Site: TDK Corporation
13-1, Nihonbashi, 1-chome, Chuo-ku
Tokyo 103, Japan

Date Visited: October 7, 1993

Report Author: G. Meieran

ATTENDEES

JTEC:

W. Boulton
G. Harman
G. Meieran
R. Shelton

HOSTS:

Fumio Uchikoba

Kenichi Aoshima

Noboru Yamaki

Akira Okamoto

Mitsunobu Goto

Akira Fukuno

Tadashi Mitsui

Kouichi Yajima

Takashi Horiuchi

Tatsushiro Ochiai

BACKGROUND

The visit to TDK was bimodal; a very interesting and useful morning in which we discussed the TDK packaging and assembly issues, and an afternoon plant tour of the ferrite manufacturing facility in Narita, which was interesting but not particularly pertinent to the packaging and assembly issues. Hence, this site report concentrates only on the morning session, and only mentions two aspects of the afternoon tour that had some pertinence to packaging and assembly. Most of the information discussed below was presented to the visiting JTEC team in the form of two interesting and extremely well-done video tapes; we did not personally see any of the processes or equipment described in these tapes.

TDK appears to be a very well run company with excellent facilities, top quality people, and excellent technology. It was formed in 1935 by the invention (so we are told) of ferrites. It appears to be a world leader and the world's largest supplier of ferrites, ferroelectrics, and other electronic materials. Its products are used everywhere, such as in floppy and hard disks. As a result of this path, TDK has become very adept at manufacturing passive and active electronic components, and in their assembly into high-quality PC boards. Indeed its manufacturing capabilities in the component area are no less than phenomenal; TDK produces very small parts in huge quantities, using state-of-the-art equipment and processing capabilities. Because there is a need to have equipment available to assemble these small components, TDK makes much of its own equipment, and, as other Japanese companies showed us, it is prepared to sell its assembly equipment to the external market (except for its latest state-of-the-art manufacturing equipment).

The quality of TDK products appears to be very high; design of products and processes is thorough and well done. Video tapes describing all this were among the best I have seen and were convincing. Even at the fairly old ferrite plant, most of the processes are automated; it appears as if yield is good and quality levels are high. However, we had the impression from the ferrite plant visit that TDK is not addressing other manufacturing issues; there was lots of inspection, lots of inventory, and little evidence of motivation to do anything differently.

Since much of TDK's product line consists of old products with very little profit margin, it is moving some of its manufacturing overseas, to China and Malaysia, and has challenged its domestic staff to come up with new, high-value-added products. It will use its current products, but will look to incorporate these into high-value products. In newer lines, such as seen on the video tape, there is little evidence of inspection, and automation appears to be 100%. It was impossible to see what inventory levels were present. The TDK philosophy seems to be, "Good enough is good enough." If something requires automation or special manufacturing processes, fine, the company will develop and provide technology and automation as necessary; otherwise, incremental improvement is satisfactory.

TDK has about 25,000 employees total, about 7,000 of whom are domestic. About 60% of its manufacturing capability was in Japan at the time of the JTEC visit. Its product line includes ferrites, semiconductors, ceramic components, recording media, metallic products, and a variety of other mostly electronic products.

All that we saw was consistent with TDK's stated company philosophy, "Respecting Originality," with its company motto, "Contribute to Culture and Industry Through Creativity," and with its employee attributes, "Vision, Courage, Reliance."

COMPONENT DESIGN AND MANUFACTURING

TDK designs and fabricates components such as small transformers, inductors, and capacitors in huge (hundreds of millions of individual pieces) quantities. It also makes hybrids, sometimes incorporating passive elements in an active circuit. Consequently, it manufactures a very broad line of products. Indeed, its product line numbers in the range of 140,000, although it is trying to reduce this to a more manageable level (30,000?). TDK clearly appears to be competitive with the state of the art, or is setting the standards for state of the art, in the fields in which it competes.

The sizes of the small parts TDK manufactures is staggering, even to those of us in the semiconductor industry, in which "small" refers to line geometry, but the chips may be fairly large. Its smallest capacitors are, for example, smaller than a grain of pepper (one sneeze and 400,000 parts disappear!). Even these tiny objects are made up of many layers of alternating dielectric and metalization, cleverly laid down in sequential sides of a chip to give a ramped staircase effect, in the case of physically small inductors.

The emphasis is on cost and size reduction. TDK seems less interested in integration; indeed it seems to be the opposite of a company such as Intel, which makes individual chips that integrate many functions and grow larger as a result of integration; TDK wants to make things as individual components and as small and inexpensive as possible. As a result, it has developed a line of assembly equipment that can handle its small and exceedingly wide line of products. This line of equipment is extremely sophisticated, very precise, and totally automates assembly of PC boards; operators can barely see, much less manually handle, TDK's line of small components. The company's equipment appears to be a decade ahead of anything comparable in the United States.

