn core. Attempting to measure the crossover [X=R] point is beyond the range capability of the 259B. For those who wish to get closer to the actual mix specifications, loss tangent charts are available from a variety of sources which precisely list these crossover points.
Split beads come in just about every length and diameter you can think of. However, those with snap-on plastic housings generally come in four internal sizes; .25, .375, .5, .75, and 1 inch, although the actual ID may vary slightly from these sizes. Lengths vary too with the larger ID stretching to 1.5 to 2 inches. Although you can buy split beads without the plastic covers, they're inconvenient to attach (or remove) and the difference in cost is not worth the extra effort to attach them.
Just as important as knowing where to install them is how to install them, and which ID size to use. Here are a few tips. It really doesn't matter if the bead is tight or loose when snapped over the cable in question. If it is too tight you run the risk of abrading the wire, and this condition should be avoided. Too loose and it won't stay in place, but a well-placed tyrap will keep it where it belongs. Where it belongs is also important.
Where to Put Them
Probably the most ubiquitous use of ferrite beads is in the form of an antenna motor (and reed switch) lead choke. This requires special winding considerations if they're going to be effective, and the Antenna Matching article explains how. They're also used quite often for common mode current suppression.
When used for RFI suppression, beads should be installed as close to the offending or offended device as possible. While this is not always easy the results are worth the extra effort. In some cases it is impossible to get close enough for optimum results. On many vehicles the fuel injector units are under the intake manifold. About the only solution is to place large beads over the cabling as close to the devices as possible.
Again, it is very important to know which mix you have. Installing the incorrect mix is a lessen in futility, and as pointed out above, may actually increase the RFI problem you're trying to suppress! Mix 31 is the one you want. Mix 43 is a viable alternative on the VHF bands. Both mixes are available from DX Engineering, Mouser Electronics, and perhaps others.
The first and most important is placing them on the power cables leading to devices which interfere with you (egress). Good examples are electronic fuel pumps, AC and cooling fans as shown in the left photo, COP (coil over plug) units, fuel injectors, IAC (idle air control) motors, windshield wiper motors, and ABS motors as shown. Place them as close to the offending device as you can.
The second is their placement over power cables for devices which you interfere with (ingress). Alternator control circuits (not the DC power out), ABS control units, cruise controls, AM/FM radios and their amplifiers, traction control systems (a headache I’ve experienced), navigation systems, and even some windshield wiper and cruise controls. Here too, they need to be placed as close to the affected device as possible.
If you own an Icom 706, and have one of the early extension cables without the factory installed beads, you’ll need one on each end as well as shown in the left photo. If you don't head this advice the radio has a tendency to lock up or shut down, and most likely erase the contents of your memories.
If your vehicle has an LED CHMSL like my last one did, you should bead the lead where it enters the trunk area. Remember, an LED is a diode and they will rectify RF and cause all sorts of RFI problems, especially at VHF when the antenna is mounted atop the trunk lid.
If you have an OBD II (On-Board Diagnostics, generation two) reader, it can be an effective tool in locating some types of RFI ingress. For example; if transmitting causes the engine to miss fire, a code will be sent to the OBD II unit which may (or may not) turn on the Maintenance Indicator Light (MIL). Sometimes, the codes will indicated a specific location. On one of my previous vehicles, the code indicated a problem with the motor controlling the secondary throttle plates. A bead placed over the motor's harness cured the problem, and the code never again showed up.
Most OBD II readers have a reset function to clear the codes which turn on the MIL. Most dealers charge $50 or more to reset them even if there is nothing wrong. In some cases, it pays to buy a reader from your local auto parts store. Typical units cost from $35 to over $100. More information about OBD II RFI is located in the Digital Electronics article.
Ford Motor Company products used to have a bad reputation with regards to fuel pump RFI. Since 2003, that's no longer the case. However, most automobile manufacturers including Ford, are increasingly relying on vehicle-wide data bus systems. These digital system can and do cause all manner of egressed RFI. The typical result is evenly-spaced birdies throughout the HF spectrum. Since the bus is wired throughout the vehicle, attempting to suppress them with ferrite beads is a daunting task. It should be mentioned that excessive ground losses increase the magnitude of the received RFI due to common mode issues. It should be evident that RFI suppression requires more effort than a few well-placed ferrite beads.
Where Not to Put Them
Where not to place them includes low impedance DC power cables. If you adequately sized your power cabling (see my Wiring article), the use of split beads (or brute force filters) is a waste of resources.
Using them in an attempt to cure alternator whine is also a waste. In most cases, alternator whine appears when there is a ground loop caused by incorrect wiring practices. It is also quite common when mag mount antennas are being used.
Real alternator whine is caused by one or more leaky diodes. Older Delco® products seem to be the worse for this malady with older model Toyotas in second place. Brute force filters may help, however ones large enough to handle a 100+ amp alternator are usually more costly than fixing the alternator, to say nothing of the voltage drop they cause.
Stereos systems can be problematic when it comes to RFI ingress. Speaker, interconnect, and power cabling are all susceptible to RFI, and shielding them is out of the question. Beads do help if installed close to the electronics as possible. However, the power amplifiers used in most models are connected to DC power any time the ignition key is inserted even if the system is turned off! Whether it is powered or not, some designs allow rectification of the offending RF. Low level audio from the speakers is one thing, but fully amplified audio is another. GM products are particularly susceptible in this regard (think OnStar®). In some cases, it maybe necessary to add a secondary power switch to the amplifier.
Placing capacitors across speaker terminals should be avoided as this can cause the offending amplifiers to go into oscillation, and destroy themselves. There is a further caveat with respect to some mobile transceivers, and that is floating outputs. That is to say, the speaker output is not referenced to ground. Thus grounding either lead can cause the audio amplifier to fail. Yet another reason not to use built-in stereo systems as the audio output stage.
Odds & Ends
Mouser Electronics also sells mix 31 beads that are not split; they're cylindrical. What's more, they're expensive. However, there are times where these beads are the best choice. Since the outside is cylindrical, they're easier to wind than split beads. Their added length also increases their reactive component by about 20% over the same size ID split bead. If you decide to use cylindrical beads, buy yourself some 3M #27 glass tape from Mouser too. Use it to cover the ferrite, in and out, then wind the wire on, and then protect it, in and out, with #27 as well.
Beads work very efficiently for most types of RFI and EMI, but they are not a cure all. Sometimes the installation of multiple beads is still inadequate to cure stubborn cases. In these few cases, it might be preferable to install discrete capacitors and chokes, as a last resort. However, be very careful where and how you install them.
Modern vehicles use a lot of solid state devices to switch and control the various functions. Haphazardly installing noise suppression capacitors and chokes could spell disaster. If in doubt, leave it out. In any case, always seek professional advice before taking this route.
I'm often ask if one type or model of vehicle, or manufacturer is noisier than some other. I'll answer that by repeating what I have covered elsewhere on this web site. Some otherwise identical vehicles are noisier than others. Further, some individual sub assemblies are somewhat noisier than their counterparts in other vehicles. I would be a millionaire if I knew why this is so. Worse, some vehicles are literally cars from hell, and no amount of noise suppression will make them quiet enough for amateur radio operation (I've had one nearly that bad). Let's hope yours isn't one of them. In any case, don't give up hope.