Charge in Parallel, Discharge in Series

I assume you mean to do this without disconnecting and reconnecting. If you have 4 cells, a 4 pole double throw switch or relay and 8 diodes can be used. Of course, this could be extended to any number of cells.

https://electronicpowersupply.blogspot.com/2014/11/make-parallel-battery-charger-circuit.html

A charger across each cell with the cells isolated with diodes is another possibility, but it might be more expensive than the switch.

https://electronics.stackexchange.com/questions/222079/separating-lithium-ion-cells-for-charging

In the OP, HTRN said "I do not wish to do this with relays because of the expense." - I'd assume that cuts out switches as well, and makes multiple chargers a definite no-no.

Sometimes a problem is constrained beyond the point where an economic solution is possible. At that point, one needs to walk away, go home, and leave the bean counters to their deliberations...

With isolated chargers, I don't see a need for the diodes.

The cheapest hardware method is to remove the batteries and connect each battery to one charger at a time. This will be the most expensive approach in time, manual labor and probably failure rates. My point here is that until one considers all of these attributes and agree on a conversion factor of each of these attributes to a common unit (often a unit of money) you'll have no idea which approach is "best".

Connecting and removing manual link bars of copper wire with appropriate lugs on or charging the batteries a few at a time (depending on the charger characteristics).

It's a simple but slightly time consuming method, really anything else involving components is going to be expensive unless you get them second hand at a junk shop or have them lying around already.

What about opposing PWM setup...? ...charge could input on one frequency and output could be on the opposite...or you could have the charging voltage level above the discharge level and have zener diode regulators....

When you're dealing with hundreds of amps, you probably should think more about safety than the expense. Just sayin'.

That having been said, knife switches are relatively inexpensive and can carry a good amount of current. With a number of switches in parallel, each switch would not have to carry the high current of charging all batteries in parallel through a single charging path. I still think that using N double throw knife switches for N cells is your best option. (N/2 switches for DPDT, etc.)

(DPDT Switch)

Er, that's actually an arrangement for "charge in parallel and discharge in parallel", Mildred.

How many cells are you thinking of?

If you used a 3 terminal system, It looks like you would need (number of battery cells -1) switches, i.e. MCT (MOS Controlled Thyristor) with low forward drop, but you are right, your float voltage for batteries when charging would be different by some small amount, as the string gets longer.

I think you need to be careful on that; series and parallel entail some opposite battery polarities.

Wouldn't it be better to have 4 switches, with the fourth one located closest to the terminal at the lower right-hand corner of your diagram?

I guess that would depend on the output circuitry of the charger, I think if a full wave bridge, would it matter?

We'd want to check the charger could stand to be floating, i.e. filter capacitors, etc.

I assumed that we could have two floating circuits for either the series or the parallel, not necessary to have a single common potential for both circuits. In industrial dc power supplies, we tend to avoid earthing either side of our dc supplies

WRONG! On discharge you have +ve to +ve and -ve to -ve. Nice try!

Ooops

I don't see you doing it without a relay/contactor or some form of solid state switch.

Below is a fairly simple diagram that I devised that uses a single multi pole contactor and a pair of diodes. You can add any number of extra batteries in series with just the addition of extra contacts.

The advantage that I see with this layout over that shown by Rixter is that

1. There are far fewer components so likely cheaper.

2. The two diodes are only in the charge circuit and therefore only carry charge current

3. There are no diodes in the battery-load circuit and therefore no resultant voltage drop.

4. One 3 pole contactor could control 4 batteries.

When the charger is energised, the relay/contactor is also energised thus opening the NC contacts and isolating all batteries from each other.

The N/O part of the C/O contact in the charger output is a failsafe to prevent charger voltage being applied to multiple batteries while still connected in series should the relay coil fail.

There will be about 1.2 volt drop in the charge circuit, a bit less if using Schottky diodes, but this could be compensated for by increasing the charger output - which may be as simple as charging AGMs on a flooded cycle - being sure to disable any equalisation program.

When in run (not charge) mode if we call top battery A, next B, etc and we call contacts top X, Middle [ Y and notY (N. O. contact for charging)], and Z then D- is solidly connected to C- and D- is also (through Z) to C+. Short circuit. Also B- solidly connected to C- and (through Y) to B+.

to me the only logical solution is to increase charger voltage and let them in series.

Most battery charger outputs are galvanically isolated from ground and input voltage connections. Therefore you can just disconnect at one point of the series of batteries to the rest of the load. Then connect each individual charger to each individual battery.

