Understanding intelligent chargers

[noelex 77]

Originally Posted by sailorboy1

So what, batteries probably experience too short of absorption way more often than too long.

This will not happen with the classic algorithm I outlined in post #2. Each charge source must go through its own absorption time. The danger with this algorithm is not that the absorption time will be too short, but the opposite problem that the absorption times will be too long. In some instances the absorption times of multiple charge sources will be applied consecutively.

The absorption time can end up too short if the charge source has the “trailing amps” feature, but this will only create a too short absorption time under some specific circumstances.

So in summary, the most common problem of multiple charging sources is an absorption time that is too long.

[john61ct]

Which is not actually "a problem" with FLA.

And in general a completely insignificant problem across the board,

compared to the fact that on maybe 95% of deep cycling cruising banks, AHT is way too short every day, thus they are never getting to Full, bank longevity murdered from routine PSOC abuse.

Yes, each owner **should** be checking, occasionally looking at trailing amps at the Float transition time. Not difficult, just slightly inconvenient.

Or use a BM / shunt to regulate all charge sources, less practical.

Either approach will solve **both** problems.

[Delfin]

Some good explanations on this thread, but it probably could have stopped with Noelex clear comments in post #2. That said, I would add a couple of points for clarification that might help....

1. Charge sources are not the same as charge source regulators - they seem to be unintentionally conflated by some on this thread. Regulators can be smart, or dumb, but the idea of a smart charger is a bit of a misnomer, since all chargers are essentially dumb. It is only the regulator that is "smart", and even then smart is a relative term that may include, as Noelex points out, things like temperature compensation.

2. The charge source, being dumb, simply puts out rectified ac current in the form of dc current at the maximum voltage the regulator allows. With an automotive alternator on a 12 vdc system, that is usually 14.4v. If you have a "smart" regulator managing the charge source, that voltage is presumably adjustable. Without a regulator of some kind, the alternator will put out its maximum, unregulated voltage, which I think with 12v alternators exceed 17 volts - not so good. The period of the charge cycle where the regulator is throttling the voltage output of the alternator is generally called the "bulk" phase, but some also refer to it as the "absorption" voltage.

3. Just because the alternator is putting out 14+ volts, doesn't mean that is the voltage that can be read at the positive terminal of a depleted battery. This is why you'll see the voltage slowly rising during the bulk charge phase with lead acid batteries, or really, really quickly with lithium batteries. During this phase while the voltage is rising, the acceptance rate on lead acid batteries also changes, meaning that the amount of current (amps) the battery will accept declines. With lithium this doesn't happen until the last few minutes of the charge cycle, which is why they charge a whole lot faster than lead acid - they accept that maximum current output of the alternator at the bulk voltage until they are almost completely full. Another term used during this phase is the "constant current", or cc phase because the regulator is offering to the battery a given amount of current, even if the battery due to a falling acceptance rate associated with the battery chemistry is unable to absorb it.

4. Once the regulator senses that the bulk voltage endpoint has been reached at the battery, the absorption phase or "constant voltage" (cv) phase is triggered. At that point, the regulator continues to put out the bulk voltage for a period of time a smart regulator has been programmed for. With lead acid batteries that can be a very long time, since their acceptance rate of current drops so much long before they are fully recharged. A common setting is a couple of hours, but that is usually too short to get a fully charged lead acid battery. For lithium, because they have such low internal resistance, no absorption phase is needed.

5. After that amount of time has passed that you programmed the smart regulator to allow for constant voltage, the regulator limits the output of the charge source to the float voltage. For lead acid batteries, this is needed due to their internal chemistry and high rates of self discharge. For lithium, folks debate whether there should be any float voltage allowed, with most acknowledging it is pointless with that chemistry, and possibly harmful to longevity.

Happy to be corrected on any of the above...

[noelex 77]

Originally Posted by john61ct

And in general a completely insignificant problem across the board,

The primary aim of “intelligent chargers” is to adjust the absorption time so the battery is not over or undercharged. I think this is a worthwhile goal, and helps prolong battery life.

If you disagree that you don’t need an “intelligent charger”, you don’t even need a multistep charger with bulk, absorption and float, just set the outgoing voltage to the bulk voltage. There is no need for a charging algorithm at all.

I think this would be a mistake.

Originally Posted by john61ct

compared to the fact that on maybe 95% of deep cycling cruising banks, AHT is way too short every day, thus they are never getting to Full, bank longevity murdered from routine PSOC

John, I think you are used to generator based systems. For most (but not all) solar systems the default absorption times are too long. Often the default voltage set points are little low, but this is seperate issue.

[Ip485]

This is a very interesting discussion for me. I am sorry to pursue the technical aspect but I find that especially interesting as information is so difficult to come by, and opinions so much at odds with this discussion.


