Understanding intelligent chargers

[Ip485]

I would like to understand how intelligent chargers work.

I think the charger is determining the state of the battery charge by a combination of measuring the voltage at the terminals and (possibly) its resistance. Is it doing so by pulsing the charge and taking an average of lots of readings microseconds apart?

I ask because when a charger is operating the voltage at the terminals is higher so it needs not to confuse itself by seeing its own charge. Then i thought you couldnt operate two chargers at once, but, as a previous discussion demonstrated, you can, and very effectively. So with two chargers one is not confusing the other, or so it would appear.

But how does this work!

Interestingly the usual search engines reveal very little in terms of more technical information.

[noelex 77]

The basic smart charger (solar or battery charger) algorithm is reasonably simple.

The charger delivers the maximum current until the bulk/absorption voltage is reached. This is the bulk stage. The charger then maintains that voltage by adjusting the output current until the voltage has been held for the absorption time (usually about 2 hours) and then the battery voltage is reduced to the float voltage.

If at any stage the battery voltage drops below the bulk return voltage (usually about 12.5v) a new cycle is initiated. For solar chargers a new cycle is usually initiated every solar day.

The above is only an outline. I have left out temperature compensation and equalisation. There are also many variations. The most common are to adjust the absorption time based on the start up voltage or to add a third “storage” voltage lower than the float voltage. Another variation used by some mains chargers is a drop to float if the output current drops below a low value.

It should also be noted that there are some very different algorithms. The popular Victron solar controllers vary the above so much that their method should probably be regarded as a completely different one. Some of the integrated systems have a seperate shunt to measure the current entering the battery that can be used to precisely control the absorption time. This latter system is best, but is the most expensive and difficult to install so is not often used.

Most algorithms work reasonably, but they first need some user input to adjust the voltage set points to the ideal value and the absorption time to a suitable number.

For most solar systems on a boat the voltage set points need to be increased and the absorption time decreased, but this is by no means universal.

[Ip485]

Thank you.


Lets say the Bank is at around 50% state of charge.


One charger is turned on.


So are you suggesting when switched on the charger then measures the terminal voltage and moves to an absorption cycle? During absorption what is the charger "doing" to determine when the absoprtion phase is complete? Is it relying just on the alogrithm or is is sampling the state of charge, and if so how?


You turn on a scond charger straight after the first. Does anything either charger is doing effect the other, and if so how and why?

[Ip485]

Further to, during the constant current stage it is my understanding that the charger is waiting to see a terminal voltage of around 14.4V before passing to the constant voltage phase. I dont understand how chargers appear to do such a good job when, for example, running two chargers, of ignoring the effect one charger may have on the terminal voltage. I also dont understand if there is a large load on the battery while it is being charged how the chargers adapt to the load (if at all). After all it seems to me the terminal voltage will be effected by a number of factors.

[noelex 77]

Originally Posted by Ip485

So are you suggesting when switched on the charger then measures the terminal voltage and moves to an absorption cycle?

The battery will start at low voltage and the charger will deliver the maximum current until the voltage reaches the bulk/absorption voltage. It will then start a timer that counts down while the absorption voltage is maintained and when the absorption time is finished the charger assumes the battery is charged (or more correctly that it is an appropriate time to drop to float) and the voltage will be reduced to the float voltage.

Originally Posted by Ip485

During absorption what is the charger "doing" to determine when the absoprtion phase is complete?

It is just counting down.

Originally Posted by Ip485

Is it relying just on the alogrithm or is is sampling the state of charge, and if so how?

No, there is no sampling of the SOC. However, the system works better than may be imagined. If the SOC is low, the charger will spend longer in bulk because the battery voltage will rise more slowly. If a load is added during the absorption phase and the absorption voltage is not maintained, the countdown pauses.

It is not perfect, but if the voltage set points are correct and you adjust the absorption time to be appropriate for your system, the charger will drop down to float at close to the right point. The initial SOC affects the time in bulk, but does not have much influence on the correct absorption time.

Originally Posted by Ip485

You turn on a scond charger straight after the first. Does anything either charger is doing effect the other, and if so how and why?

The simple answer is that multiple charge sources do not normally interfere with the correct charging of the battery to a significant extent. The battery is only interested in the correct voltage.

An analogy that may help, is imagine pulling on a rope and trying to keep the gauge reading 20lbs of tension. If a second or third person comes to help, if they have the same instruction the tension on the rope will remain the same. One person may be doing more work than the others, someone else may even stop pulling altogether because from their angle it looks like the gauge is reading 21lbs, but despite these differences the tension will be correct no matter how many people are pulling on the rope.

However, the above is a simplified view. There can be some minor undesirable consequences, although they are rarely significant. The most common is that if the voltage set points don’t match (and they will never be perfectly aligned) the absorption time can be applied by each source. So the absorption time can be applied by each charge source consecutively. So with multiple charge sources, setting a shorter absorption time is often better. It is also possible by adjusting the respective absorption voltages to mitigate this effect. The second problem is a feature present in a small number of chargers where the algorithm will drop to float if the current is low and the absorption set point is maintained. This can cause one charger to drop to float earlier than it should. This has no effect on the battery, providing the other charge sources can maintain the absorption voltage. If the other charge sources are turned off the algorithm will be incorrect. Unfortunately, the popular Victron chargers have this “feature”.

