Charge-in-place system for light project

Hello,

I have an idea for a more ambitious mod than the little tweaks I've done so far, but first I need to get things straight so I'll ask the questions in their relevant threads. The final project will probably be a Mag3D/4D with multiple MC-E leds (3?) but I want each led to be powered by it's own battery pack for maximum runtime -- in essence to have 3x3 18650 for example, each 3x pack powering one led.

One big problem with a setup like that however is that the large number of batteries become a chore to take out and place back in -- that's why it would need a "charge in place" system, something like a ballancing laptop brick that can charge several 18650's at the same time and keep them all balanced.

Is something like that feasible? I am prepared to mod existing stuff and don't mind spending a little more money (within limits ) but I wanted to ask here first about that, maybe someone else thought about it before and may have some pointers. If it's been discussed somewhere else please point me there, I searched the forum before but to no avail...

Thanks in advance!

not much difference between

|{======]{======]{======]---|>---@----|
|______________________________________|

|{======]{======]{======]---|>---@----|
|______________________________________|

|{======]{======]{======]---|>---@----|
|______________________________________|

and
|..............|..............|...............| (battery charge taps)
|{======]|{======]|{======]-|--|>---@----|
|{======]|{======]|{======]-|--|>---@----|
|{======]|{======]|{======]-|--|>---@----|
|_________________________________________|

with <| being the led and @ being the driver thing, and of course connecting the driver as needed (not shown).

the reason i mention this Parelel setup is it would be easier to charge 3X parellel/series set , then 3 seperate series sets, that will be at different levels.
this way you can concentrate on charging each trio and balancing the parellel series sets, and not with 9 freaking batteries at differnet levels.

this still allows you to have 3 seperate drivers running 3 seperate leds, even at different drive levels. and notice I didnt connect the 3 leds as parellel so they are each using a seperate driver, then depending on the driver , they have a negative or postive "ground" or connection that just goes right on through.

it is possible using common care, to parellel and parellel charge li-ion batteries when charging with a voltage max, with ni-mh and using a v-drop charging this wouldnt work, but li-ion charges to a maximum voltage, so the parelel sets would remain mostly the same voltage charging and discharging.

so you can parellel, then you just make sure you balance the series on charge, which will become even more important, especially after many cycles and time.

so if you use this method, you have 3 parellel sets, 4 wires needed to fully balance charge.
then you have to apply proper maintance of the "pack" , insuring that your not charging ruined cells, overheating, putting under pressure, and other li-ion safety things, so the pack is not Permanent , it would have to be checked to insure it it not dead from age , and stuff.
it wouldnt hurt to toss a few thermal fuses in there too, beings you will have the space to do that.

Gosh Vidpro, excellent answers and info! I salute your battery and circuit knowledge. Thanks for sharing.

Stuff like this is why this forum is a hit!

Jeff O.

Hi, sorry for not replying sooner, I was out of town for a few days and only had little time to check up on ye old Net addictions...

Anyway, thanks VidPro for the informative reply, I think I understand the gist of your proposal -- basically to drive the leds with cells in 3S but charge each of those separately in 3P with ones from the other leds? Sounds like an inteteresting idea, however I worry about inbalance, as you've hinted as well. I think it would be easy to have one cell get discharged sooner and somewhat cumbersome to keep checking them up?

In my original post I mentioned laptop chargers, but then I realized that the laptop "brick" is basically just a 12V adapter, and all the smarts of charging are actually embedded in the battery pack (hence the high price). I must say my original inspiration for this mod came after contemplating my laptop's pack, enlightened with the knowledge that it must contain 9x 18650 cells. They are arranged in a triangle shape, and the output is 11.1V/85Wh which translates to around 2500mAh per cell in a 3S3P configuration. So I figured -- why not have such a system for a flashlight? It provides ample power and is only about the size of a 3D.

