Noise Suppression Filter Guide

The secret story behind the birth of chip ferrite beads (part 1/2)

14/12/2012

["What is ferrite?"]

It was May 1985. I was riding in the Raicho limited express train from Fukui bound for Kyoto on a business trip to the Head Office (Murata Manufacturing Co., Ltd.). That was my first business trip since joining the company a month earlier. On the train, the General Manager of our department explained to me the purpose of the business trip: "It seems that we will be able to make ferrite surface-mount components, so we are going to hear about them." I remember asking what is now quite an embarrassing question: "What is ferrite?" The General Manager replied, "Rusted iron," which I accepted with a simple, "I see."
That may have been my first encounter with ferrite.

[What is multilayer multilayer ferrite with a low Q used for?]

Development of multilayer multilayer ferrite started by aiming for a Q value equivalent to that of winding types, however ferrite with high-Q characteristics could not be created easily, so we were left searching for something that we could use. Since low-Q ferrite is advantageous as a noise filter, it was decided to investigate commercialization as noise filters.
Development and prototyp manufacturing were performed at the then Murata Manufacturing Head Office (currently the Nagaoka Plant), and product development, the was performed at Fukui Murata Manufacturing.

[Product development by trial and error]

At the time, the EMI Group was an organization comprising less than 20 people that had just split off as an independent section from the MLCC Division. I had been assigned there as a new employee, and was engaged in product development. This was the first chip component product to be developed by the EMI Group, and at the time, Murata Manufacturing had only two types of surface-mount components: multilayer multilayer capacitors (GR series) and chip coils (LQH series).

There were so many things that I didn't know: "What characteristics are required?" "What should the outer dimensions be?" "What about the internal structure?" "How should reliability be evaluated?" "How should surface mounting be evaluated?" "What needs to be investigated for surface mounting in the first place?"
The situation was such that I could not be concerned with how it looked to be asking even such elementary questions as "What is a noise filter?" I remember going around fearlessly asking questions of various departments and people. Looking back, I am keenly aware that it was "newbie" privilege.

In terms of characteristics, we aimed for characteristics equivalent to those of ferrite beads with lead wires in the BL01/02 series which were commercialized at that time. We wanted a size that would be easy to use mounted in the serial interface lines of the target applications. As a result, it was decided to use 4516-size outer dimensions, which are long with a narrow width.

In terms of design, multilayer multilayer ferrite at that time had a higher firing temperature than current products, and when 100% Ag internal electrodes were tested, the Ag dispersed or vaporized during firing, resulting in frequent open defects. Therefore, Ag/Pd prototypes were created in an attempt to suppress Ag dispersion and vaporization. We wanted to go with 100% Pd to stabilize conduction, but with the addition of some structural countermeasures we were able to achieve prospects for mass production of Ag/Pd products to reduce costs. There were some concerns about the reliability of Ni/Sn plating for the external electrodes, so the commercialized products used a thick Ag/Pd film. However, this later led to a complaint.

The evaluation samples were finished around this time, but the issue of "what to do about the characteristics guarantee?" remained. At the time, only the outer dimensions of ferrite beads with lead wires were guaranteed, so this was of no use as a reference. The developed products were noise filters, so it was quickly decided that impedance should be guaranteed, but the question was at what frequency? There were no official standards, and also no preceding manufacturers, so there was nothing we could use as a reference. Ultimately, we decided on our own to guarantee the impedance (Z) and DC resistance (RDC) at 100 MHz, which is within the 30 to 300 MHz band for which noise was an issue at the time, and commenced various reliability evaluations.

Design of the embossed tape also finished around the end of commercialization, and a mounter was borrowed from Komatsu Murata Manufacturing to perform the mounting evaluation. However, it was found that the mechanical centering method^* used as the positioning method at the time had problems with the narrow 4516-size dimensions. This issue was cleared with the cooperation of various parties, and the products were provided to customers together with the mounting method.

In this manner, product development also finished, in this case merely meaning the end of prototype manufacturing on a prototype line at the Head Office, with the cooperation of many different people, and the first BLM41/31 series products underwent in-house design examination in autumn of 1986.

^* Mechanical centering method: A positioning method that mechanically restricts the component center based on the package outline.