Mechanical design of the HP 4980 Network Advisor - protocol analyzer - Technical

Hewlett-Packard Journal, Oct, 1992 by Kenneth R. Krebs

As mentioned previously, the display housing and display housing rear-cover molds were designed with two sets of inserts to accommodate the color or monochrome LCD display. The color LCD has the larger active area opening and requires different mounting boss height and location.

The handle-mounting piece and the front-panel rear cover piece are about the same size and weight. As a result, they were put into a family mold to save on both mold cost and part cost. This later had the disadvantage of changing two parts when only one, the handle mount, needed its thickness reduced. Additional work was then needed on the frontpanel rear cover to compensate for the reduction in the mold separation line.

During the lab prototyping phase, we soft-tooled the entire box using urethane molds for parts over 40 square inches and aluminum tools for the smaller parts (a process limitation). The vendor worked directly from IGES** translations of our 2D, undimensioned HP ME 10 drawings. The parts had molded-in color and texture but were soft and capable only of limited structural and temperature testing. The smaller parts, molded in aluminum tools, were molded in their proper materials, which allowed thorough structural testing of these parts. While this process produced highquality parts that were very useful in our evaluation, the process was costly and took much longer than the vendor estimated.

With an all-plastic enclosure, we knew that shielding technology would become a critical factor. From the outset, we planned on vacuum-deposited aluminum for this job. This technology was available in very few places. Also, the size of some of our parts (requiring up to 600-ton presses) limited the molder selection greatly. However, one molder had up to 1000-ton presses and the abihty to do vacuum deposition in-house. This vendor selection later proved to be unfortunate for two major reasons. First, they relied heavily on cutting cavities in the parent steel. Many of our parts have deep ribs and intricate features which, later experience would show, would have been more easily polished and otherwise tuned had they been formed using mold inserts. While more costly up front, inserted molds would have saved us time and money during the tool tryout and tuning phase of our product development. In some cases, attempts at changing the parent steel, especially polishing of deep ribs, caused mold damage that could not be fixed, leaving some parts permanently out of specification. In these cases, unfortunate and costly design changes were necessary.

The second reason was that we found out very late in the tool tryout phase that the molder could no longer supply us with molded parts. During the ensuing turmoil we initiated our contingency plan of moving the molds to PAFC, which had recently installed a new 550-ton press capable of molding our larger parts. This transition would have been significantly more painful if our matedais department had not planned for it.

Environmental Testing

The 2-to-3-micro-m-thick vacuum-deposited aluminum coating proved to be an ineffective shielding technology with which we experienced flaking problems, especially during humidity testing. Our search for an alternative included nickelbased and copper-based paints, electroless copper plating, and zinc arc spraying (ZAS). After tests and evaluations of the controlability of the processes, we opted for ZAS, a capability that PAFC had in-house. While this has proven to be more costly than vacuum-deposited aluminum, it has allowed us to meet our targeted CISPR 11' radiated emissions specification. Initially, because the silicon ZAS masking fixtures are soft and flexible for easy cleaning, the force of the spray would sometimes blow the mask away from the part allowing zinc overspray to get on the cosmetic surfaces. Reinforcing the masks brought the process under control.