Current Controlled Battery Charger

[Cloud9]

Hi All,

I am working on a project that will run an alternator to generate a charging voltage for a battery bank. I will be controlling the speed of the alternator to get my desired charge voltage (after the rectifier).

The charge voltage from the alternator/rectifier is fed to a battery bank via a FET switch so I can connect/disconnect with my microcontroller. The battery bank has a current shunt in series with the ground so I can measure the charge current in the bank.

My question is can I PWM the FET to control the amount of current? It seems by doing so, ultimately I would only be affecting the voltage (and the current by law)

What is the most effective way to control the current of this circuit? I would like between 2-4A and the charge voltages can range from 14V-55V depending on the battery config. So wattage from about 56-220W.

I am assuming simple resistor limiting will be super power hungry.

Thanks!!

 

[MrAl]

Hi All,

I am working on a project that will run an alternator to generate a charging voltage for a battery bank. I will be controlling the speed of the alternator to get my desired charge voltage (after the rectifier).

The charge voltage from the alternator/rectifier is fed to a battery bank via a FET switch so I can connect/disconnect with my microcontroller. The battery bank has a current shunt in series with the ground so I can measure the charge current in the bank.

My question is can I PWM the FET to control the amount of current? It seems by doing so, ultimately I would only be affecting the voltage (and the current by law)

What is the most effective way to control the current of this circuit? I would like between 2-4A and the charge voltages can range from 14V-55V depending on the battery config. So wattage from about 56-220W.

I am assuming simple resistor limiting will be super power hungry.

Thanks!!

Hi,

Well you say you will be controlling the speed of the alternator to get your desired charge voltage, but then you say that you'll be using an FET switch to control the charge.
So which is it you want to do?

 

[MikeMl]

Charging lead-acid requires requires you to control the voltage at the battery terminals; the current will be what it will be as determined by the state of charge of the battery.

Hundreds of millions of automobiles cant be wrong. Do what they do in their Voltage Regulator. Compare the actual battery voltage to the voltage required to charge (See here, Read the section "Basics"). If the battery voltage is too low, apply 12V to the field of the alternator. If the battery voltage is slightly higher than the charge voltage, switch the field current off. The inductance of the alternator field determines the rate of field current increase and decrease. It is just like the thermostat in your house; a simple bang-bang control system.

A 12V automotive alternator will charge battery bank ranging from 12V to 48V, but your charge controller will have to know how many cells there are in your battery bank...

 

[Cloud9]

@MrAl - I will be doing both. The speed will affect the voltage - the FET is there just to turn on/off the load. I figured if there was a way to use it to limit the current, then great.

@Mike I appreciate the link but I understand the basics pretty well. My concern is to address the following possibilites:
Bad Battery
Improper connections
Possible new future battery technologies

Plus when the battery bank is powering a load instead of charging, I can limit the current instead of popping breakers.

You say a 12V alternator will charge a 55V bank. How? Do you mean tap into each one, basically changing the series connection to parralell? It just seems the overall cost and complexity of this was going to be alittle much. Instead I chose to manipulate the speed to get up to 40VAC (since I'll get x1.414 after rectifing and smoothing)


I do appreciate both of you taking the time to get in this discussion.

 

[Cloud9]

On a side note, I'd like to mention that for the load/charge switching, I was going to use MOSFETS but for the Amps and Power, it had to be an N-channel high side which was going to be too difficult to get the gate drives worked out in my short time. Ultimately, I am using SSR modules from Crydom for the switching now.

For my current mesasurement I went with a .01ohm resistor to get 10mV/A reading with a max drop of 500mV at 50A. Sound feasible? If drop at max 50A = 500mV, then if P = IV then 50A*500mV = 25W for my shunt resistors min power rating. Correct? (I will be over spec'ng to 30-50W for good measure)

I am measuring with a 12bit ADC with a 2.4V reference. I am thinking that will give me resolution of up to about 1/17th of an Amp, which is definately more than what I even need. My theory is rusty so feel free to correct or point out any caveats. I used ADC_VAL = (0.01vin/2.4vref) * 4096resolution, So adc result of 1A reading is 17.

 

[MikeMl]

...
Plus when the battery bank is powering a load instead of charging, I can limit the current instead of popping breakers.

You say a 12V alternator will charge a 55V bank....

You are making it way too complicated. An automotive alternator system supplies "loads" in parallel with the battery it is charging. An alternator is intrinsically current-limited due to magnetic saturation. An alternator rated to put out 60A will put out only slightly more than that into a dead short... (and it will do it without killing itself).

