The CRT Screen Line placed the phosphor screen on the inside of the CRT and involved about 10 major manufacturing processes. When electrons strike the phosphor screen the "green" visible trace that is typical of CRT's is produced.
The original concept was to duplicate the existing manufacturing line and then to replace the operators with a series of robots. An alternative architecture was presented that integrated a clean room conveyor system with varying pieces of process equipment. The alternative was accepted with a significant cost savings. The final result was approximately 52 machines of various sizes and complexities.
To present a variety of different bulb shapes and sizes so that there was a consistent interface with the equipment, a universal carrier was developed. The carrier, (shown on the right below) was relatively inexpensive but was required to repeatedly survive 450 Deg C process temperatures:
When Potassium Silicate was dispensed into a "cushion" of Barium Acetate the result is a precipitate that forms the phosphorus screen on the inside of the CRT. The mixtures needed to be precisely controlled. Contamination was a primary concern. In previous designs, long lengths of distribution tubing was suspected to attribute to contamination, so an important aspect of the configuration was to keep the distribution system as short as possible and to move the bulbs by the filling station. The chemical mixing and dispensing equipment is shown on the Right.
The Chemical mixture was required to "settle" for 30 minutes. At the end of this time the remaining solution was carefully decanted as a thin phosphor layer remained attached to the inside of the face plate. After decanting, the phosphor layer was dried. The dispensing, settling, decanting and drying operations were all built into a single machine. The air dry nozzles, (shown at the bottom-center of the picture) tracked along with the constant belt speed, (shown in the lower center of the picture) via a servo system:
The next step was to visually inspect every screen. A rotary inspection table, (shown below) presented the bulb to the inspector. If the bulb was accepted it would continue on to the next process, (to the right). If a defect was detected the bulb would be routed to an automatic reclaim station and then back to the input of this production line, (towards the top of the picture). This approach was adopted because the final output quality of the process was less than the overall quality requirement of the production line. The automatic rework-rerouting required very little additional labor cost while meeting the overall yield requirements.
A microscopic cross section view of the phosphor layer would appear to look like a series of mountains and valleys. Ultimately a thin aluminum layer will be deposited that is supported only by the tops of the "mountain" peaks. To accomplish the aluminum "sheet", a lacquer layer is deposited to create a network of "bridges" over the "valleys". The first step to create the "bridges" was to fill the "valleys" with water. A very thin coating of the lacquer was then "floated" on the bottom surface of the water. The amount and distribution of the water and lacquer was controlled by spinning the bulb at the correct speed and time. After the lacquer is sprayed and spun, a set of drying nozzles entered the tubes to "set" the lacquer. Immediately exiting the spray station, the parts enter into the lacquer dryer, (shown on the right of the picture below) where the water was evaporated out from under the thin layer of lacquer.
Once the water was removed from under the lacquer, aluminum was vapor deposited onto the lacquer surface. The Aluminum Vapor Deposition station is shown below. Along the back of this station is a "fixed" linear robot. This robot received unprocessed parts from the left, loaded it into the available "aluminizer" and then forwarded the aluminized bulb onto the Lacquer Bake Out oven, (shown towards the back of the picture).
The load station over the top of each of the aluminizers could hold two bulbs. One position for the bulb in process and a second position for handing off and receiving the next bulb. This approached optimized the production time.
Another major feature was the number of aluminizers. The process flow rate only required three units. However, previous aluminizers were relatively undependable and required minor maintenance about every 20 "shots". To avoid interruptions in the production line an additional aluminizer was added. This allowed for any three units to be in operation while a 4th station could be removed, maintained and replaced.
The aluminizers themselves represent advanced applications of technology. Each aluminizer incorporated a complete high vacuum system including a roughing pump, a "Cryo" pump and all the necessary instrumentation. The engineer that designed these units earned a patent for a very unique heating process which used an electron beam and a graphite crucible. While the classical use of tungsten filaments required frequent maintenance, the graphite crucibles processed thousands of parts before needing to be replaced. The E-Beam approach has since become an industry standard.
Once the aluminum layer was deposited, the lacquer was "burnt" out from under the aluminum. This processed left a very thin layer of aluminum supported only the "peaks" of the grains of phosphor screen.
The advantage of the aluminum layer was that since it had close to a mirror finish a much larger percentage of light was reflected forward when electrons struck the phosphor layer. Not only was the light generation efficiency improved but the aluminum layer also dissipated heat which helped prevent burning spots in screen.
Similar to the Frit oven, the technique to closely control the temperature profile in the oven was to implement a number of heat control zones. This oven had 10 zones, which allowed for a tight profile to be obtained in less than 30 minutes upon startup. Each zone was comprised of a temperature controller, (shown below) a heater and a re-circulation fan:
Since the overall process time through the oven was approximately 3.5 hours, an equipment failure could present a potential loss of product and production time. To address this issue, the oven was specified to have an individual electrical disconnect switch for each fan and heater, (shown below). This allowed safe repairs to be made while the oven remained in production:
Because the lacquer oven process took 3.5 hours, considerable logistic issues could occur at the beginning and end of production shifts. The logistic challenges were solved by incorporating a lacquer Oven Storage module. This module stored slightly more bulbs than the oven could hold. This allowed the oven to continue to operate after the end of the shift. Instead of the finished modules being sent onto the next production line, they would automatically be shifted into the storage module.
Approximately 3.5 hours after the shift ended the storage module would be full, the oven would be empty and everything would shut down. The next morning, the oven would pre-start approximately 30 minutes before the shift, but the oven would be empty.
During the first 3.5 hours of the next shift, completed parts would be withdrawn from the Oven Storage module and sent onto the next process without interruption in the flow of parts. About the time the Oven Storage module was empty, parts would begin arriving from the Oven. The end view of the storage module, (empty) is shown below:
The overall line included approximately 750 PLC inputs and outputs, motors, sensors, solenoid valves etc. Three large electrical control cabinets were distributed throughout the production line. 4x4 electrical gutters were also installed throughout the line to allow for easy installation of the electrical wiring to the field components, (shown along the top of the below picture).
The electrical gutter approach greatly improved the initial installation as well as accommodating any changes or additions after the production line was installed.
An addition item of interest may be the air exhaust system, (shown by the grey PVC pipe in about the middle of the picture). All compressed air was exhausted into this exhaust system. Models of direct acting valves were selected to help preserve the air quality of the clean room.
More than 30 operators were involved with the original manual production line and the yields averaged about 60%. The completion of the automated line reduced the number of operators to only 4 with greater than 98% yields.
Once the general architecture of the production line was defined, a set of specifications were written for each machine module. This greatly assisted in the coordination of a team of between 3 to 6 engineers and designers who worked together approximately 18 months until the line was commissioned.
The cost of engineering was approximately 15% of this $1.5M project, (early 1980 dollars). Coordination included working with numerous inside and outside shops and resources.