2 Differently Scaled current sensors for Space Vector Modulation Motor Control

[Glenn0010]

So correct me if I'm wrong but in sensorless control using space vector modulation you measure the current to know the rotor's position and hence you can know how to drive it

As I underetand At low rpm the problem is the current is small and hence you cannot measure it properly.

To solve this issue, lets say you, have shunts as current sensors. Couldn't you connect 2 differential op amps to each shunt? One would have a high gain to get an accurate current measurement at low speed while the other would have lower gain to measure at high speed

Would this work? And you would do away with hall effect sensors

[Rerouter]

Why not just use the hall sensor for feedback?

I am not fully able to understand your question, but at a guess its to use a small 3 phase motor like one for a quadcopter as a servo motor.

At near 0rpm. You really do need feedback, and a hall sensor may work so long as it meets your requirements for final position accuracy.

What the electronic speed controllers for quads do is time the zero crossing of the generated emf of a coil. Which is 0 at 0rpm. And very tiny at rpms above that. This can still be used at low rpms as long as the system has a fair amount of inertia. As its predicting when the coils are aligned to ramp up the current in a given coil.

[Leo Bodnar]

So correct me if I'm wrong but in sensorless control using space vector modulation you measure the current to know the rotor's position and hence you can know how to drive it
...
To solve this issue, lets say you, have shunts as current sensors. Couldn't you connect 2 differential op amps to each shunt? One would have a high gain to get an accurate current measurement at low speed while the other would have lower gain to measure at high speed

Are you sure you you are using right terms?
Sensorless means no sensors at all, not even Halls.  And then you use BEMF to find out the current position.

And yes, dual gain opamps were used in such configuration forever but I am only aware of DIY products.  To be honest, it is pretty useless unless you are making a drive that can control huge range of motor currents.  If you size it to a particular motor 10 or 12 bits are more than plenty.

Leo

[filssavi]

If it were that simple everyone would be doing that...
Unfortunately you are confusing 2 different problems

The current sensing is needed for field oriented control to control torque in both sensors and sensorless controls.

To estimate rotor position in sensorless mode you use a flux observer, by sensing the back-emfs, at low rpm those are weak and that is the problem

This problem can't be solved by gain alone first of all at some point (sooner than later) you'll run into noise issues.
Another problem is that at 0 rpm you don't know how the rotor is oriented so you need some other technique (high frequency injection)