Customer Focus

TDK might be expected to be a leader in high-temperature superconductors, resulting from its extensive expertise in ferroelectric device technology. But a technical as well as business perspective arose when superconductors were discussed during the JTEC visit. On the one hand, TDK spent several years and lots of money looking at high-temperature superconductors; however, there were several technical problems that suggested that it need not bother, and work on these materials has ceased. Even more important, our hosts stated that even if all the technical problems were solved (materials synthesis, defects, hysterisis, whatever), the devices would still need to be operated below room temperature, so TDK would likely not get involved. The customer base would not be expected to want to use cooled packages, and since TDK makes products for its customers, it is not likely to get reinvolved. In this sense, TDK looks at technology to help develop markets, but, clearly the market rules, not the technology.

TDK's ceramic substrates are quite thin, say about 5 microns, and it has solved the cracking problems normally associated with such thin brittle layers. It then builds up 70 or 100 layers, making these small, 1 pf capacitors. It has solved cracking problems by carefully selecting metalization materials (e.g., nickel), closely matching the thermal expansion coefficients of the materials, and firing at relatively low temperatures.

Assembly Automation Equipment

Initially, equipment was developed in-house to handle only TDK products, under the assumption that the company could get a better lock on the market this way. People who bought their equipment to assemble some TDK products would find that they could only use it on TDK product lines. This strategy back-fired. It was too focused on TDK product lines, and the competition from companies such as Matsushita, which offered a more "open architecture," caused management to change this strategy. Their present line of equipment, made by a company called Avimount, accounts for 7% of TDK's business.

Cost Basis

TDK has calculated that labor costs in Japan are on the order of two seconds per yen. Since a multilayer capacitor costs about one yen, it is impossible to inspect products. Hence, TDK has adopted the philosophy of zero defects. Indeed, current figures show a defect rate on the level of 1 to 10 PPB. On assembled products, there is a defect level of 10 to 100 PPM. Clearly, TDK drives all non-value-added costs out of the manufacturing process; U.S. companies might invest in inspection and testing, while TDK invests in process improvement.

TDK is ISO certified, but only because of U.S. customers. There appears to be no great drive in Japan for standards as there are in the United States.

Technology

Our hosts showed us an interesting expert system, of which they seemed quite proud. This is a system that designs products by having a customer choose magnetic properties, and the expert system provides the details of how to make the product, e.g., composition, firing conditions, etc. Or the customer can choose a composition for the three-dimensional phase diagram, and the magnetic properties and appropriate processing conditions pop out. This is a neat piece of software, with some excellent graphics, and it appears to be in use.

TDK has an interesting way of making rare-earth iron magnets. It fires (sinters) magnetic bodies made from fine powders with intentionally lower oxygen content in order to get rid of nonmagnetic rare earth oxide impurities. It has introduced an hydrogen absorption and desorption process to ingots so that the ingots become brittle. Since brittle ingots need less impact for fine pulverization, they can prepare fine powder by applying lower oxygen content - a neat way to use a failure mechanism to help solve a problem and simplify a process.

The JTEC team got some idea of the persistence of Japanese companies in pursuing worthwhile ideas and technology, and the inverse, the rapid discarding of even good ideas that do not fit their marketing molds. At Narita, we saw a (new then) green tape line for metalized ceramics that TDK had been pursuing for seven years and was only then being readied for mass production. This line is for glass-ceramic composites, with a low ( about 900 degrees C) firing temperature. The material is low-cost and strong, but does not have the thermal conductance of pure ceramics such as alumina. But TDK sees a market and is going to start proliferating the technology in the near future.

This is to be contrasted with the rapid elimination of the high-Tc superconductor program, which simply did not meet TDK's customer profile, even though it seems to have totally appropriate technology and expertise, and the technology seems to have promise for some markets.

Indeed, this seems to be an area of some contention; how does a company like TDK take advantage of new opportunities? TDK is excellent in technology, but seems to be less strong in developing new markets that may be a bit out of the path of its existing markets.

SUMMARY

The impressive features about TDK are its technology for mass producing a large variety of very small, high-quality products. As a consequence, it has developed a very strong customer consciousness and a wealth of supporting tools (such as assembly equipment for mounting these small parts). It has a strong company focus on producing quality products that require minimal inspection, on satisfying the customer with the range of products it produces, on maintaining a product line compatible with its technological expertise, and on persisting until the technology problems associated with its product lines are resolved.

Any new products must meet the above criteria. As a result, TDK is a very strong force in Japanese electronics manufacturing and has an impressive track record ranging from innovation and invention (of ferrites) to implementation and marketing. Its electronic products line is broad as well as deep, and is state-of-the-art. Whether or not TDK will be as strong a force in marketing assembly equipment remains to be seen, but there is little doubt it is capable of strongly entering this market.



Published: February 1995; WTEC Hyper-Librarian