[wingover]

In boat world, electric propulsion partisan, have to face heavy sarcasm for fuel fans, My EV...is in fact EB, slim 40 feet sailing trimaran.
EV forums are a peaceful places ! that's why i am here !

As regenerating is as important as the propulsion in a sailing boat, motor will run 24h/24h when sailing. 1000 hours of use is reached only in a few months, so i think PM brushless motors are almost compulsory for my application.
Looking at actual applications of this kind of motor and sensorless controllers, it seems they are used a lot in aerospace and military applications. This let me think it is a good technical choice for bulletproof applications.

Also, for a direct drive low revering (compulsory for efficency) of the propeller shaft i will use a 24 to 12 poles motor giving full speed at 900 to 1200 tr/mm. What will be the effect on the controller ? I guess this helps ?

My application runs under 50 volts and motor will be 4 to 5Kw, but for the moment i dont not know if it will be AC synchronous, asynchronous, PMAC...as it is a little confused in my head about advantages of each as far as efficency, weight and regeneration efficency. I found a 24 poles motor from Perm in Germany, but for the moment this motor is only qualified as a wind generator and they ask for time to do more test to use it as a motor. So if you have ideas about other motors...let me know !

The question i need to answer now, is about the performance of the controller for reading informations with the BEMF (Back ElectroMagnetic Force) from the motor and then giving the good orders. I have been told that all the controllers are not the same on this aspect . Some do it well, some get lost under certains situations...
Is it true ?
In what situation will the controller be lost or do not execute correctly the orders ?

I found this first answer that is giving usefull informations

Originally Posted by abudabit

Here is how it was explained to me:

With no hall sensors in the motor and the motor not turning there is no way for the controller to know the position of the rotor (as far as I know). The way controllers know the position of sensorless brushless motors is by detecting back emf in the various windings. When not turning there is no back emf. When slowly turning there is very little back emf. To detect the position during slow or no turning the controller cycles pulses through the windings in a circular pattern. This is very inefficient because it is basically blindly (but in a useful pattern) running electricity through the windings.

Once the back emf becomes large enough to detect then the sensored and sensorless can become equally efficient. But in stop and go traffic, or during a stall, or starting from a stop, or crawling extremely slowly, etc. it is massively less efficient. And those are the times you need efficiency the most because those are the highest torque applications in an ev.

I could be incorrect as I have never worked with an unsensored brushless.

What is your experience with sensorless motors ?
Is it preferable to forget about sensorless as the global reliability won't be affected using hall sensors ?

Thank's !