So after reading your suggestions and giving this some more thought, I believe there are basically 3 options:

1. Have the batteries + protection circuit inside the light, with +/- connectors like the ones described in the above drawings, and a single external 3S charger. The charger would basically "see" each 3P pack as one battery in series and charge it accordingly, and -- at least in theory -- the batteries should ballance each-other within a pack. But it would still introduce slight variations in time that would need to be checked and corrected.

2. Take a charger PCB and mount it inside the light, or build the PCB if necessary. This would present the advantage of only having to connect the light to an external 12V source, just like a laptop, and the rest would work just like in case 1. However, there is still the problem with possible unbalances, and this solution is also not flexible at all -- what if one wants to add one more led + driver + battery pack? It would require a whole different PCB, plus charging more than 4S is becoming increasingly difficult, with 5S chargers rare and expensive. 3S ones on the other hand seem to be readily available and cheap...

3. So here goes the final option -- more expensive, more complicated and more space-consuming, but offering maximum flexibility: have 3 charger PCB's mounted inside the light, one for each 3x pack, and then each pack can be a 3S again, basically a full independent light with everything included (charger, battery pack, driver, led) and only incased in the same body. Of course this means obtaining 3 charger PCB's and it would require more space inside the light (so probably the Mag body won't fit and a custom one will need to be built), but to me this seems to be the best option. I would actually like to have the ability of adding led's to the light as time comes, maybe have 3xMCE and 2xR2 like some monsters I've seen here, which this design would allow easily.

Of course I may be completely off base about this, and if so please correct me! As of now, I will revise my project in order to start with the smallest possible combination and grow from there:
- get a small Mag, probably 2D/3D since the body will need to be custom anyway and I would only use the head
- build a body out of PVC pipe (I've seen some lights here that use that, and it is within my "modding" skills) and get hold of a proper heatsink
- get 1 MC-E + driver, 3x18650 (protected) and gut a 3S charger, connect these and place inside the light.

If all this works properly, I'll then move on to doing the same x3 times in order to obtain the original idea

What do you guys think?

ok.
there is a load of space available down the center of the round cells, when in a round tube, i have used that space before to put the protection in. potentially one could even use that space for a charge curcuit, but the problem of most charge curcuits outputting heat :-(.


only thing i dont understand is what the difficulty of a 4 wire power connector on the back of a light, and any cheap RC balance charger being able to balance it. sure its not AS easy, but then you get readouts on all three batteries, so over TIME, you can see what is occuring inside, without 4 wires, you cannot SEE inside. if you can see what is happening inside, then you gotta disasemble it again :-(

VidPro said:

only thing i dont understand is what the difficulty of a 4 wire power connector on the back of a light, and any cheap RC balance charger being able to balance it. sure its not AS easy, but then you get readouts on all three batteries, so over TIME, you can see what is occuring inside, without 4 wires, you cannot SEE inside. if you can see what is happening inside, then you gotta disasemble it again :-(

Click to expand...

Yes, I was thinking about that as well, the need to have some leds or something outside to be able to see the charging status. And the issue of heat you brought up is another problem :-/ hmm...

Well then, going back to your proposal, I agree it is not that difficult to simply have a balancing connector on the light together with the power connector, it's just more cumbersome (those tiny pins are finicky and can easily get bent, I've had it happen with a RC charger). So if we go that route... how safe is charging 3P instead of one battery on each lead of the balancer? I've checked the instructions on one right now and it clearly states to not do that, but of course they could be overly cautious... From what I read the nP batteries would balance each-other when charged, but what happens over time? Now that I think about it, with one of those 2x5S charger/balancers one could charge 2x5s3p = 30 cells at once? I mean it would rock to be so, even if it takes a while

to get rid of those flumsy connectors, just use an XLR-connector. They are available up to 7-pin, this should do. You can just mount one in the cap of a Mag D... just as is done in audio (microphone) booms....
If this is too big, use a Lemo connector, for example one from the 2S-series. Rugged, small, at least 10-pole, and able to withstand some current

Timmo.

mgc8 said:

how safe is charging 3P instead of one battery on each lead of the balancer? I've checked the instructions on one right now and it clearly states to not do that,

Click to expand...

