Can a motor controller be seen as a Current controlled source?

On the DC bus side, you will need to take into account the internal resistance of the batteries because the supply voltage dips as current increases because of this internal resistance.  When the system is in regen the voltage climbs relative to this internal resistance as well. 

The regen in most controllers is a side effect of the flyback diode on the switching device the controller will regen unless the IGBT mosfet or whatever actively keeps the potential of the buss close to the potential of the reactively generated voltage.  

As to treating a motor controller like a current controlled source it would be a very simple simulation.  It would be treating the system like a constant load like a pump, and it could work where if the current commanded falls below a certain point current will flow out of the controller onto the bus.  

Another thing to take into consideration is the ripple on the DC bus controllers cause by switching lot of amps.  Some component blow fuses for no apparent reason because of this and need inductive reactance added to the input path to bring the ripples to an acceptable level. 

Good luck,

forrest2192002 wrote:

 

On the DC bus side, you will need to take into account the internal resistance of the batteries because the supply voltage dips as current increases because of this internal resistance.  When the system is in regen the voltage climbs relative to this internal resistance as well.

So, if there is no battery connected but a super capacitor then the voltage keeps rising (provide the initial voltage of the super cap is at bus level and the voltage level at the controller terminals is higher)?

At the moment I am ignoring ripple from the motor controller, I am just looking at the curretn profile in and out. This is what I need for my simulation.

cheers

Hello,

I don't think that the capacitor and the CCS are standard Simulink blocks - can you tell me where they came from ?

Michael Kellett

Thats right they are blocks from the SimPowerSystems toolbox. Which is an expansion to the simulink block set. It allows for simulation using a more electronic approach. I was initially looking to build the whole lot in simulink but then the bi-directional nature of a converter would become problematic. In this case, I can use Simulink for the control strategies while taking a electronic approach to the circtuiry elements.

Hello Dirk,

I'm sorry but I can't really help you with the PowerSystems blockset.

(I don't have it and I've never used it so I would be shooting in the dark !)

Michael Kellett

Hi Michael,

I understand. But the question was not about the toolbox, but about whether a motor controller can be seen as a variable current source with a capacitor. I think it can based on a setup of an inverter.

An inverter has basically very little resistor  but with a capacitor in front it can supply high current to the motor (in an induction motor current is the defining factor not so much voltage). Under regen conditions the motor basically becomes a voltage source (l believe the voltage raises as a consequence of the back emf increase). Any current into the capacitor raises the voltage of the capacitor (in my assumptions I am assuming that there are no problems with saturation in the motor).

According to these assumption I should be able to see the motor controller as a variable current source.

Is this more or less correct or are there very large discrepancies in my thinking?

Cheers

Hello Dirk,

I'm unconvinced that you can model the controller this way but it does depend on how you control the current.

Your capacitor is being charged by a constant current but that means that the power going into the capacitor increases as the voltage across it rises.

So I think that with the model you have shown me the energy in the cap would increase according to t^2 for constant power from the motor - so the current controlled source is making energy.

I'm not sure - it's all in the blockset - in real life there are no current controlled sources

Try running the model - do you get sensible results ?

Michael Kellett

The way I see it under normal conditions:

The throttle is pressed and the motor controller starts to move the motor (through FOC or whatever). This makes that the motor controller demands a current from the battery (the input to the contrller is a capacitor for some smoothing and voltage measuring).

So, so far a current controlled source should work. Simulations show a reasonable match to other simulated outcomes in literature.

Under regen conditions (and this is where I am having some doubts, but nothing that realy would stop use of the current controlled source).

The brake is pressed and assuming that we are not going to recover all the energy but only a part of it. The system is going to react in reverse with current flowing out of the controller. The controller I am using on my dynamometer has an "over voltage under braking" protection which monitors the voltage across the input capacitor and when that goes to high it starts burning of current. In other words,

So I think that with the model you have shown me the energy in the cap would increase according to t^2 for constant power from the motor - so the current controlled source is making energy.

The energy that flows into the capacitor is then further transported to -for example- a battery pack.

I got the idea from a paper I read, but under buck (regen) conditions I was encountering some behaviour that I could not explain (was not explained in the paper and I was surprised the author had not made a note on that behaviour). The author connected a buck converter to a bus (with the bus simulated by a curretn controlled source) and only used current control to control the current through the converter. But I found that if the input capacitor isnt monitored the voltage will go sky high or drop to input voltage and the converter will stop functioning. The explanation to this effect is simple: because of the switching behaviour of the converter you do not know exactly how much current you draw out of the capacitor and how much will flow into it. And the tipping point is very steep. i.e. too much current and the capacitor voltage will drop; too little current and the capacitor voltage will keep rising (as per your quote again).

So, all in all, I think it can be done but I wondered why this wasn't reported in the paper. Something I considerd fairly important to the overall functioning.

My question, here was to find out if more people had looked at this before and if there are serious objections to the use of a current controlled source as a motor controller (in simulation).

cheers,

I'm sorry but I don't think I can help any more   - without seeing the complete model and some actual data I'm just guessing what you are doing.

Michael Kellett

Lol think I just found my answer:

http://www.dspace.com/en/inc/home/products/hw/simulator_hardware/simulator_specific_hardware/electronic_load_module_30v.cfm

quote:

It can work as both a current sink and a current source to provide bidirectional current flow, i.e., it generates or consumes real current on ECU motor outputs.

I think this answers my questions.

I'm happy if you're happy but the dSpace device is an electronic load intended to simulate a motor not a controller.

Michael Kellett

True, but what is an inverter but a set of switches that create a path for the current to flow (there is not voltage involved - assuming ideal switches equals no voltage drop). The input capacitor would then see only a sink or a source of current.

True, but what is an inverter but a set of switches that create a path for the current to flow (there is not voltage involved - assuming ideal switches equals no voltage drop). The input capacitor would then see only a sink or a source of current.