Hi Alfredo.

The failure you describe is typical of an electronic control module that does not have the engine temperature signal or this is very different from that of the air. Not having certainty of the actual temperature of the engine coolant, lights the electro -ventilator, commanding the relay to close (avoid engine overheating)
Now I see that you measure the 5V and mass reference well. Now you must be careful with the themist used by Ford and Mazda, they are in many cases of positive coefficient or PTC. That is, the thermistor resistance rises with the increase in temperature and does not fall as in the NTC themers used by the vast majority of engineers designers of modern electronic control systems.
I recommend measuring the voltage drop with the circuit running, both in the signal line and in the mass, ideally with an oscillograph. You can have a thermistor or temperature sensor changed, which does not correspond to the one that uses your car.
The other, much less likely, is the spoiled control module, you already know an analog signal, must enter the ECU, filter and/or amplify, then digitize. The ECU Motor only processes digital signals ... If the signal enters the ECU Motor and by the scanner you do not resibe the temperature data or it has a subtituto value of -40 degrees (hence the widening of the imjection pulse can be. Test the above and how many. Greetings from Chile.