[john61ct]

Did not read it all, but want to agree with the essential point

You can put as many disparate charge sources of whatever types, in parallel and active concurrently as you like on a bank, without a problem.

Ideally similar setpoints, but NBD if not.

As long as

1. none on its own would harm the bank, and

2. total amps is not so high as to harm the bank

Yes, some algorithms may be suboptimal, but overall it comes out fine in the end.

[john61ct]

Originally Posted by Ip485

Further to, during the constant current stage it is my understanding that the charger is waiting to see a terminal voltage of around 14.4V before passing to the constant voltage phase. I dont understand how chargers appear to do such a good job when, for example, running two chargers, of ignoring the effect one charger may have on the terminal voltage. I also dont understand if there is a large load on the battery while it is being charged how the chargers adapt to the load (if at all). After all it seems to me the terminal voltage will be effected by a number of factors.

You are giving the chargers way too much credit.

The transition from CC Bulk to CV Absorb is not under their control.

Whatever the setpoint is, Bulk is striving to reach it, delivering max amps available and as per bank acceptance.

Once setpoint is reached, all the regulator does is prevent V from rising further.

It is the bank chemistry, resistance, SoC that determines the trailing amps level.

In an ideal world, the transition from charging voltage (Absorb) to Float is based on an endAmps setting,

so it does not happen until the batt mfg spec for 100% Full is reached, often .005C - .02C,

but that requires a BM/shunt, few mains chargers work that way.

Most sources use a dumb eggtimer for Absorb Hold Time, hopefully user adjustable.

Sometimes a more sophisticated algorithm.

It is up to the user to calibrate AHT as needed to ensure 100% Full is reached if possible, at least a few times a week for bank longevity.

[john61ct]

Originally Posted by Ip485

I ask because when a charger is operating the voltage at the terminals is higher so it needs not to confuse itself by seeing its own charge.

The bank and charger output are one circuit.

The bank's isolated voltage is low, the charger's Open Circuit voltage is high, connect them together, and if the wire is fat enough, the "negotiated" voltage is the same at both ends.

What V that is depends on bank SoC, resistance, AH capacity and charger's Amps output.

Higher quality chargers have a dedicated Voltage Sense wire to more accurately measure Bank Voltage, separate from the charging circuit.

[medsailor1]

I too would love to understand more , as I just don't grasp half of what I should know .
Here my example . To day we started off with an 12.5 v ( SOC 70%) the battery bank is 450AH , we had no wind so we had to motor for 10hours and 10 mins ,
at the same time over that time our solar panels (200w) put in 780wh or so my Mppt says , and the max V reached 14.26 .
The Victron is set to Absorption at 14.8 at one point amps drop to 2A with both the alternator and the panels .
What I don't understand motoring for that time plus what the panels where putting in why didn't the voltages reach 14.8 and the Victron entered Absorption stages .

[noelex 77]

Originally Posted by medsailor1

What I don't understand motoring for that time plus what the panels where putting in why didn't the voltages reach 14.8 and the Victron entered Absorption stages .

I would suspect the alternator voltage set point was too low.

[medsailor1]

So basically I never going to get the batteries to Absorption stage , unless I put on a bigger Alternator ?

[john61ct]

Solar was only 50-something AH, right? And was that panel output or controller output into the bank?

You need a solid external VR setup to get good charging of deep-cycling batts from an alt.

Also too-thin wires, bad connections?

Measure your V & A, ideally count AH over time too.

[noelex 77]

Originally Posted by medsailor1

So basically I never going to get the batteries to Absorption stage , unless I put on a bigger Alternator ?

I suspect the problem is primarily with the volage regulator. If the regulator settings are too low no matter how large the alternator the alternator output will be reduced.

10 hours of motoring even with a low output alternator should result in a higher SOC. An output of only 2A further suggests the problem is not the maximium output of the alternator, but rather that the alternator output was reduced prematurely by the alternator regulator.

However, if you want to charge the batteries quickly, when you are motoring for shorter periods, the combination of a smart alternator regulator, with the correct voltage settings, combined with a good quality alternator that is capable of producing high currents, even when hot, is hard to beat.

[john61ct]

Yes the **heat** factor is critical.

Many stock alt/VR setups just drop the voltage level, rather than current, in overtemp conditions,

almost completely stopping effective charging.

A good VR like Balmat MC-614 can be calibrated to keep putting out your desired setpoint, just de-rating amps when things get hot.

[medsailor1]

Originally Posted by noelex 77

10 hours of motoring even with a low output alternator should result in a higher SOC. An output of only 2A further suggests the problem is not the maximium output of the alternator, but rather that the alternator output was reduced prematurely by the alternator regulator.



.

Sorry I didn't explain properly .
The Alternator was pushing 35A later in the day the battery monitor was showing only 2A was going into the batteries,
The combination of MPPT and the Alternator , so am I right in thinking that the batteries where almost charged and if so why hasn't the voltages gone up enough for the MPPT to go into Absorption?
Or what am I missing ?