Having read all the posts carefully I still feel that there isnt a complete answer to the supplementary question I posted which is that if the charger is sensing the terminal voltage, why doesnt the second charger (where two are being used) get confused.


Moreover a marine electrical engineer is of the view that the charger will "see" the terminal voltage which will be much higher than the battery voltage when the bank is already being charged.


From a practical point of view I discharged my bank to 60% today (per the BMV). The voltage was indicated as 12.4V. I then switched on one charger, and the voltage was indicated at 12.7V, and then the second about 5 minutes later, and the voltage rose to 13.0V. So, at least in theory, the second charger "saw" a higher voltage than the first.


Clearly both were charging the bank because if either were turned off the BMV indicated a significant fall in the charge.


The questions;


Does this mean that the second charger will go to float much earlier than it would were it the first charger in the system? In other words when it started it "saw" a higher volts because of the effect of the first charger, so perhaps the alogrithm would determine that the charge time would be less.


Is the electrical engineer correct when he says the charger is "seeing" the terminal voltage, and that this must effect a second charger?

[Delfin]

Originally Posted by Ip485

This is a very interesting discussion for me. I am sorry to pursue the technical aspect but I find that especially interesting as information is so difficult to come by, and opinions so much at odds with this discussion.


Having read all the posts carefully I still feel that there isnt a complete answer to the supplementary question I posted which is that if the charger is sensing the terminal voltage, why doesnt the second charger (where two are being used) get confused.


Moreover a marine electrical engineer is of the view that the charger will "see" the terminal voltage which will be much higher than the battery voltage when the bank is already being charged.


From a practical point of view I discharged my bank to 60% today (per the BMV). The voltage was indicated as 12.4V. I then switched on one charger, and the voltage was indicated at 12.7V, and then the second about 5 minutes later, and the voltage rose to 13.0V. So, at least in theory, the second charger "saw" a higher voltage than the first.


Clearly both were charging the bank because if either were turned off the BMV indicated a significant fall in the charge.


The questions;


Does this mean that the second charger will go to float much earlier than it would were it the first charger in the system? In other words when it started it "saw" a higher volts because of the effect of the first charger, so perhaps the alogrithm would determine that the charge time would be less.


Is the electrical engineer correct when he says the charger is "seeing" the terminal voltage, and that this must effect a second charger?

I believe what you are observing when the second charger kicks in is the effect of increasing the amount of current available to the charging battery, resulting in an increase in the terminal voltage of the battery and likely the acceptance rate. This increased terminal voltage will affect the first charger in the sense that it will shorten the time to reach the bulk voltage setpoint, but other than that, I don't see any other effect. Incidentally, I believe the increase in terminal voltage when the second charger comes on line is a function of the lead acid chemistry. I could observe that on my AGM bank when both the alternator and the charger in the inverter were on line, but I don't see it when charging my lithium bank. With lead acid, both voltage and acceptance rate increased; with lithium only the acceptance rate increases because more current is available to accept with both sources charging.

[noelex 77]

Originally Posted by Ip485

Does this mean that the second charger will go to float much earlier than it would were it the first charger in the system? In other words when it started it "saw" a higher volts because of the effect of the first charger, so perhaps the alogrithm would determine that the charge time would be less.

Generally, no. For the classical algorithm, the start voltage will make no difference. The charger makes no attempt to judge the SOC. The charge algorithm is not dependent on determining the initial SOC. The algorithm works reasonably independently of the initial SOC.

There are some chargers (noticeably many of the Victron models) that use the starting voltage to modify the absorption time. In this case, a second charge source that is already putting charge into the battery will act to potentially shorten the absorption time. This is not a bad compromise. The main danger of multiple charge sources is absorption times that are too long, so an algorithm that shortens the absorption time if multiple chargers are present is often sensible, rather than detrimental.

[Ip485]

Originally Posted by noelex 77

Generally, no. For the classical algorithm, the start voltage will make no difference. The charger makes no attempt to judge the SOC. The charge algorithm is not dependent on determining the initial SOC. The algorithm works reasonably independently of the initial SOC.

There are some chargers (noticeably many of the Victron models) that use the starting voltage to modify the absorption time. In this case, a second charge source that is already putting charge into the battery will act to potentially shorten the absorption time. This is not a bad compromise. The main danger of multiple charge sources is absorption times that are too long, so an algorithm that shortens the absorption time if multiple chargers are present is often sensible, rather than detrimental.

but surely, if the battery's SOC was around 80%, the last thing you would want is the charger to carry out a full bulk charge? If the alogrithm is indpendent of SOC on what is it based, as something must be determining how long the charger spends in each phase?

[noelex 77]

Originally Posted by Ip485

but surely, if the battery's SOC was around 80%, the last thing you would want is the charger to carry out a full bulk charge? If the alogrithm is indpendent of SOC on what is it based, as something must be determining how long the charger spends in each phase?

