Initial work on the project was to construct an idealised logic, based on the logic diagrams given in the Report which are incomplete. This was done with the aid of an Editor/Simulator named ELSIE which was written explicitly for this purpose.
Using the simulation facilities in ELSIE we have been able to complete this logic and the replica is largely based on that. The logic diagrams are not set in stone and are reviewed as the rebuild progresses and we reach a better understanding of how things were. Often photographs will provide information as to the number and types of valves on individual panels, and when this conflicts with the logic we have resolved the conflict by re-structuring parts of the logic.
At one point in the development of the machine a number of drawings, produced towards the end of the life of EDSAC, came to light. These have proved to be encoraging as they have confirmed a number of our deductions. They also provided details of component values in places where we had been unable to determine these.
As the simulator is intolerant of timing errors, the logic contains a number of delays which are probably not needed in the real machine, as stray, and deliberately added, capacitance will smooth out many of the glitches that the simulator finds to be a problem.
Note: In my parlance, modules are subdivisions of the logic, roughly based on logical units. In some cases they correspond to individual panels, in others they may consist of several panels.
The logic modules can be viewed here.
Page created by Bill Purvis, last updated 15th December, 2015