In an automotive system with a 60A alternator, if the total load (lights, AC) is 40A, there is only 20A left to charge the battery. The controller knows nothing about loads or current; it just keeps the system voltage at 14.2V +-0.1V! If the total load current were to be 70A, the alternator supplies 60A. and the remaining 10A is coming out of the battery. The controller sees the sagging voltage, and just applies full-on field current to the alternator.

You can think of an alternator as a constant-current source, with an open-circuit voltage of ~120V. That is why a nominal 12V alternator can charge a 55V battery, with one caveat. The diode-stack in the alternator may not have a PIV rating that will support that.

 

[Cloud9]

@Mike - I think that I have left out alittle to much info about the project. First, this will be a commercial project so longevity is vital. I know there is a fine line between too-complicated and well-designed in this case. I want to *try* to consider any potential problems and incorporate solutions into the design. There will be 2 seperate battery banks, one charging, and one powering a load. The goal of the current limiting is not only for the charge side, but could be useful on the load side as well. We cant make any assumptions about the load because we plan to allow the customer to hook up whatever kind of load they want.

 

[ronv]

Hi All,

What is the most effective way to control the current of this circuit? I would like between 2-4A and the charge voltages can range from 14V-55V depending on the battery config. So wattage from about 56-220W.

Seems like the best way to do it would be with the field voltage rather than trying to control the speed or trying to pwm it.

 

[tcmtech]

I don't see the point of using a SSR or anything else to disconnect the alternator from the load being that when the rotor field current is turned off the alternators output drops to zero anyway. Its a pointless control device that just adds complexity and cost to the system.

As far as regulating the output current that can easily be done with a combination of voltage and current feedback loops.

What is the mechanical power source that will be driving the alternator and why is it charging one battery bank while a second one supplies power to something else?

I hope this is not another over unity fail device as in battery A powers the motor that drives the alternator that charges battery B until its full then switch's around and battery B now powers the motor that drives the alternator to recharge battery A again.

 

[tcmtech]

You can think of an alternator as a constant-current source, with an open-circuit voltage of ~120V.

That doesn't make any sense at all being alternators have voltage regulators that keep the output voltage constant while the load current can vary from zero up to whatever the upper limit is of the alternators output. Thats why alternators have voltage regulators not current controllers in their regulation systems.

From your description it would be like saying your homes electrical system is a constant current source that varies its current between 0 and what ever the breakers trip or the fuses bow out at. Well that not a constant current its constant voltage source with a upper current limited by some means.

In my book anything that tries to keep its voltage stable while the current can vary by any amount under the designed upper limits not a constant current source.

A constant current source is just that a source that once a specific current value is set it tries to keep it there regardless of what its voltage has to do from zero volts up to whatever its upper output voltage limit is.

Constant current source definition, https://en.wikipedia.org/wiki/Current_source
Constant voltage source definition, https://en.wikipedia.org/wiki/Voltage_source

 

[MikeMl]

That doesn't make any sense at all being alternators have voltage regulators that keep the output voltage constant while the load current can vary from zero up to whatever the upper limit is of the alternators output. Thats why alternators have voltage regulators not current controllers in their regulation systems.

Have you ever used a clip lead to bypass the voltage regulator to apply full 12V field excitation to a spinning automotive alternator, or witnessed what happens if the switching transistor inside the voltage regulator shorts?

The alternator's field current goes to ~2A (determined only by the ~6Ω DC resistance of the field windin ). The alternator goes to full rated output (50 to 100A, depending on the alternator rating). If there is a 12V battery across the output of the regulator, the battery will absorb the current as charging current. If the condition persists, you will massively overcharge the battery, boiling the electrolyte out of it, destroying it. While the battery still has some electrolyte covering the plates, it will hold the voltage down to about 15.5 to 16.5V, but what happens after the last bit of the electrolyte is gone? The voltage at the output of the alternator will soar, going as high as 140V, destroying any 12V lamps and electronics designed for only 12-14V. The runaway alternator is effectively a 60A constant-current source with a compliance of >100V.

An automotive alternator is a current-controlled current-source. The equation that describes its operation is Io=k*If, where Io is the output current, If is the field current, and the gain constant k is about 30, meaning that 2A of field current produces 60A of output current. In an automobile, the Voltage Regulator senses the battery voltage. It shuts off the field current when the battery voltage is >14.2V. It turns on the field current when the battery voltage is <14.2V. The deadband between the switching points is tiny (~10mV). It is the inductance (>1H) of the field winding that determines the rate and duty-cycle ratio of the field current ON and OFF times, but in most automobiles, the effective PWM rate is about 30 to 50Hz. The function of the Voltage Regulator is to sense the battery voltage, but control the alternator's field current (not field voltage)! The field voltage at any given instant is either 12V or 0V (bang-bang control system, like the thermostat in your house). It is the average field current that dictates what the battery voltage is...

