Ferrite Bead Usage: How Much Is Too Much?

  • Thread starter Thread starter likephysics
  • Start date Start date
  • Tags Tags
    Bead
Click For Summary
SUMMARY

The discussion centers on the usage of ferrite beads in ARM9-based embedded system designs. While some designers advocate for multiple ferrites across various components like NAND and DDR, the consensus is that excessive use can complicate PCB layout and create split reference planes, negatively impacting high-speed signal integrity. Designers often include more filtering elements than necessary due to EMI concerns and established practices, leading to potentially inefficient designs. A recommended approach is to start with a conservative design and gradually remove ferrites, rather than adding them post-design.

PREREQUISITES
  • Understanding of ARM9 architecture
  • Knowledge of PCB layout principles
  • Familiarity with EMI mitigation techniques
  • Experience with ferrite bead applications in circuit design
NEXT STEPS
  • Research best practices for ferrite bead placement in embedded systems
  • Learn about high-speed signal integrity and its relationship with PCB layout
  • Explore EMI testing methods for embedded designs
  • Investigate alternative filtering solutions to ferrite beads
USEFUL FOR

Embedded system designers, PCB layout engineers, and anyone involved in optimizing EMI performance in electronic circuits.

likephysics
Messages
638
Reaction score
4
How much is too much, when it comes to using ferrites.
Some of the boards I designed previously did not even have single ferrite. We pretty much never used ferrites. Compliance was not a problem.
Now I design ARM9 based embedded system boards. There are ferrites for everything - NAND, DDR, CPU, CPU core. Are so many ferrites really needed?
worst of all, there is no justification. They continue to use it just because its been used previously or it was used in the reference schematic.
One ferrite after the power supply is great. But for every rail?
It basically creates split planes. So no single reference plane for high speed signals. Makes layout hard, specially when DDR2, NAND are present.
 
Engineering news on Phys.org
EMI is difficult and very much experimental, so what one designer does results primarily from his habits.

Because EMI worries take so long to solve, and a ferrite bead or decoupling capacitor or voltage regulator is cheap, experimented designers use to put more than strictly needed, in accordance to hard-learned systematic practices, to save stress and time - and leave them in place when the design works, instead of trying each and every one.

As a result, typical designs have more filtering elements than needed. Possibly many more.

If you wish, you could try to suppress some and take the very long time to check if the circuit still works perfectly. Remember that a digital circuit's error-free operation is very long to check - at every supply drift and temperature and timing skew.

But don't first design a simplified PCB, run into trouble, and try to improve it later. Take the paranoid design first and remove gradually some elements instead. If not, you'll fail.

And you knwo what? I suggest that you evaluate how much the beads cost and how much time you'd waste (months and several runs of design+cable+test) if it runs wrongly.
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 8 ·
Replies
8
Views
923
  • · Replies 1 ·
Replies
1
Views
3K
Replies
14
Views
5K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 11 ·
Replies
11
Views
4K
Replies
1
Views
5K
Replies
22
Views
24K
  • · Replies 18 ·
Replies
18
Views
5K