Erik is not the only one. Check out Tatiana van Vark. Here's a picture of what is in her DINING ROOM, the complete electronics suite from a Vulcan bomber!
Yes, Tatiana is the queen of collecting this kind of stuff and she has an excellent page!
There's also a video of her picking up a Litton inertial measuring unit and moving it around, there's 3 racks of gear to support it.
Well, I have a similar video of my "baby" which is somewhat newer - look at https://www.youtube.com/watch?v=-EQqfxiGgd8 beginning minute 5:00. The LN-3 you can see in Tatiana's video is from the 1960ties (F4, F104 and many others) and all computation is done mechanically using gears and ball resolvers for Sinus/Cosinus. In the FIN-1010 shown in my video being from the 1970ties, some calculation (e.g. integration of rate to orientation) is still done mechanically within the gyroscopes, platform alignment and maintaining the position of the platform as well as flipping the gimbals is accomplished by an analog computer (OpAmps and so on). Last but not least a bitserial digital computer is supervising the analog computer and changing its "program flow". Additionally the digital machine contains routines for calibration of gyro drift etc. The digital computer also solves the navigation equations of direction and distance to goal. So the fascinating thing is the combination of different technologies and the outstanding precision needed and achieved by the mechanical instruments. Altough the digital machine consists of only ~200 TTL chips, it is a 32 bit machine and delivers 33 navigation oututs per second. BTW: This enabled the Tornado aircraft to fly autonomously 200 ft above ground at supersonic speeds passing certain preprogrammed waypoints... Erik.