sure dont do that, WHEN charging ni-cd or ni-mh USING a V-drop alogrythm (the usual one they use) but a li-ion proper alogrythm for charging is a voltage maximum, so all 3 of the cells will be at the same Voltage, WHERAS with ni-mh they pulverise the battery till it is in an overcharge state, which causes a voltage drop, then terminate when the computer sees this voltage drop, with 3 cells reaching end of charge at different times the computer cant (properly/easily) detect the v-drop, it will not properly terminate.

same charger charging, but they (usually) use 2 completly different charging styles (beyond the voltages) for the li-ion vrses the ni-?? chemistries.

you could say they are kinder to li-ion by it not reaching an overcharge state, and ramping up to a set voltage max allows the parelell Trio to reach a similar VOLTAGE.

note: that does not (definatly) define its actual capacity or charge state at that voltage, it just the max it will be at, and it wont DROP at the end of the charge, so the trio will Stay the same voltages, so there will be minimal cross charging or abrupt changes or wild differences, making the voltage max charging work good for parellel. its all about the charge alogrythm chosen, , , and the way they expect that chemistry to be charged.

and on the other hand, prefer not to SERIES charge li-ion without balancing, because one cell could be very low, or doesnt hold or take a charge, and the other cells could therin be overcharged.

hmm, what we need is pictures again.

bad examples:

li-ion with BAD cell
|{==2v=Bad==]{===6.3v===]{===4.3v===]|
|_dum charger that thinks 12.6v will work___|

say long into time, this one cell has gone all ultrafire on us and wont take a charge , the brainless charger only knows 4.2v Per cell, and doesnt have any idea what is happening. as one cell is overcharged (exagerated voltage) and has probably gassed, because one cell will no longer hold a charge the bad one, the OTHER one vents.


li-ion still with bad cell, but charger is tapped, balance charge

|{==2v=Bad==]|{===4.2v===]|{===4.2v===]|
|_____________|____________|____________|

the tapped balance charger , is checking the voltage independantally, so while the bad cell still exists, the other cells are still checked to see if they are overcharging. Even if the charger is to stupid to notice that the first cell isnt taking a charge, and is heating up, the other 2 cells are not overcharged, those 2 cells do not vent.
still a bad situation depending on the rate of charge and heat and all, but no overcharge, and readings on the machine would show something going wrong, and probably the machine would say something besides, at least when it could not balance it. still the bad cell can overheat.


in parellel, it will never look like this

|{==2v=Bad==]|
|{===4.2v===]|
|{===4.2v===]|
|_4.2vcharge__|

this isnt going to happen, all the cells will be pretty close to the same voltage levels. if one is bad, it will bring the WHOLE parelell portion down with it. (been there, done that, seen that)
IF a person was trying to charge 3X 1C for the parellel set (in this case) they would likly start the bad cell on FIRE, because of its voltage its getting all the charge. so again charge rate is STILL important.

in parellel it looks like about like this:

|{==2v=Bad==]|
|{===2.1v===]|
|{===2.1v===]|
|_4.2vcharge__|

the other 2 cells are trying to charge, and are capable of charging, but the bad cell is just blowing the power off in internal resistace between the plates, and other bad stuff. the bad cell takes ALL of the power of the charge going to all 3, but doesnt charge. because the other 2 are physically connected they will just be parked, or being drawn down by the bad one. the bad one is neither taking the charge, nor holding it. likly long term scenario all 3 cells will be discharged to low and die also.

now combine that with this and those, and i dont know if it makes a picture anymore

|{==2v=Bad==]{===6.3v===]{===4.3v===]|
|{==2v=good=]{===6.3v===]{===4.3v===]|
|{==2v=good=]{===6.3v===]{===4.3v===]|
|_dum charger that thinks 12.6v will work___|

this still wont work, obviously, and now it has 2 problems one bad cell heating up and 3 cells (or even 6) venting from overcharge.


