I've read an idea posted several times, that the low inductance of the larger BLDC motors leads to controllers that need to manage >1kA of peak phase current. I've done some ballbark napkin estimates of what these motor winding inductances might be, and come up with a 100 uH - 5 mH range for devices like the big turnigy motors and the colossus. A 30 kVA UPS operates with only a few dozen uH of output inductance at 20 kHz or less, and we don't have a problem controlling them at all. Why do RC controllers have a problem with it? What am I missing? Does anyone have a real, hard winding inductance measurement on a machine in this class that you would be willing to share?
BLDC inductance ballbark & high phase current
- Thread starter jdb
- Start date
Ricky_nz
10 kW
I believe the issues are because RC controllers are cost optimised far more than even UPSs..
My motor controller project is quite capable of controlling current into very low inductances with no problems but it has phase current sensing and a fast micro. Two things that a RC controller would not likely have.
I can control the current in into a short circuit enough to avoid hitting hardware current limit
and thats at 12KHz so the only inductance is in the wire used for the short. I can go faster, its just not necessary yet.
Edit:
Turnigy 80-100 180Kv 13uH - 27uH rotor position dependant and 12 milliohm across 2 phases of the delta windings.
My controller project: http://endless-sphere.com/forums/viewtopic.php?f=2&t=23205
My motor controller project is quite capable of controlling current into very low inductances with no problems but it has phase current sensing and a fast micro. Two things that a RC controller would not likely have.
I can control the current in into a short circuit enough to avoid hitting hardware current limit
and thats at 12KHz so the only inductance is in the wire used for the short. I can go faster, its just not necessary yet.Edit:
Turnigy 80-100 180Kv 13uH - 27uH rotor position dependant and 12 milliohm across 2 phases of the delta windings.
My controller project: http://endless-sphere.com/forums/viewtopic.php?f=2&t=23205
Jeremy Harris
100 MW
Part of the problem is that the vast majority of RC controllers have no form of current limiting at all. Even the top end ones only have a very crude way of measuring phase current (some do monitor the voltage across the turned on FETs I believe).
It's not just phase winding inductance that matters, a low inductance is OK if the winding resistance is so high that it effectively limits peak current. The thing that's challenging for a controller is driving a combination of low resistance and low inductance.
There's also a fair degree of misunderstanding about just how high the phase current can get under normal circumstances. It's easy enough to estimate it, if you know the supply current limit and the motor applied voltage (or the battery voltage and duty cycle). For example, with a max rpm of 6000 at 48V, that's being loaded so it an only run at 2000 rpm, with a supply current limit of 50A, will have a phase current of around 3 times the supply current, or 150A. Power is always conserved (less losses) so the phase current can be estimated for any given supply current and motor rpm.
The worst case is undoubtedly when the motor is running slowly and is highly loaded, with no supply current limiting. This can never occur with an RC model aeroplane, because of the cube law relationship between power and rpm for a propeller. It can easily happen when using one of these motors and controllers on a bike, though, which is where the problems occur. With no supply current limiting, the controller will try and draw as much current as it can and will also try and efficiently convert supply voltage to phase current. If the situation was as above, but with no limiting, the supply current could easily be 200A, meaning that the phase current could be around 600A.
Jeremy
It's not just phase winding inductance that matters, a low inductance is OK if the winding resistance is so high that it effectively limits peak current. The thing that's challenging for a controller is driving a combination of low resistance and low inductance.
There's also a fair degree of misunderstanding about just how high the phase current can get under normal circumstances. It's easy enough to estimate it, if you know the supply current limit and the motor applied voltage (or the battery voltage and duty cycle). For example, with a max rpm of 6000 at 48V, that's being loaded so it an only run at 2000 rpm, with a supply current limit of 50A, will have a phase current of around 3 times the supply current, or 150A. Power is always conserved (less losses) so the phase current can be estimated for any given supply current and motor rpm.
The worst case is undoubtedly when the motor is running slowly and is highly loaded, with no supply current limiting. This can never occur with an RC model aeroplane, because of the cube law relationship between power and rpm for a propeller. It can easily happen when using one of these motors and controllers on a bike, though, which is where the problems occur. With no supply current limiting, the controller will try and draw as much current as it can and will also try and efficiently convert supply voltage to phase current. If the situation was as above, but with no limiting, the supply current could easily be 200A, meaning that the phase current could be around 600A.
Jeremy
bigmoose
1 MW
http://endless-sphere.com/forums/viewtopic.php?f=30&t=23059&hilit=current&start=180
Third post down I posted some measured numbers for three motors I had my hands on.
Third post down I posted some measured numbers for three motors I had my hands on.
Even without phase current measurement, an open loop V/f controller should be able to drive any of the motors that bigmoose documented (thank you, sir!). That way, the motor controller acts somewhat more like a current-source inverter rather than a voltage-source inverter.
Ricky_nz
10 kW
Provided it has sufficient knowledge about the motor.jdb said:
A V/Hz controller would need to know more about the motor than what is typically configured on an RC controller to accuratly control the motor.
The RC controller manufacturers have decided that their market wants cheep and easy to setup and some things had to be thrown out to achieve this.
