balance charging battery pack method - will it work? SERIES

[Mechatrommer]

just buy this...
http://www.hobbyking.com/hobbyking/store/__64345__Turnigy_Accucel_6_80W_10A_Balancer_Charger_LiHV_Capable.html

if you are cheap arse, this...
http://www.hobbyking.com/hobbyking/store/__58319__Turnigy_B6_PRO_50W_6A_Balance_Charger_Blue_.html

they seems to share common FW interface like the clone balance charger that i own bought from china ebay. so i smell an OSHW leak for this, maybe this...
https://github.com/stawel/cheali-charger
but i dont know. i'm not interested in buiding it myself, maybe an opportunity to the OP.

disclaimer: i'm not affiliate of hobbyking, just an occasional customer. mostly i will buy from ebay china if i have better deal.

[user1235423643]

ok again... i dont want to use a external specific charger.


is it possible to use something like this with DIODES so not to short out?

[ajb]

No. 

[user1235423643]

why not?

[ajb]

Because in order to keep the chargers from shorting out you'd need a diode between the two batteries with cathode at the positive of the lower battery and anode at the negative of the upper battery.  In that orientation, the diode will also keep the batteries from discharging.  That's what sleemanj was telling you.

You could use an active switch between the two batteries, but that's not going to be a simple solution like you keep asking for.

[user1235423643]

hmm ok, so why cant the diode be put in line with the negative out of each charger which will still the back feed of + ?  or on all the + which ever

also option of using cheap dc to dc isolators?


http://electronics.stackexchange.com/questions/234086/new-layout-3cell-charger-with-tp4056

[ajb]

hmm ok, so why cant the diode be put in line with the negative out of each charger which will still the back feed of + ?  or on all the + which ever

There is no way to add diodes to the circuit to prevent shorting out either the battery or the charger without also preventing the use of the batteries to power a load.  The diodes you've drawn there will not even protect against shorting out the chargers, because they will all be forward biased when the chargers are powered on!

also option of using cheap dc to dc isolators?

Yes, as has been mentioned repeatedly, using isolated DC to DC converters or isolated wall warts would be the only way to make your proposal work.  However, if you go the isolated DC-DC route, be aware that all of the really cheap DC-DC converters are only good for a watt or two.  If you want to charge the battery at 1A, you will need at least a 5W converter.  This will maybe not be so cheap, by the time you buy one per cell.

[Siwastaja]

Your circuit is just perfect and will work perfectly, no need for diodes, just buy and connect everything together and you'll see! Your idea is so brilliant that I suggest you test it using as big cells as possible, for a great demo.

[user1235423643]

thanks can you please explain why the diode idea wont work?
what i have been told is the feedback of - and + on the common outputs will short but i dont understand how if diodes are used. thanks

hmm ok, so why cant the diode be put in line with the negative out of each charger which will still the back feed of + ?  or on all the + which ever

There is no way to add diodes to the circuit to prevent shorting out either the battery or the charger without also preventing the use of the batteries to power a load.  The diodes you've drawn there will not even protect against shorting out the chargers, because they will all be forward biased when the chargers are powered on!

also option of using cheap dc to dc isolators?

Yes, as has been mentioned repeatedly, using isolated DC to DC converters or isolated wall warts would be the only way to make your proposal work.  However, if you go the isolated DC-DC route, be aware that all of the really cheap DC-DC converters are only good for a watt or two.  If you want to charge the battery at 1A, you will need at least a 5W converter.  This will maybe not be so cheap, by the time you buy one per cell.

[sleemanj]

thanks can you please explain why the diode idea wont work?
what i have been told is the feedback of - and + on the common outputs will short but i dont understand how if diodes are used. thanks

Because the modules are not isolated, the IN- and OUT- are effectively directly connected so you will create short circuits, eg...



Actually that's only because you drew one of the diodes backwards (right most)

But even if you corrected that error, you can see without using isolated supplies, the charger board is still shorted on it's output, if the B+ has a route to the B- (even if it can't get to the battery because of your diode) then it is shorted out.



If you provided this circuit with isolated supplies (and didn't connect them on the supply side as you showed) it would be fine (as has already been said to you), if you don't, then you shorted out your supply, the boards, and at worst, the batteries.