There is a way to do this automatically. High current, IR loss, and cost are just engineering challenges. I've often wished I had someone who wanted to invest in producing it...

Interesting paper here....

http://www.fsec.ucf.edu/en/publications/pdf/FSEC-CR-1292-01.pdf

It seems to me this could be done with a number of FETs turned on/off to perform the task, in addition, they could be used as synchronous rectifiers - this would get rid of all the diode drops and which are a large loss at low voltages.

Just get FETs with a high enough rating to resist the PIV

<...still trying to find...expense...>

It depends upon the value of time. If a designed, installed, switching network is undesirable on the basis of outgoing money expenditure on equipment, one could always go the no-expenditure route: simply manipulate the connections between charged batteries using available hand tools to achieve the desirable configuration for discharge. Reliability then depends upon the dedication and skill of the individual doing the manipulation.

Why this option is still being sought might forever remain a mystery.

The forum could not possibly evaluate among the options.

Thanks to everyone for all the valuable feedback.

I realize that what I am asking is not currently feasible within the constraints that I have imposed.

Instead, I am now searching for a method to automatically balance the charges on the batteries while being charged in series.

Perhaps to bypass an individual battery when it reaches full charge, and pass the current on to the next battery in line.

Automatic temperature compensation would also be nice, but not a deal breaker.

Battery balance was my main reason for desiring to charge in parallel.

I appreciate all constructive feedback and look forward to possible solutions to this question.

I guess that you are aware that a 12v lead acid battery already comprises 6 cells in series which you have no chance of balancing individually. Lead acid cells have a degree of tolerance to overcharge as they tend to gas off excess energy input, so a bit of imbalance is of little concern.

Having said that, balancing individual batteries in a series string can give some advantage as the charger only measures the total string voltage (as it does on a single battery) and goes to float when that parameter is met, this can leave some units over or under charged especially if there are deficient batteries in the pack. The longer the string, the higher the charger voltage required, and the higher the out of balance possibility becomes.

I use a Marsen HA02 balancer on my LiFePo4 pack in my RV. It keeps the cells balanced to within 10 to 20mV which would give far more accuracy than you could ever wish for with lead acid technology.

It can handle up to 4 x 12v batteries in series, but multiple balancers can be overlapped to manage even more batteries in series

I was thinking something along the lines of a 7.4 volt zener,across each 6volt battery to trigger a transistor or scr to bypass the battery after it reaches the desired level.

But all of the batteries will be at the same voltage if they are wired in parallel, well except for the IR drop of the wires and presumably I→0 as a battery becomes charged.

That was the objective of my original question,but after considering the difficulty of this to achieve,I am considering another method to equalize the voltages on all batteries.

So my question now is:How to achieve balance of all batteries while charging in series in an economical and effective way.

I have seen the commercial version presented by spades,(Thanks,Spades!),but I am seeking a simpler method,perhaps by using appropriate zener diodes to bypass the battery when a desired voltage is reached.

Perhaps an op amp with high output power transistors?

I am assuming that these are Lead Acid batteries.

If the charger is set correctly, you should have a maximum charge voltage of 2.45V per cell or 7.35V across a 6V battery. This is lower than your suggested cutoff voltage of 7.4V and so would achieve nothing in that regard.

If we assume that one of the batteries is defective (maybe a shorted cell) so that its lower terminal voltage causes the average voltages of the other series connected batteries to rise higher than the permissible maximum, then your individual battery voltage cut offs would operate to prevent that, but you would then create other problems:-

When charging an LA battery, there is a phenomenon called "surface charge" where the chemical reaction on the surface of the plate takes a finite time to "soak" right into the plate and thus fully charge it.

To achieve this, the charger must maintain a constant voltage for a period of time to allow that absorption to take place completely (that's why it's called the absorption stage).

The first "Bulk" stage is a constant current one where the cell voltage will rise to about 2.4V in roughly 4 hours. The battery will now only be at about 70% SOC The second stage (absorption) is a constant voltage phase and continues until the cell reaches about 80% SOC - this can take another 6 or 7 hours whereupon the charger goes to "float" mode and the voltage tapers off to maintain the battery SOC.

The point that I am making here is that if you try to terminate the charge based solely on voltage rise at the terminals, you will end up with a battery that is far less than fully charged.

In the event of a faulty battery in the string as mentioned above, your good batteries would have been disconnected before they even got started.

