The first step the team took in restoring the PDP-1 was taking an inventory of the system. The CPU had three 4K banks of core memory (for a total of 12KB!), a cosmetically-intact and largely-functional front panel and several peripheral devices. Work began with the CPU.
From previous restoration experience, it was known that capacitors in electronic power supplies can deteriorate with age. The first task was therefore to test and "reform" these components and then test the power supplies themselves while disconnected from the system. Once certain that the power supplies were functioning correctly, they were re-connected to the CPU and power was applied.
In parallel with power supply testing, the team replaced burned-out lamps and switches in the front panel—the only means of communicating with the machine when first turned on. The three large cooling fans in the bottom of the CPU were also re-built and tested.
Power was applied. Since core memory can retain its contents even when power is removed, the team archived the contents of memory for later analysis. They used the front panel to read in the first few words into memory and then entered in a simple program using the front panel switches to dump the rest of the memory’s contents to a custom interface built to allow this data to be relayed to and stored on a personal computer. This data has not been analyzed yet but could be rare historical code.
At this point, the team began running diagnostic tests (with modern-day logic analyzers and oscilloscopes), fixing any errors they found. Typically, the simplest and most effective repair was to replace one of the CPU’s circuit cards (known as a "System Building Blocks"). When spares were not available, the module in question was analyzed in depth and repaired. At times, this was made more challenging as some of the components were not longer made.
With the CPU functioning at a basic level, work then focused on restoring the peripherals required to make a full system.