Also how will the devices sense the cell voltage if they are behind diodes.  Answer: not well.

[user1235423643]

thanks for the reply. but how is it flowing back if the diode is blocking the +/- flowing into the -/+?
from what i was told the - or + of the battery would short the output of the charger

thanks can you please explain why the diode idea wont work?
what i have been told is the feedback of - and + on the common outputs will short but i dont understand how if diodes are used. thanks

Because the modules are not isolated, the IN- and OUT- are effectively directly connected so you will create short circuits, eg...



Actually that's only because you drew one of the diodes backwards (right most)

But even if you corrected that error, you can see without using isolated supplies, the charger board is still shorted on it's output, if the B+ has a route to the B- (even if it can't get to the battery because of your diode) then it is shorted out.



If you provided this circuit with isolated supplies (and didn't connect them on the supply side as you showed) it would be fine (as has already been said to you), if you don't, then you shorted out your supply, the boards, and at worst, the batteries.

Also how will the devices sense the cell voltage if they are behind diodes.  Answer: not well.

[sleemanj]

thanks for the reply. but how is it flowing back if the diode is blocking the +/- flowing into the -/+?
from what i was told the - or + of the battery would short the output of the charger

It has been explained to you multiple times, there is no isolation in to out,  there is no other way to explain this, the only way you can make your plan work is with some manner of isolated supplies for each board, if you try anything else at best it won't work, at worst something will  explode.  Be my guest to try, but at least do it outside so you still have somewhere to live after the fire.

[user1235423643]

ye ok dude, thats isnt answering why and what is going on for it not to work.

how can it short if the diodes are blocking the polarities mixing. if you cant explain or dont know how to answer, leave it for someone who can. thanks

thanks for the reply. but how is it flowing back if the diode is blocking the +/- flowing into the -/+?
from what i was told the - or + of the battery would short the output of the charger

It has been explained to you multiple times, there is no isolation in to out,  there is no other way to explain this, the only way you can make your plan work is with some manner of isolated supplies for each board, if you try anything else at best it won't work, at worst something will  explode.  Be my guest to try, but at least do it outside so you still have somewhere to live after the fire.

[cs.dk]

ye ok dude, thats isnt answering why and what is going on for it not to work.

Answer people like that won't help you..
That is drawn in the post right up above

[user1235423643]

right.... how does that explain how the + is flowing back up the - if there are diodes??? it should only allow flow in one direction?? so again if you cannot help explain dont bother to comment. thanks

ye ok dude, thats isnt answering why and what is going on for it not to work.

Answer people like that won't help you..
That is drawn in the post right up above

[CraigHB]

The attached drawing in reply number 13 shows how to do what you are wanting to do.  I'm surprised nobody emphasized it within the good number of replies afterward.

Basically you use a single cell charger on each cell with a virtual ground as shown in the drawing.  This is not the same method as used in the video posted in the original reply.  That method is used by hobby chargers which may charge at rates well above 2C.  Batteries used in powered models may have acceptable charge rates up to 10C.  Hobby chargers use a pack charge (only two wires) then during the final CVM phase of charging the charger balances cells as shown in the video.  There's a few reasons why hobby chargers balance this way, but none of them apply to your application.  The best way to charge series cells in your case is by using the method shown in the diagram of reply 13.

One thing to keep in mind is the power you have available to charge with.  You mentioned 5V a few times so I assume you're using a USB source.  Those are typically limited to an Amp with specialty chargers going as high as 2.1A such as those used for iPhones and Blackberries.  Since voltage has to be boosted somewhat above terminal battery voltage, charge rate will be reduced accordingly.  For example a typical limit for a USB port would be 5W.  At 20V you'd be limited to a 250mA charge rate neglecting losses.

[ajb]

The attached drawing in reply number 13 shows how to do what you are wanting to do.  I'm surprised nobody emphasized it within the good number of replies afterward.

