Is an alternator a voltage or a current source?

[soldar]

A conventional car alternator has a rotor which is magnetized with current and induces voltage or current in the stator which is, let's say, the secondary. Three phases and all that so diodes rectify the output.

A regulator provides feedback and control. It senses the output voltage and regulates the rotor current which in turn stabilizes the output voltage against variations in speed, load, etc.

Summary: the output voltage is maintained near constant by adjusting the rotor current which the regulator does.

My question is this: without a feedback control regulator, just the electro-mechanical machine is it closer to a voltage source or to a current source?

We know because Mssrs, Thevenin and Norton told us that mathematically voltage and current sources are kind of the same thing and can be mathematically converted. A current source has a very high output impedance and a voltage source has a very low output impedance. Ideally zero in both cases.

Imagine I have a car alternator, with no regulator, running at constant speed, feeding a constant current into the rotor, and with a load and the voltage and current at the load are whatever they are, constant, obviously.

My question is, if I vary the load does the alternator behave more like a voltage source and the voltage remains with little or no change or does the alternator behave more like a current source and it is the current that tends to become unchanged?

Is an alternator a current multiplier? Input n amps and get n multiplied by p at the output? 

Or is more like input n amps and get B volts output?

Again, assume constant rotational speed because, obviously, the torque and input power will change.

[trickysolution1976]

In general, alternators are categorized as voltage sources.

[CaptDon]

Because the voltage output from the windings can be exactly predicted by the number of lines of force cut per second it would be considered a predictable constant voltage source with a constant RPM input. The voltage will sag predictably by the winding resistance and a couple of magnetic effects caused by the current being taken from the windings, sort of a demagnetizing force as a simple explaination. It would not be considered constant current because the current will vary directly by the load applied remembering the predicted voltage would be constant and only sag due to the above mentioned reasons. The voltage regulators in years gone by had one circuit to limit the maximum output current to protect the windings from burnout but other than max current protection didn't care about the charge current. The voltage circuit was either a two step lowering the available charge current as the battery reached full charge voltage and shifting above and below the set point by a hysterisis factor or the circuit produced a PWM in either a solid state fashion or with a set of 'chattering' contacts on a voltage regulator relay. You should look up the old Delco Remy generator, alternator and voltage regulator field service manuals that explained all of the theory. Knowing not only how to service these systems but the theory behind how they worked and how they were controlled was part of becoming a Merchant Mariner with Engineering certifications. When at sea 'Murphy' is out to get you and 'MacGiverisms' are the norm until you make it to the next port.

[soldar]

OK, thanks. I always wondered about this.

So at a minimum RPM it already provides enough voltage to charge the battery and as the RPM increase the regulator decreases the rotor current in order to maintain voltage to float the battery because if the voltage tried to rise too much the battery would demand so much current that the alternator would burn itself.

[TimFox]

Any practical generator can be represented as a voltage source or current source (Thévenin or Norton equivalents, respectively) with an output resistance.
Those theorems have the important result that the output resistances for the two cases (series for voltage model and shunt for current model) are equal.
We characterize the practical generator as voltage or current source by comparing the output resistance to the expected load:  if the output resistance is much less than the load resistance, then it is more appropriate to use the voltage model, but if it is much higher than the current model is more useful for analysis.
For linear circuits, the choice between voltage and current source is up to the user, since they are equivalent, but usually one or the other is easier to use in the entire circuit.
As CaptDon explains for the alternator, usually the voltage-source representation is more useful.

[ejeffrey]

Also with any motor or generator, the mechanical side is part of the circuit.  A electric generator at constant RPM is a low impedance voltage-like source.  But if you drive it with a less stiff mechanical drive that looks more like a constant torque drive then the generator will act more like a constant current source as the RPM will vary with load.

[soldar]

I realize a complete electro-mechanical model is going to be quite complex.

When electrically loaded the mechanical shaft will be loaded and depends on whether it is constant speed, constant torque, constant power. In a car constant speed is closest for this purpose as the load on the motor is not too large.

Also the current in the rotor is going to vary because of counter EMF. etc.  And all this doesn't even begin to take into account magnetic and resistive losses.

That is why I specified fixed variables. Everything else being equal and the spherical cow...

It would be interesting for me to see a mathematical model of an alternator.

[Simon]

I think that in practice an alternator should be considered a current source. The reason is that unregulated the alternator will output more than 12V or 24V, otherwise the regulator would be unnecessary and the regulator is to guarantee the voltage not the current.

I am thinking here mostly of your car alternator, not the alternators used to power the grid. So your engine can run from hundreds of rpm to thousands, a whole order of magnitude variation. So voltage output cannot rely on the Kv of the alternator as at 1000 rpm you would get too little voltage with the correct voltage only supplied at say 6'000 rpm. Or you could have the voltage you want at 1000 rpm but at 6000 rpm you will have 6 times nominal. Clearly neither extreme is suitable and no in between solution will work either.

So much like your linear voltage regulator the voltage is controlled with current. Current, not voltage produces the magnetic field in the windings, and magnetic fields I believe primarily produce current in a winding. Thus your Alternator is a current to current conversion with a maximum compliance voltage that depends on the rpm of the alternator.

The actual voltage control is done by ensuring that there is enough current in the stator to produce the current that the load will be using at the rated voltage. The control though is done through voltage monitoring. Voltage too high, put in less current. Voltage too low put in more current. It's basically a crude switching converter with the mechanical energy converting to electrical.

Put it this way, I observe over the time I have been doing electronics that to control voltage you act on a current and to control a current you act on voltage.