The Proper Split Beads to Suppress RFI
Last Modified:
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 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 a modern amateur transceiver is loaded with them. But it isn't their use as RF transformers or baluns or ununs were concerned with in this article, but their use as RFI suppressors, and primarily in split bead configuration.
Split beads are tubular chunks of ferrite split along their length and typically mounted in a plastic enclosure. 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.
I digress: Mix 43 has been touted as the best all around material and it is still a good choice as it is readily available from a variety of sources. A mix 43 split bead has an initial permeability of 850. Mix 31 has an initial permeability of 1,500. What's more, mix 31 offers considerably better low frequency attenuation than mix 43. It's attenuation at 100 MHz is less, but this is not usually a concern as egressed noise seldom exceeds 45 MHz.
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 50 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.
The basic application I just described, a one turn inductor, will suffice is some cases. However in more stubborn cases or where more RFI suppression is required, additional beads may be installed. This is particularly true with tank-mounted electric fuel pumps. If there is enough lead length, multiple turns should be used. If not, you'll need multiple beads.
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 (me included) to seek relief by purchasing surplus units from after-market sources. The problem is, you don't know what you're getting. Some of these surplus split beads are mix 67 (ui=40) which is virtually worthless for HF RFI suppression. Sometimes you might find a bargain if you had a way to check the mix. This takes some rather expensive equipment and since initial permeability can vary as much as 20% between batches of the same mix, you can't be 100% sure even then. What’s a poor ham going to do? Relax, cause here's a way to get close.
Nowadays, the MFJ 259B Antenna Analyzer has become almost ubiquitous in the modern ham shack. And believe it or not, 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 inductive value will be approximately 500 ohms. Pushing the mode button three times will bring up the reactance 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 hasn't 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 [X=R] 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's 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.
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.
Using beads on existing wiring seldom allows more than one turn through the bead, but if you're installing the wiring, leave enough length to make two or more turns. The reason is simple. The amount of the effective resistance is proportional to the square of the turns. That is to say, 2 turns is 4 times that of 1 turn. Put another way, it would take 4 one turn beads to equal 1 two turn bead.
Again, it's very important to know which mix you have. Installing the incorrect mix is a lessen in futility. Mix 31 is the one you want. Mix 43 is a viable alternative especially on the VHF bands. Both mixes are available from DX Engineering, Mouser Electronics, Palomar Engineers, and perhaps others.
Where to place them is also a hotly debated issue. There are three aspects to this and for clarity I want to separate them.
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 in the right photo. 
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 as shown in the photo bottom right. Remember, an LED is a diode and they will rectify RF and cause all sorts of RFI problems.
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 Check Engine light. 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 Check Engine light. 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.
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 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.