Yes, conceptually that will work, but not with the TP4056 chargers the OP wants to use, which have an absolute maximum input voltage of 8V.  That isn't really even going to be enough to be able to do two cells in series.  For the overall system to not be horrendously inefficient you'd need to use some sort of switching charger topology. 

right.... how does that explain how the + is flowing back up the - if there are diodes??? it should only allow flow in one direction?? so again if you cannot help explain dont bother to comment. thanks

If you don't understand the problem from the diagrams and explanations that have already been provided, I suggest you try working through it yourself.  Start with the application schematic from the TP4056 datasheet, noting in particular that the negative terminal of the battery is connected to the charger's ground.  Place two of these schematics together, and connect them up the way you're proposing.  Now trace out where current flows through the circuit, starting from your positive supply.  Be sure you understand how diodes work so that you can identify whether or not current will flow through them.  (We'll leave aside the problem of placing diodes in series with the battery that was explained in your other thread for the moment.)  Post your work here, and I'm sure people will be willing to help you understand.  The forum best helps those who help themselves.

[user1235423643]

hello
what i am not understating is the issue is suppose to be the mixing of the + and - of the battery's  with the output of the chargers.
ie. backfeeding + into - and - into +
a diode is suppose to allow flow in one direction? 
if i am blocking the directing what is the problem then?


there must be a youtube video that can help explain this better to me?
thanks

The attached drawing in reply number 13 shows how to do what you are wanting to do.  I'm surprised nobody emphasized it within the good number of replies afterward.

Yes, conceptually that will work, but not with the TP4056 chargers the OP wants to use, which have an absolute maximum input voltage of 8V.  That isn't really even going to be enough to be able to do two cells in series.  For the overall system to not be horrendously inefficient you'd need to use some sort of switching charger topology. 

right.... how does that explain how the + is flowing back up the - if there are diodes??? it should only allow flow in one direction?? so again if you cannot help explain dont bother to comment. thanks

If you don't understand the problem from the diagrams and explanations that have already been provided, I suggest you try working through it yourself.  Start with the application schematic from the TP4056 datasheet, noting in particular that the negative terminal of the battery is connected to the charger's ground.  Place two of these schematics together, and connect them up the way you're proposing.  Now trace out where current flows through the circuit, starting from your positive supply.  Be sure you understand how diodes work so that you can identify whether or not current will flow through them.  (We'll leave aside the problem of placing diodes in series with the battery that was explained in your other thread for the moment.)  Post your work here, and I'm sure people will be willing to help you understand.  The forum best helps those who help themselves.

[user1235423643]

thanks for the info, im not looking to run off 5V usb either a power suply.  also looking to run of a boost\buck so i can charge from any good supply voltage,
do yo have any more info about the image in post 13?

basically i want a very cheap way to charge mixed 18650s in series so they are balanced and the charge time is not wasted on discharge \ balance. thanks

The attached drawing in reply number 13 shows how to do what you are wanting to do.  I'm surprised nobody emphasized it within the good number of replies afterward.

Basically you use a single cell charger on each cell with a virtual ground as shown in the drawing.  This is not the same method as used in the video posted in the original reply.  That method is used by hobby chargers which may charge at rates well above 2C.  Batteries used in powered models may have acceptable charge rates up to 10C.  Hobby chargers use a pack charge (only two wires) then during the final CVM phase of charging the charger balances cells as shown in the video.  There's a few reasons why hobby chargers balance this way, but none of them apply to your application.  The best way to charge series cells in your case is by using the method shown in the diagram of reply 13.

One thing to keep in mind is the power you have available to charge with.  You mentioned 5V a few times so I assume you're using a USB source.  Those are typically limited to an Amp with specialty chargers going as high as 2.1A such as those used for iPhones and Blackberries.  Since voltage has to be boosted somewhat above terminal battery voltage, charge rate will be reduced accordingly.  For example a typical limit for a USB port would be 5W.  At 20V you'd be limited to a 250mA charge rate neglecting losses.

[ajb]

hello
what i am not understating is the issue is suppose to be the mixing of the + and - of the battery's  with the output of the chargers.
ie. backfeeding + into - and - into +
a diode is suppose to allow flow in one direction? 
if i am blocking the directing what is the problem then?

You're not blocking the problematic current with the diodes oriented as you've drawn.  If you oriented the diodes to block the problem current, you would also block the charging current.  Draw out the circuit and see for yourself.  You're also continuing to ignore the fact that diodes between the batteries and charger are a bad idea, as pointed out in your other thread.

basically i want a very cheap way to charge mixed 18650s in series so they are balanced and the charge time is not wasted on discharge \ balance. thanks

Balancing doesn't waste any time at all.  It doesn't matter if you charge each cell separately at 1A or the entire pack at 1A, the time it takes the whole pack to finish charging will be determined by the time it takes the last cell to finish charging.  For all of this effort, you will not save any time.

