In one B737 Classic simulator I flew, actuation of the fire warning switch on the simulator instructor panel would cause an immediate overheat warning light (with appropriate Master Caution/Overheat lights on). Approximately five seconds later the fire warning switch actuated. This was useful to the pilot as even a few seconds early warning of an engine fire was helpful.
Yet another 737 Classic simulator was wired up so that the same actuation of an engine fire warning on the instructor panel instantly produced a fire switch illumination.
The difference in philosophy between the two simulators is quite significant. In the case of the first simulator in order for the simulator instructor to "arrange" a fire warning close to V1, it meant the overheat light would come on several seconds earlier than the selected fire warning thus alerting the crew. The Boeing advice is to reject below 80 knots and continue if an overheat warning occurs after 80 knots.
On the other hand, a rejected take off up to VI is advised for a fire warning. So, in the first simulator, the crew are faced with a period of uncertainty - especially if they (rightly or wrongly) assume in that particular simulator that the overheat warning will be followed by a fire warning. By rejecting the take off below V1 if the overheat light illuminates, there is less chance of running off the end of the runway if the take off performance is runway limiting.
I would be interested to know the applicable philosophy in other B737 Classics. In other words in the real thing (engine fire) would you always get an overheat warning first? If that is a correct statement, then is the second simulator mentioned above, where a fire warning actuation is not preceded by an overheat indication, indicative of faulty fidelity?