|{==2v=Bad==]|{===4.2v===]|{===4.2v===]|
|{==2v=good=]|{===4.2v===]|{===4.2v===]|
|{==2v=good=]|{===4.2v===]|{===4.2v===]|
|_____________|____________|____________|
balance charger

this still has the one problem, and that being if the user thinks the 1C rate as ok, then Trippled that for the three cells, your still pounding the one bad cell with a lot of heat.


of course what we really want to see at the end of the charge all the time is more like this
|{===4.2v===]||{===4.2v===]|{===4.2v===]|
|{===4.2v===]||{===4.2v===]|{===4.2v===]|
|{===4.2v===]||{===4.2v===]|{===4.2v===]|
|_____________|____________|____________|
balance charger

and of course these examples are HIGHLY exagerated, and the effects of Protection are not shown.
but still unprotected:
with a brainless series charge a slight imbalance in the series WILL still toast the other cells, the series mates only have to go to 4.4-4.5 and they are gone .
a minor capacity difference in the parellel cells, they will just work a bit unevenly, but EACH parellel packs voltage will remain about the same, they are connected
a minor capacity difference in the series cells, will mean the voltages will vary on charge and discharge, the balance charge being the thing that reset that each charge.

so now i gotta toss protection on each cell (or parellel set), or have "pack" protections.

li-ion totally wacked out of balance bad cell

|{==2v=Bad==]{===4.2v===]{===4.2v===]|
|_dum charger that thinks 12.6v will work___|

ahhhhh, now the protection will disallow the overcharging of the cells, they can be out of balance, but the protection will still keep overcharge from occuring, the cells will not vent. the bad cell will still be heating up depending (again) on charge rate.


second scenario as above, but with balance charge
PROTECED li-ion Still wacked out of balance, but charger is tapped, balance charge

|{==2v=Bad==]|{===4.2v===]|{===4.2v===]|
|_____________|____________|____________|

as long as this cell is BAD, not just off in capacity/voltage this pack still aint going anywhere. that is why i hate cheap junkey cells that dont last time.

so now your saying, What do i need BALANCE for, if protection is enough?
well the above examples were of a BAD cell, a terrible cell, a dead cell, one aged to long, a cheap piece of junk.
i should have discussed protection first, then balancing, then bad cell.

above i was displaying a BAD cell, one that is completly dead or ruined.

so now i back up a bit, to minor offsets that balancing fixes.

|{===4.1v===]{===4.25v===]{===4.25v===]|
|___dum charger that thinks 12.6v will work___|

even protected, this series example can occur, balancing taps balancing charger would have charged these in similar voltage (even if the capacity is different).

|{===3.9v===]{===4.20v===]{===4.20v===]|
|___dum charger that thinks 12.6v will work___|

even protected, this series example can also occur, even if the protection cuts at 4.20v , the protection CUT-OFF the whole series set, one of the batteries was not finished charging at all.
all the protection knows is if one cell went under or over the voltage, once that occurs the protection OPENS on one cell, once the total "curcuit" is open, the chargeing will stop.


this unbalanced pack (even protected) being discharged

|{===2.5v===]{===3.6v===]{===3.6v===]|
|___Discharge load of some kind___________|

because the unbalanced pack cells were not fully charged, the discharge ends unbalanced too, in the above the protection will cut-off at the cut-off point, because the cells capacity (not fully charged) was exhaused before the other cells.

SERIES balancing during the charging would "fix" that, even if there are slight capacity variations, each full recharge will reset them all to the same voltage.
|{===4.2v===]||{===4.2v===]|{===4.2v===]|
|_____________|____________|____________|
balance charger

at each discharge there will still be variations in the voltages, especially neer the end, protected or not balanced or not, when the capacity runs out it on the series set, they are not going to be balanced, that is re-done at each balance charge, so you Start out with them even and full charged.

then parellel the cells, and you have basically the same thing, with more capacity, as long as the end of charge termination is a max voltage, the parellel cells are connected, and good, the parellel will work just like a single cell with more capacity.

ok, now Ni-Mh parellel charging, gosh i need an editor badly.

