Quote from vilas:
"On the other hand the first fatal crash of A320 in 1982 India was similar because the crew switched only one FD OFF and remained in OPDES and imagined they were in speed mode. Had they switched off the other it would have switched to speed mode."
That's correct, except for the date. The A320 did not enter airline service until 1988. To avoid any confusion, the accident vilas is referring to is the Indian Airlines accident at Bangalore in 1990, as recently mentioned on another thread
here.
Sadly, the experience that we in the launch-customer airlines had previously gained may not have been passed on. A number of incidents involving the speed dropping well below VLS, including some in which alpha-floor had been activated, had led to our SOP of switching both FDs off for visual approaches.
The problem results, as vilas indicates, after a transition from the normal idle-descent phase to a visual profile when one or both FDs are left on in IDLE/OPDES mode and the PF pitches up in manual flight above the FD pitch bar. In those days, unless the crew selected SPD mode, the A/THR would remain in IDLE mode unless and until it was disengaged by the crew or by alpha-floor - regardless of the speed selected by the crew.
Following the accident, a protection was introduced by which the A/THR changes automatically from IDLE to SPD mode if the IAS falls below VLS (vilas may tell us the precise figure, which I don't have to hand).