How was the mechanical system 'de-coupled' from the electrical channels thus that any movement from the control column was 'ignored' by this channel ?
I'm sure Christiaan or Dude will have better explanations, but in brief :
In mechanical back-up control demands were fed to relay jacks which acted as force amplifiers so the pilot was unaware of control run friction. The autopilot also fed into these relay jacks. This meant that the control precision and ability to harmonise control forces given by the electrical control system was not degraded by mechanical system shortcomings.
In electrical signalling there was a dead space at the Powered Flying Control Units (PFCU) to allow for the difference between mechanical and electrical commands produced by autostabiliser activity. Variation of this dead space with flight condition gave the autostabiliser authority limits. Autostab. was not available in mechanical signalling when the PFCU servo valve was locked to the mechanical control system.
So the control column movements were never 'ignored' by either system, but the mechanical system never 'saw' the autostabiliser commands.
Cheers
Clive