The time in the bulk phase is determined by the battery. If the SOC is high, the voltage will rise to the absorption voltage quickly and the charger will only spend a short time in bulk.

If the SOC is low, it will take longer for the charger to reach the absorption voltage.

At the end of the bulk phase the battery will be at a reasonably similar SOC independent of the starting SOC. Thus a fixed absorption time that is initiated after a variable bulk time works reasonably (not perfectly) to ensure at the end of the absorption time the battery is fully charged (or more correctly this is an appropriate point to drop to float).

This is the fundamental principal of the charging algorithm.

[sailorboy1]

Originally Posted by noelex 77

So in summary, the most common problem of multiple charging sources is an absorption time that is too long.

And just what is "too long" and what is the " common problem"?

I have FLA T105s for my house batteries. I have my alternator regulator and my solar controller set to stay in absorption for 6 hours (or 2 amps for the solar). Been that way for over 2 years and haven't seen any sign of damage.

I do have to add about 1-2 cups total of water to the 4 of them every 1-2 months.

[Ip485]

Originally Posted by noelex 77

The time in the bulk phase is determined by the battery. If the SOC is high, the voltage will rise to the absorption voltage quickly and the charger will only spend a short time in bulk.

If the SOC is low, it will take longer for the charger to reach the absorption voltage.

At the end of the bulk phase the battery will be at a reasonably similar SOC independent of the starting SOC. Thus a fixed absorption time that is initiated after a variable bulk time works reasonably (not perfectly) to ensure at the end of the absorption time the battery is fully charged (or more correctly this is an appropriate point to drop to float).

This is the fundamental principal of the charging algorithm.

OK that makes sense. So the charger is referring back to the bank and measuring the voltage? However, if the voltage appears higher because a second charger is running then surely one or other charge would go to float a lot sooner, and the remaining charger would register a fall in the SOC and remain in bulk for longer?

[noelex 77]

Originally Posted by sailorboy1

I have FLA T105s for my house batteries. I have my alternator regulator and my solar controller set to stay in absorption for 6 hours (or 2 amps for the solar).

Flooded lead acid batteries, such as yours, are the most tolerant of an excessive absorption time. Nevertheless, you will have a longer battery life if the batteries are charged optimally.

A 6 hour absorption time is likely to be too long for maximum battery life. The exception is the Victron solar controllers that do not use a normal measurement of absorption time. A 6 hour setting on the Victron controllers is likely to be around one hour of real absorption (although this is only a very rough rule of thumb) and this can be a suitable number.

Have you measured the acceptance current of the batteries at the end of the absorption cycle? If so, what was the value? This is the best method. When combined with the battery bank size, it will tell you if the absorption time is appropriate.

[noelex 77]

Originally Posted by Ip485

OK that makes sense. So the charger is referring back to the bank and measuring the voltage? However, if the voltage appears higher because a second charger is running then surely one or other charge would go to float a lot sooner,

The voltage is the battery voltage so it will be the same for both battery chargers (ignoring possible minor differences due to voltage drop).

Originally Posted by Ip485

the remaining charger would register a fall in the SOC and remain in bulk for longer?

The chargers do not attempt to measure the SOC.

[john61ct]

OP, you keep operating on the false assumption the charge source is controlling voltage during Bulk or current acceptance during Absorb.

It is not.

Even at 80% SoC starting point, 6 hours of charging may be required, even with a 200A current available to a 100AH bank.

Capiche?

Now with a high-CAR chemistry in that scenario, Bulk CC may only last thirty minutes, and the remaining 5.5 hours is spent in Absorb.

Now say the current is 15A with that 100AH bank. Bulk CC may last three hours, the transition occurs at a higher SoC, so time spent in Absorb is only three hours to get to .005C endAmps.

The charger no matter how "smart" **has nothing to say** about those above variables.

It knows the voltage setpoints, and has an AHT algorithm, often a straight egg-timer. And that is it, wrt the basics

until you get into temp compensation, shunt-based transition to Float, relatively abstruse stuff.

____
And again, see my previous statements about multiple charge sources active concurrently.

There are so many variables and moving parts trying to imagine how/why there might be problems.

In real life, there is no problem for you to worry about,

from thousands of direct first-hand reports, dozens by very well-respected experienced expert members across many forums

[john61ct]

Originally Posted by noelex 77

The chargers do not attempt to measure the SOC.

+1 +1
Yes OP, chargers never know anything about SoC.

That would require a shunt / battery monitor in place, and it is only going to signal Stop Charging based on endAmps measurement,

and that's very rare, maybe .001% of high-end installs.

It is up to **you** the owner, to ensure AHT is set long enough to get to endAmps, at least a few times a week.