previous | contents | next

manual switch as a variant of the display as indicated in Solution 4. What is the maximum allowable event rate?

3. Define the action of Ksub(Process-cell-overflow) for Figure Hist-6 and draw the flowchart of its behavior.

4. Carry out the design indicated in Solution 1 to increase the update rate (i.e., decrease the time between events). This is accomplished by polling the update- request process within the Kmacro(Display). What is the maximum allowable event rate?

5. Design the control parts for the two-Bus scheme as indicated in Solution 3.

6. Design a two-Bus histogram recorder using a K(arbiter). Minimize the time spent in the common control section. What is the maximum allowable event rate?

7. Plot the cost and performance for the various solutions.

8. Compute and display the integral of the distribution. Form the cumulative probability distribution: FF(X[J]) = sum, 1= 0 to J, of F(X[I]) for all J = 0,1,....,N such that FF(X[N]) = 1. This process would be initiated by pressing a K(manual evoke).

9. Compute the mean of the distribution described in problem 8.

10. Compute the median of the distribution described in problem 8.

11. Using the T(paper tape) described in the transducer section, punch the contents of the 16 bit switch register, followed by the 2*1024 8-bit characters in the histogram memory. Punch a 16-bit sum check of all this preceding data;

KEYWORDS: Monitor, control, waveform analysis, multiplexing

This problem involves the design of a system for the analysis of multiple, but similar waveforms. The process from which the waveforms arise is a continuous, multiple station coating mill (see Figure CTM-1).(1) The coatings are on sheet stock such as paper, cloth, plastic, and tin which move through the mill. Since this is a continuous process it is difficult (too costly and impractical) to test the thickness of the coating at completion. However, controlling the coating thickness is extremely important because material will be wasted if too much is applied, and similarly, the quality of the product may be unacceptable if too little is applied. We will discuss and propose the design problem for this task, but will not present detailed solutions.

A special purpose digital system, which we shall call a coating thickness monitor (CTM), can permit the nondestructive testing of the coating thickness. In this method, coating thicknesses are measured by beta gages that provide an electrical (analog voltage) output related to the mass of material in the gage measuring gap. Gages are placed before and after each coating station as shown in Figure CTM-1.

A PMS diagram of a possible general structure for the CTM is given in Figure CTM-2. The various beta gage voltages are selected via an input S(multiplexor). The results of analyzing the beta gage waveforms are displayed at the various T(display)'s via an output S(multiplexor). In general, for an n station coating mill, there would be 2n input beta gages and n T(display)'s. Thus the main emphasis of this design is to show how a single system can appear to be n, independent systems by time-sharing a single CTM.

----------------------

1. This problem was derived from Jurgen (1970).

219

previous | contents | next