As noted in my previous blog regarding PDS design, the whole power supply conundrum is one that has been plagued over the years with a number of erroneous rules-of-thumb; “black magic” design rules and confusion as to what does or doesn’t work.
One of the most controversial topic areas focuses on the use of ferrite beads as a means of controlling and containing EMI. There is conflicting information concerning the use of ferrite beads and it is difficult to ascertain which information is valid and which is not. The real challenge is that the erroneous information looks to be valid because of the large amounts of data associated with it. To add to the confusion, in the application notes of some ICs, the component vendors will recommend using ferrite beads as a means of eliminating EMI.
To help mitigate this confusion, this blog will address several topics including:
• What are the origins of ferrite beads?
• What is the history of their use?
• Why has it been assumed that the use of ferrite beads constitutes a valid design rule?
• What actually happens when a ferrite bead is put in series with the power lead of an IC?
• What do you do when an IC vendor specifies the use of ferrite beads?
As a result of the discussion presented here, it will be demonstrated how the use of ferrite beads in series with a power lead of an IC does not eliminate or contain EMI but, in fact, degrades the performance of the PDS.
The Origins of Ferrite Beads
To address the first point of confusion, ferrite beads are not beads. They are little inductors. The thing that people refer to as a bead is actually a toroid. (A toroid is a coil of insulated or enameled wire wound on a donut-shaped form made of powdered iron. It is used as an inductor in electronic circuits especially at low frequencies where comparatively large inductances are necessary. They have been used forever as the cores of transformers). To maintain consistency, ferrite inductors will be referred to as they are in the rest of the industry—ferrite beads.
Ferrite beads are surface mount components much like other components such as resistors and capacitors. And, they are available in the same sizes as these other components. A typical ferrite bead package is illustrated in Figure 1. Notice that the word bead is surrounded by quotation marks as this part is not, in fact, a bead.
Figure 1. Typical Ferrite “Bead” Package
In terms of composition, ferrite beads are made from a ferromagnetic material commonly referred to as a ferrite. This material behaves like an inductor made from a coil of wire. The attractiveness of this component is that it has a relatively high inductance in a small form factor. Typically, these components are not specified by the amount of inductance they have, but rather by their impedance at a particular frequency. As shown in Figure 2, the impedance of a ferrite bead is a function of frequency much like an inductor with the impedance being quite low at low frequencies, rising to a high point and then dropping off.
Figure 2. Typical Ferrite Bead Impedance vs. Frequency
The History of the Use of Ferrite Beads
The origins of the use of ferrite beads in PCB designs harkens back to the late 1980s when custom CMOS devices finally switched fast enough that they created frequencies in the EMI band. EMI technicians stuck the ferrite beads in the power leads of the devices and the EMI went away because the part could no longer switch fast enough to create the frequencies that were in the EMI band. Thus, the ferrite bead was a band aid. It worked and it stopped the part from making the noise but it also prevented the part from working at speed (i.e. switching fast). When ferrite beads were first utilized, speed, in terms of fast edges, was not an imperative so that is how the use of ferrite beads got its start.
To further illustrate this, when a ferrite bead was placed in series with the power lead of an IC, a circuit, such as the one depicted in Figure 3, was created.
Figure 3. IC with Ferrite Bead in Power Lead
The frequencies involved in radiated EMI range from 30MHz to 1GHZ for most products. When the IC attempted to draw power at high frequencies from the power supply, it was prevented from doing so by the impedance of the ferrite bead. As a result, there were no high frequencies on the IC package to cause an EMI problem. This is one of the two ways to control EMI—eliminate the source or eliminate the antenna. This technique worked as long as the IC or ASIC were not expected to operate with fast edges or fast clocks. Prior to the use of 130 nanometer ICs, most circuits ran slow enough that there wasn’t the need for a very low impedance source at high frequencies. This was actually a case of PDSs working in spite of bad habits rather than because of good engineering practices.
What Do Ferrite Beads Really Do?
What actually happens when you place a ferrite bead in series with the power lead of an IC is that the performance of the PDS is degraded as seen by the device increasing its output impedance. It’s important to remember that a power supply is expected to be a voltage source, meaning that no matter how much current is drawn from it, the output voltage remains the same. In other words, power sources are expected to have zero or very low output impedance at all frequencies in order to properly do their job. As noted above, eventually, the speed of ICs increased to the point that inserting a ferrite bead prevented them from operating as they should. The reason was that the PDS output impedance was too high. The proposed solution was to add a capacitor after the inductor as shown in Figure 4.
Figure 4. IC with Ferrite Bead and Capacitor in Power Lead
This solved the operating problem but brought back the EMI problem. Then, the method recommended for implementing this circuit was to cut an island in the Vdd plane. This is not a valid alternative either (see Reference 1 at the end of this blog).
