Datacon's Multi-Chip Die Bonder 2200 apm

Application	Keywords	Products	Details
Datacon's Multi-Chip Die Bonder 2200 apm	semiconductor, flip-chip bonding, automation technology	BM 15 A BM 20 A MN 12

 In 1999, the increase was 75%, and for the next few years — depending on the development of the generally volatile semiconductor business — a growth rate of 20% to 40% is anticipated. Because of the excellent development of its business, this specialist in multi-chip and flip-chip bonders now employs more than 310 persons. And only 1 year after moving into newly built company premises, in Radfeld/Tyrol, Austria, it has already expanded its office and production space to 9,600 sqm. (Further information concerning the historical development of the company can be found in Marketing News No. 51, September 1999.) The company premises, which were initially occupied in March 1999, were expanded by 3,600 sqm one year later. During the spring of 2000, Datacom’s European Technology Center, located next to the main company building, was inaugurated. A center for comprehensive packaging solutions, it was established together with partner company Kulicke & Soffa. Apart from the Datacon technologies for die bonding, flip-chip bonding and automation technology, K&S solutions for wedge bonding, ball bonding, manual bonding and wafer cutting are available for practical application trials. For this reason, Datacon also refers to the center as a further step on the way to “total packaging solutions.”

Multi-Chip Die Bonder 2200 apm
The new generation multi-chip die bonder, the 2200 apm series, was designed to meet the demands of the advanced packaging market. It replaced the PPS 2200 series, which was described in the previous Marketing News. The goal of the new machine was primarily the enhancement of positioning accuracy from 25 μm @ 3 sigma to 10 μm @ 3 sigma. This quantum leap in accuracy also placed new demands on the mechanical heart of the machine, the threeaxis positioning system, which is
installed suspended and which has the task of positioning the bond head both dynamically and with the highest precision.
The new 2200 apm series for multi-chip and flip-chip applications with an enhanced positioning accuracy of 10 μm @ 3 sigma for the advanced packaging market. Because of these demanding requirements, Datacon was compelled to start a project with an outside partner for the development of a suitable three-axis system. During the spring of 1999, the specification was submitted to various suppliers, with a very tight time budget. (The prototype was scheduled to be operational in December 1999!) Following the first project presentation, the SCHNEEBERGER team decided to name it “The Magic Flute,” after the opera by W. A. Mozart — an appropriate designation for this demanding project and also a suitable reference to the customer’s Austrian location!

The Engineering Process as a Success Factor
Engineering costs and the expense of prototype development are frequently not taken into consideration by most system suppliers. The profits from series production are supposed to generate a payback within 12 to 18 months. In this case, the anticipated high expenses for development and engineering activities, the tight time schedule and the competitive situation left a few questions unanswered:

• Who will pay for our expenses, if we do not obtain the order for series production?
• Can we maintain the time schedule with the resources at our disposal? Will we have to rely on outside resources?
• And will the customer honor our efforts and expenses to the degree necessary for us to be successful versus the competition?

Because Datacon decided to have prototypes built by several suppliers, SCHNEEBERGER faced a significant competitive situation. Thus a solution had to be found to take into account the development costs. Early in the project, it was also clear that we did not have the necessary resources to develop the prototype, with the demanding performance criteria, in such a short time period. Therefore the decision was made to commission an engineering partner during the development phase. This made it even more important to provide a clear separation of expenses between the prototype and the series production phase. During the summer of 1999, we submitted a quotation for development, manufacture, tuning and testing of three prototypes to Datacon. In order to increase their chance of success, the competition decided to charge the same price for prototypes as for production units. Nonetheless, in October 1999, Datacon awarded us the order for the development, manufacture and installation of three prototypes. Apart from the supply, tuning and testing of the mechanical three-axis carriage with linear motors and optical distance measuring systems on all axes, the system also included the complete motor and encoder cabling; the cabling and pneumatic lines and hoses for all systems, which later might be located on the bond head; as well as an electric and pneumatic control unit, located on the Z-axis. The initial tests on prototype #1 were completed in the middle of December, and they not only verified our system concept, but also identified a few weak points. A reworked and improved prototype #2 exhibited the required performance of this three-axis system, especially its good thermal stability.

The Much Sought After Objective — the Order for Series Production
At the end of April 2000, Datacon released the production order for the first 100 series machines, with an obligation to purchase a minimum of 300 systems per annum. Technologically we were able to convince Datacon with the results of the first bonding tests, which were carried out with glass chips. Organizationally we convinced them of our competence in manufacturing production quantities of a demanding project like the “Magic Flute” with consistent high quality. However, we only had 12 weeks’ time left to deliver the two first systems from the pilot production series. Simultaneously, the series production logistics and the assembly line and
personnel required for it had to be built up and readied such that up to 8 systems per week could be delivered in less than 20 weeks. With respect to the purchasing logistics, the concern was to build up the sub-suppliers in accordance with our
ramp-up in the shortest possible time. There was hardly any
leeway remaining for initial difficulties and problems.

Above all, the deliveries to the customer had to be implemented as smoothly as possible. The requirement was that, based on a long-term forecast, a defined number of systems had to arrive at an exactly predefined day. To achieve this, the following solution was implemented. On a predefined weekday, and in accordance with the forecast, the corresponding number of systems are loaded and delivered directly to the customer the same day in a single shipment. There are no delays, because the goods reach the customer without any transfers. Even if the assembly operation during peak demand periods cannot supply the required quantities, a small number of finished systems are maintained in stock, ensuring a smooth flow of product. Success Factors and Customer Benefit The decision to separate the development and engineering costs for the prototypes from the series production and also to invoice these separately proved to be the right one. From the experience gained, it is important to ensure that both the customer and the supplier benefit from this engineering
process as follows:

• The development and prototype team can fully concentrate on the development of the prototypes in accordance with the time schedule and with the required performance characteristics. They are not distracted by the fact that other customer projects might be more promising and therefore are not subject to the risk of changing priorities. In addition, the motivational aspect for the important persons involved must not be underestimated.
• If decisive competences and/or resources for a successful engineering and acquisition are lacking, then these can be brought in from outside (providing this has been previously included in the budget calculations). Valuable know-how is transferred to the company and the development time is shortened.
• As a result of the financial ties to the supplier, the customer also finds it easier to support and participate in the development during a substantial effort. The customer/supplier relationship turns into a genuine partnership.
• Finally, the successful development of the prototypes, aswell as the closer customer interaction, increases the probability of success for the contract for series production.

After the series production deliveries had commenced, at the end of October 2000 we had already received the follow-on order for 200 systems, and the forecast for the annual quantity required was increased to 360 pieces. New projects are already in process. Datacon is also addressing the requirement of the semiconductor market for the larger 300 mm wafers. This is manifest in the form of the so-called apm+, a multi-chip die bonder based on the 2200 apm series, suitable for the processing of 300 mm wafers. We are currently building three new prototypes of a three-axis system, which essentially is based on the “Magic Flute” butwith an extended x-axis stroke. As a leading supplier of flip-chip bonding equipment, Datacon wants to continue to expand its technological leadership and for this reason is working on the development of a completely novel flip-chip bonder. For this project, SCHNEEBERGER also supplied a prototype XY table with linear motor drive, and an optical distance measuring system, including cabling. Thus our collaboration with Datacon continues to be exciting!