3 Perfectally good ni-mh cells in parellel.

|{=====1.48v====]|
|{=====1.48v====]|
|{=====1.48v====]|
|_V-drop. charge___|
|_Voltage 1.48V____| <-- computer sees voltage going up

in this scenario, these ni-mh cells are pretty much charged up, they are about to reach the overcharge state, where they will V-drop.

|{=====1.47v====]| <-- this cell v-droped as it reached overcharge state
|{=====1.49v====]|
|{=====1.49v====]|
|_V-drop. charge___|
|_Voltage 1.485V___| <-- computer Still sees voltage going up

oops TO BAD, the stupid computer doesnt SEE the V-drop occuring, because all it sees is the Packs whole voltage, and the voltage (depending on the charge rate again) still was going up.
OH, OH, seems that one is fully charged, and V-dropping, but the other 2 arent charged yet, so the DEPENDANCY the computer has on a visable v-drop is not seen.

|{=====1.46v====]| <-- this cell v-droped as it reached overcharge state
|{=====1.47v====]| <-- this cell v-droped as it reached overcharge state
|{=====1.49v====]|
|_V-drop. charge___|
|_Voltage 1.48V____| <-- computer finnally sees V-drop

ok so now the second battery V-dropped finnaly, and the computer sees the indicator for stopping the charge. in the Mean time, the first battery was overcharged a lot.

depending on the rate, the number of batts, in series and parellel, and the sensitivity of the charger and all, i have seen severe V-drops and large overcharges because the computer looking for a CHANGE, doesnt see that change because there is more than one cell there.
sometimes, the other cells are still going UP, when one is going down.


Series
One still going up and the other finished and going down:
|{=====1.49v====]|{=====1.49v====]|{=====1.48v====]|
|________poor charger trying to see voltage go down_______|
|____________voltage 4.45_____________________________|

..charged v-dropped
|{=====1.48v====]|{=====1.49v====]|{=====1.49v====]|
|________poor charger trying to see voltage go down_______|
|____________voltage 4.45_____________________________|

this Same thing happens in Series TOO with ni-mh chargers , if they would stop or slow down at a set VOLTAGE then finish up by topping off slowly.
or if a slow , never overcharges rate is used , or a combo of both of those styles is used, then an invisable v-drop doesnt mater.

this is a huge problem with V-drop charging as used for ni-mh and ni-cd.
with a multiple series parellel set like 12X there can be a voltage difference of a whole volt or more.

parellel & series ni-mh big problem
|--------Series set V-dropped hard 14.3V-------| This is overcharged V-dropped from 14.9V long ago
|--------Series set still charging up 15.1V-------| this is now sending power into that.

above ^ the .8v differential has not only the charger hitting the overcharged battery, but the 15.1v series set, is now pumping power into the set that is already overcharged, making the situation worse still.
IF we had stopped the charge at ~14.5v (or so) neither parellel set would have v-dropped . instead the overcharged (v-drop) series set, is still being overcharged, and its parellel teammate is pushing power into it ALSO.

So as you can see with ni-mh PARELLEL charging, the end of charge V-drop is not seen easily, and BECAUSE of the v-drop the differential between parellel cells gets worse and worse.

with li-ion, they dont do V-drop, they stop at a specific voltage, and therfore the huge differences , in both the charger, and how they charge the chemistry.
hmm wouldnt it have been easier to say : Ni-?? charging voltages DROP on end of charge, li-ion is not allowed to reach that state or condition, so the variation at end of charge doesnt occur normally.


with that many pictures, mabey a video of it occuring would be better gee who could make a video , i wonder.

VidPro, what can I say, other than :bow:
One can get that info by browsing around here or various RC forums, but you've done a great job putting everything together in a concise and very useful way! Thank you!

Maybe the last few posts could be copied to a sticky thread in this part of the forum? I'm sure they would help anyone seeking answers in the wonderful (yet sometimes puzzling and even dangerous) world of batteries and chargers... As for me, I've got my work cut-out, at least as it comes to the power requirements for my little project