[CraigHB]

im not looking to run off 5V usb either a power suply.

That makes things a lot easier then since you have a range of voltages and currents available limited only by your power supply.  You can use whatever rate you want at whatever voltage is required.  Though there is the limitation of the charger controller chip you use which is typically an Amp for the linear type.

There is the option of using a buck type charger controller which can support higher currents.  Usually when charging 18650s a 1 Amp rate is adequate though it will only be about 1/3C for a typical 18650 cell.  The charger controller you mentioned is the linear type which is limited by the power dissipation of the chip, that being the difference between input voltage and battery voltage times rate.  From my experience a linear charger controller having a thermal pad soldered to a ground plane tops out about an Amp when driven with a 5V supply.

It's a pretty simple thing to put charger controllers in a ladder to charge a series pack of Li-Ion cells.  I've not actually built a charger like that so I can't speak from experience.  I've done a number of single cell chargers though.  The poster above stated that voltage tolerance for each chip in the ladder must be greater than input voltage, but I don't believe that's the case.  The ground for each chip is referenced to the plus minus connection of the cell in the battery it's charging so it's never going to see full input voltage. 

The simple linear controllers you mentioned should work in a ladder.  If there's any doubt about voltage tolerance you can always put a charger front end in front of each chip such as the TI BQ24316DSJ.  Or you can use a charger controller with a higher voltage tolerance.  There's also the option of a buck type charger controller to support higher rates.

[ajb]

The poster above stated that voltage tolerance for each chip in the ladder must be greater than input voltage, but I don't believe that's the case.  The ground for each chip is referenced to the plus minus connection of the cell in the battery it's charging so it's never going to see full input voltage. 

Look at the diagram in post #13 again.  The bottom charger in the ladder sees the full 20V input.  You could do something like have a different step-down converter for each 'rung' of the ladder such that each charger only saw 5V or whatever at its input, but unless all of those converters are setup to track the battery voltage, the input voltage to each charger will be larger when the pack is discharged, and a pack with many cells in series could wind up violating the max input voltage of one of the middle-ish chargers. 

But there's actually another, more insidious caveat with that topology, which is that the charging current for any of the upper batteries must flow through all of the batteries below it.  That means that if one of the lower cells reaches full charge first, the upper charger will wind up overcharging the lower cell as it attempts to finish charging its cell.  The only way to avoid that is to use two-quadrant chargers that can sink current as well as source it to prevent overcharging or--surprise surprise--use an additional balancing circuit.  The LT part linked earlier in the thread would work, since it's two-quadrant.  It's also a switching topology, so it would be more efficient as well.  The TP4056 is NOT two-quadrant, so on its own would not be a good choice for this topology, even if its max input voltage weren't a problem.  The overcharging problem is compounded by the fact that, since the charging current through any battery will be the sum of the charging currents supplied by all of the chargers at or above its level in the ladder, if all the cells are similar the lowest cells will always finish charging before the upper cells--if they don't die from excessive charging current first.  Again, you could avoid all of these problems by using a separate isolated supply for each charger+cell, or by using a single charger for the overall pack plus a balancer.

[user1235423643]

"Balancing doesn't waste any time at all.  It doesn't matter if you charge each cell separately at 1A or the entire pack at 1A, the time it takes the whole pack to finish charging will be determined by the time it takes the last cell to finish charging.  For all of this effort, you will not save any time."


that make no sense at all....
did i understand the video wrong??
https://youtu.be/wIbHLacozFo?t=368
where he explains the balance charger will charge in parallel and once a cell reaches full, it STOPS and discharges them.  (there for not charging and wasting time discharging)

and when charging a series pack in parallel is each sell getting - 1000mah / 3 = 333mah  or each cell is getting 1000mah l (max) ??

again the point is to find a way to charge a mix 18650 that may not be the same condition, quickly with out wasting time to discharge.
even a cleaver way to put a switch between each cell to isolate them while charging is a more simple option.


the battery packs will be running boost converters anyway and using recycled battery's

forget the diagram #13 is was just a example i found, i am looking for a solution's not an excuse. thanks