Notice that in Figure 4, the capacitor is called a “bypass capacitor” with quotes around it. The reason for the quotes is to call attention to the fact that this capacitor is not bypassing noise, rather it is serving as a source of high frequency charge so that the ASIC can again switch rapidly. A much better name for these capacitors is “coulomb buckets” as they are functioning as local storage devices (see my previous blog POWER PLAY—SUCCESSFULLY DESIGNING POWER DELIVERY SYSTEMS, PART 1, for further information on coulomb buckets).
Why Is The Use of Ferrite Beads Put into Application Notes?
It should be noted that in a high speed ASIC, the inductor and the capacitor form a low pass filter preventing high frequency noise from getting to the component from the power subsystem side of the system. This is the reason given in most application notes for placing ferrite beads in series with the power leads of PLL (phase locked loop) devices and other “analog” type circuits including a high speed Serializer/Deserializer (SERDES).
IC vendors have typically recommended the use of ferrite beads in their applications notes for a couple of reasons. First, the author of the applications note will say, “We’ve always done it this way and if you don’t follow our application note, we won’t guarantee that the circuit will work properly.” If such a statement is made, it’s reasonable to ask if the application note is followed exactly, will the vendor guarantee the circuit will work. Often times, the answer is “no.” This certainly doesn’t leave you with much of a comfort zone in using your particular selected IC.
The second reason a vendor will give for specifying the use of a ferrite bead is that the bead is there to block noise in the power subsystem from getting into the sensitive circuit. At Speeding Edge, we have seen examples of this in actual test circuits. The noise is indeed blocked but the circuit performance is likely to be degraded due to poor power delivery to the circuit being protected.
Figure 5 shows the output waveform of a 3.125 GB/S serial link with a ferrite bead in the power lead of the output stage.
Figure 5. 3.125 Gb/S Serdes Output with Ferrite Bead in Power Lead
Figure 6 is the same output with the ferrite bead removed and the power lead connected directly to Vdd. As can be seen, inserting the ferrite bead actually made the circuit perform worse than with no ferrite bead.
Figure 6. 3.125 Gb/S Serdes Output with Ferrite Bead Removed from Power Lead
The circuit for Figure 5 was recommended by the supplier of the part, without first checking to see if the advice was sound. The waveforms shown were actually taken from an evaluation board supplied by the vendor. In terms of blocking noise from the power subsystem, this became treating the symptom rather than solving the problem. The problem was that there was noise in the power subsystem because it was not designed correctly.
The key thing for an IC vendor to understand is the power delivery needs of an IC. This includes the maximum delta I that the circuit may demand of the power delivery system as well as at what frequencies, and the maximum allowable delta V (ripple). Without this information, it is impossible to design a working, reliable PDS.
In reading the specifications for a component, such as an operational amplifier, one of the specifications is the power supply rejection ratio. This is a measure of the amount that variations on voltage of the power supply voltage affect the output of the device. It is possible to make such measurements for digital ICs and PLLs. The idea that ICs are just “logic” and don’t need this level of characterization is left over from the days of TTL when there was such high tolerance for Vcc variations that it was not necessary to account for them.
In actuality, an IC vendor needs to be able to advise users on how to create a functional power system. Any time there is a recommendation to add a ferrite bead in the power lead of a device, four questions must be asked of the IC vendor:
Is there a problem that can be solved by adding a ferrite bead?
Does the ferrite bead actually solve the problem?
Can I be sure that the addition of the ferrite bead does not create a new problem (such as that shown in Figure 5)?
Is using a ferrite bead the best way to solve the problem?
Our experience has been that after answering questions one and two, ferrite beads are eliminated from the design. Whenever we have encountered an applications note that recommends the use of ferrite beads, we have called the IC vendor/author and asked the above four questions. In no instance have we found that answering these questions results in the agreement that the addition of a ferrite bead is a good idea.
If after the foregoing process the vendor still insists on the use of ferrite beads, it’s imperative to insist on seeing a test circuit in which the component is used exactly the way it is intended to be used in the new design. If no test circuit exists, it is good to be suspicious. In one instance, when we were having trouble in getting a microprocessor to work properly, we asked to see the test circuit used to arrive at the applications note and the specifications for the part. We were told there wasn’t a test circuit and never had been one. When we asked, “how do you find out if the part works correctly?” The response was, “we give them to our customers and they tell us if they work!”
At Speeding Edge, our experience has been that the use of ferrite beads has been the result of a knee- jerk reaction, a band-aid or a case of holding onto bad practices rather than doing good engineering. As Lee Ritchey, President of Speeding Edge notes, “In the 40+ years of designing high-speed computer systems and networking products, I have never used a ferrite bead in the power lead of a device whether it is a PLL or an analog circuit—all of which have functioned to their specifications and passed appropriate EMI and ESD tests. Instead, I have determined what the ‘ripple’ requirements of a circuit are and designed the power delivery system to meet those requirements.”
Ritchey, Lee W. and Zasio, John J., “Right The First Time, A Practical Handbook on High-Speed PCB and System Design, Volumes 1 and 2.”
Would you like to find out more about how Altium can help you with your next PCB design? Talk to Altium
About the AuthorMore Content by Kella Knack