 

[tcmtech]

Yes I have done the direct connect thing many times experimenting with making home made alternator welders.

Once you take the voltage regulator out and put a continuous power to the rotor then yes it becomes a constant current source instead of a constant voltage source.

Calling it a constant current source while it still has its stock configuration designed to put out a stable constant voltage regardless of what current draw may be on it is what I am getting at as at minimal sounds ignorant or improperly informed.

Do you refer to your home AC power or portable generators or anything else that by design puts out a constant voltage (Unless modified or broken) while the load current varies anywhere from zero up to to what ever limit the device or circuit can supply as a constant current source as well?

 

[MikeMl]

The alternator is a constant-current source, PERIOD. That is what it does; that is how it works.

It is normally used as part of a charging system. It is the function of the Voltage Regulator (Alternator Control Unit) to make the charging system act like a constant-voltage source. If you separate the the alternator from the VR, and want to use the alternator as a component of another system, you need to understand how the alternator itself works.

If you go back to the OP's proposed system, he was planning to build a battery charging system using an alternator with a homebrew charge controller. If he is going to design his own controller, he needs to understand the attributes of the alternator. For Example, until I brought it up, he didn't know that a 12V alternator could act as a constant-current source and charge a much higher voltage battery bank.

I presented a mathematical model of how an alternator behaves as a circuit element. Either dispute that, or you are just quibbling about how to use words to describe something...

 

[tcmtech]

Then why does a properly functioning alternator it try to put out a constant voltage, usually around 14.2 VDC and any amount of amps up to its designed working load limit while the current from it varies?

As far as your equation goes I am not going to bother arguing with mathematics that are being used in the wrong application! That's not my fault.

Taken from wikipedia,

Automotive alternators require a voltage regulator which operates by modulating the small field current to produce a constant voltage at the battery terminals.

For the rest of those who doubt, https://en.wikipedia.org/wiki/Alternator

 

[MikeMl]

Then why does a properly functioning alternator it try to put out a constant voltage, usually around 14.2 VDC and any amount of amps up to its designed working load limit while the current from it varies?...

Divorced from the Voltage Regulator, it DOESN'T.

 

[MrAl]

Hi,

One of you guys is talking about the alternator itself and the other is talking about the entire alternator system. The alternator has a basic operation in and of itself i think Mike has described, and in a feedback loop (the entire system) the system is able to regulate the voltage, but the alternator itself does not regulate voltage...it could never do that on it's own because of the varying mechanical input power.

 

[tcmtech]

Exactly. Alternators, generators, linear power supplies/IC's, and so forth are referred to in their functions as a whole device not as to just what their individual parts or sub systems do and work at.
We don't call them transformers, capacitors, resistors, inductors, or diodes even though their very function is to transform one form of energy or voltage to another or because they may very well contain such a device in one or more of their sub systems or components.

In the case of automotive type alternators they are configured as constant voltage output devices and that's why they are classified by their output voltage first and their peak amps limits second.

Now relating to what the OP wants he has given very limited information on what type and size of batteries he will be attempting to charge and no information relating to the brand and model of the base stock alternator he plans to use nor the mechanical power source that will be driving the alternators mechanical input.
Given what he has suggested so far we all need more relevant info to work with before being able to suggest how to go about designing a dual function regulator system for his design.

Once we know the type/class of batteries plus their AH size, what base alternator he plans to modify, and what power source he is planning to use to generate this power we can start working on a theoretical design.

 

[MikeMl]

tcmtech is still arguing semantics. I didn't say an alternator should be called a constant-current source, I said that an alternator acts as a constant current source.

 

[4pyros]

I will be controlling the speed of the alternator to get my desired charge voltage (after the rectifier).

Ya but the speed controls the frequency not the voltage.

 

[tcmtech]

I still don't see how it putting out a constant voltage and variable current during its normal operating conditions could remotely be considered acting as a constant current source.

Do you also consider our power grids to be constant current sources powering our homes being they use alternators as the primary power sources?

If using the wrong terminology, formulas, and theories for a basic electronics related actions, functions, or devices in their normal intended design state of being is semantics then yes I am being nit picky!

Just because its possible to make any common op amp IC work as a astable square wave function generator does not mean that it can be called a 555 timer IC!

So anyway I am curious as to what sort of battery systems the OP plans to charge at 55 VDC and 4 amps.
My math suggests that for a LA based battery system that would be a 46 volt battery system which is not a common voltage I am familiar with 42, 48, and 54 volts DC are the closest I am familiar with and still even those are generally associated with industrial equipment, electric vehicles and large UPS type power systems of which all tend to have battery bank capacities rated in the many tens to many hundreds of AH values which at a 4 amp input would barely qualify as trickle charging.

Curious to see where this leads.