Even if there were no defective cells and you were able to satisfactorily "bypass" a fully charged battery, what do you now do with the excess voltage that is being provided by the charger - for example a 24V charger across 4 x 6V batteries is fine, but now there are only 3 batteries on that 24V - your over voltage protections will now isolate all of those batteries with increasing rapidity and you will be nowhere near full charge.

You could introduce a "dummy" load into the circuit but, as the charge current in the circuit varies, so will the voltage drop across your dummy load - this will give false voltage indications to the charger, thus detrimentally affecting its ability to perform as designed.

Why does this matter?

1. LA batteries tend to sulphate and die young if not kept in a state of, or returned to, full charge as soon as possible after a discharge cycle.

2. Assume a 100Ah battery that you can safely (with regard to adequate cycle life) discharge to 50% SOC, so you can get 50Ah from it before needing to recharge. Now charge it to only 80% SOC, and you have only 30Ah available - but you will have even less available power because the lower terminal voltage will result in a higher current draw for the same power.

The commercial cell balancer that I linked to has no affect on the charger, it just compares the voltages of all of the cells and redistributes from one to the other as needed with no regard to the actual values. It won't protect against defective cells, you would need a fully fledged BMS for that but you are then outside the realms of simplicity and cheapness.

Thanks for the feedback and valuable insight on my questions.

Your points are well taken and seriously considered.

I have learned a lot from this forum.

I can purchase the battery balancer you linked to directly for around $35 USD, which is cheaper than I can even begin to consider building myself,using an Analog Devices LT 3305 IC and the associated external components required.

Thanks!

Lots of pallet jacks are charged in series (4 x 6V or 4 x 12V). In series, the charging current is equal for all 4 batteries.

That would be simple.

This old needle(noodle) is not a sharp as it used to be.Please illustrate a circuit for me.

Thanks!

Let me answer that with a question. At what voltage is a battery considered "good" while charging at a particular (specified) current and at what voltage is a battery considered "bad" with that identical current?

Sometimes you just have to get technical to answer a technical question.

The batteries in question are 6volt flooded cell lead acid batteries with a normal charge voltage of 7.4 volts.

There are 6 of these in series.

The current will vary on each battery,but the final voltage is the desired setpoint for the finished charge.

The battery is considered fully charged at 7.4 volts.

An ideal setup would be to monitor specific gravity on each cell,but I don't want to go that far.

A battery itself is a series of cells,and each cell is a little different,so the final output voltage is a rough estimate of the total charge of all the cells.

The constant current is the problem.A weak cell in one battery(low voltage) could increase the current to all batteries in series,thus overcharging some batteries.

And likewise,a battery that reaches full charge before the others will cause some batteries to be undercharged.

Difference in manufacturing tolerances and materials can also cause an imbalance in the charge requirements for the batteries.

Even an equalizing charge does not always balance all of the batteries to the same level.

Charging the batteries in parallel won't solve that problem (likely even worse), but charging them separately would work.

Battery voltage is highly dependent on battery chemistry. In parallel, as the cell voltage increases through it's charging voltage range, it will naturally reduce it's charge current preventing overcharge, assuming the charge voltage is within an acceptable range and each cell has some associated current limiting resistance. This works very well in parallel. Cells that reach charge before the others simply draw less charging current and wait for the other cells to come up.

In series, the same charge current must go through all cells. Therefore, those cells that near full charge first, raise their voltage, reducing the voltage on the ones less charged. If the charger voltage is just right, the less charged cells will not reach full charge, causing them to discharge early and possibly be exposed to reverse current. If the charger voltage is too high, the ones to first charge will over charge.

Thanks to everyone that contributed to this forum.

All feedback is sincerely appreciated.

I have seen single 2 volt cells for traction motors in heavy industrial equipment,and it allows individual cells to be balanced or replaced.

It would be nice to find smaller batteries(deep Cycle,etc.) with taps for each cell,and a integrated charge balancer on board the motor controller.I am not talking individual cells of a small size,but the the ability to access the cells in a typical deep cycle battery.

Of course it would require an individual wire to each battery,but that could be accomplished with a plug, and side mounted to minimize corrosion.

Imagine a 6volt battery with large current carrying conductors and smaller balancing wires to each cell.

Certainly the cost would be higher initially,but probably more economical in the long run.

500 charge cycles is the normal life cycle of a plate type deep cycle lead acid battery,and the concentric plate batteries are around 1500 cycles.(Depending on depth of charge,discharge,of course).

I believe that real time continuous cell balancing could improve this considerably,possibly as much as 33% or more.