Project: Replace 18650 cells in laptop battery (w/pictures)

Izits

100 W
Joined
Jan 1, 2017
Messages
152
Location
United States of America.
My laptop uses a standard 6-cell Dell T54FJ battery. Looking for a replacement, I was astounded to find a new one costs $140. Thanks to what I learned at Endless-sphere.com, I replaced the cells instead of buying a new battery.

Equally dismaying was the fact that my old Dell battery only uses cells with 2.7Ah capacity. I had already been buying 3.5Ah cells for flashlights so the solution seemed obvious. (new Dell batteries now use 2.93Ah cells)

Genuine Dell battery price: $140
Genuine Dell battery capacity: 5.59Ah

x7 Panasonic NCR18650GA cells: $40.75 (+one extra)
Final pack capacity: 7.18Ah

These are the stock cells. Notice the sheet metal tabs between each pair of cells for balance charging.
1kNBcHT.jpg

T9FbeQb.jpg




I don't have a spot welder and didn't really want to buy one. So I opted to solder my new cells together. The trick was to solder wires to both ends of every cell without making the cells sit too far away from each other. There is a little leeway in the plastic case for length but not much. The only available space between the cells was where the button wasn't. So I was careful to solder the wire on the perimeter of the cells.

DtLwkK0.jpg

HL3RZUz.jpg

G1J2g9B.jpg

nxeN6xE.jpg

5WuZqRq.jpg




I reinstalled the foam insulation spacer between the cell sets and added some special protection anywhere I thought there was the remotest chance of a short being created through the shrink wrap of the cells. Finally, I soldered the balance charging tails back onto the circuit board.

7TxPnRw.jpg

b0vKnKY.jpg

TRhB48m.jpg




Here is the finished case after being snapped back together. I didn't mention what an ordeal it was to get the two halves apart in the first place. Just keep prying away at it, they do eventually come apart and with surprisingly little damage. All said and done the battery worked like a charm.

I doubt this project would have occurred to me were it not for the great sharing of knowledge here at Endless-Sphere. My humble thanks to everyone.

2pgKuzY.jpg
 
I spend a lot of time on my laptop, and when the original battery died, I happened to see in some Amazon listing a replacement that had 12 cells instead of 6. Also, the cells used are spec'd on cheapest price, not longest run-time. I have purchased 12 of the newer 3500-mAh cells to replace the stock 2000-mAh cells, so I will be expecting roughly 3 times the stock run-time afterwards.

Thanks for posting this. The replacement battery-pack sellers charge a fair price, but...they don't spec name-brand cells, and they don't use the max-range cells as an option. 12 new bare cells from Samsung/Panasonic/etc will cost the same as a 12-cell replacement pack using generic 2000-mAh cells.
 
Nice work

Could you post your laptop ad battery models? I've already purchased the cells, just have to finish my spot welder to replace them, but I've read somewhere that some BMS's can't be disconnected and reconnected or they enter a protection mode that bricks it, if we have the same model it would be one thing less to worry. My plan was to use some old 18650 that i have to keep the BMS alive while i replace the cells
 
Dell T54FJ battery. I had the circuit board disconnected for at least a week. Never heard of a bricking problem. It sounds like an invented boogyman to discourage people from servicing their packs instead of buying new ones.

Pedrodemio said:
Nice work

Could you post your laptop ad battery models? I've already purchased the cells, just have to finish my spot welder to replace them, but I've read somewhere that some BMS's can't be disconnected and reconnected or they enter a protection mode that bricks it, if we have the same model it would be one thing less to worry. My plan was to use some old 18650 that i have to keep the BMS alive while i replace the cells
 
Any recent experiences here?..

Just put new Samsung cells in my 8858x Dell laptop battery and no output from the BMS, all cells are 4.0v with 12v at the Tails.

I have also read, if BMS is disconnected a internal fuse renders the board useless (bricked) please someone tell me this isn't true?.. spent a few hours now messing around trying to get a reading at the pins...

Also tried a few reset YouTube tricks to no avail
 
well, it wouldn't be a "fuse", but it can easily be firmware in the bms chip(s).

some oem ebike batteries do this too, so that once power is lost to the bms chip it resets to some default state that will not operate normally until it is "reprogrammed" or enabled by some external hardware/software connection intended to only be available at the factory.

sometimes people get around this by recelling the packs in a process that never disconnects any old cells until the new ones are installed, making sure all new ones are at the same voltages as teh old ones during the process.


but if any old cells were actually dead, unsafe to recharge, the bms could disable itself permanently (at least until the factory resets it) to prevent fires, and wont' renable itself just because the cell now reads higher than minimum safe voltage.
 
Thanks for the reply Amberwolf and confirming my findings.

Crazy we're locked out in replacing quality cells for ourselves, I've built a few packs in the past,ebike, drill batteries etc. Would be nice to have the option. God knows what I'll be buying online now, hesitant to buy genuine Dell?? Be interesting to know if none OEM packs have a more forgiving PCM/BMS?..

I'm guessing Dell are protecting there interest's and happy to kill off batteries at 600 cycles, I'd probably do the same if I was selling millions and millions off these packs all over the world.

Oh well, it was a quick fun build, some you win some you lose, and Dell got me on this one.

Thanks again Amberwolf for putting the time in
 
it's also possible that the bms just died from esd or whatever, of course.

no way to know if any particular bms will lock you out until you disconnect the cells, so you are safer replacing any cells in any pack by parallleing the new ones before removing the old ones. ;)
 
it's relatively simple, as long as none of the cell (groups) are dead or overdischarged in the old pack. if they are you can't do this.

the only non-simple part is that sometimes a bms location or or pack size / shape leaves no room for extra wire lengths, which you have to have to do this.

additonaly if there are sensors inside the old cell block that you have to leave in the pack you will have to remove them form the block without disconnecting them from the bms as you remove the old blcok in a later step below, and reinstall them in hte new block again without disconnecting them from the bms.

bascially you just need to ensure no old pack wires are ever disconnected from the bms at any time, until the new pack si completely installed.

physically build the new cell block just like it needs to be for the pack size/shape/etc.

charge (or discharge) the new cell groups to match the old cell groups.

connect the main negative of the new cell block to the main negative of the bms or pack, where the old cell block's main negative is conected to.

connect the next most positive cell group wire of the new cell block to the bms connection point of it's equivalent on the old one.

repeat until you reach the main positive of the new cell block, and connect that to the bms connection of the main positive of the old one.

now all cells are connected, and the bms has no idea that they are there.

begin cutting the old cell block away from the bms / pack connections, ensureing you do not disconnect anything from th enew pack.

phsyically remove the old cell block

test to be sure the bms is allowing charge/discharge as normal (this may not be possible on some packs iwhtout first doing the next steps)

phsyically intall the new cel block, routing the (longer) wires as needed.

reassemble the casing.

use as normal.



in some laptop packs, and perhaps fairly advanced packs in other devices, it's still possbile that the bms has been keeping track of cell capacities, and may actually shutdown pack output based on counting charge output rather than a cell-level lvc, in which case it will still only have the same capacity it last had with the old pack...if that is the case there's not much you can do about it. the good news is it won't keep dropping, it'll stay at that capacity until the new cells age enough to drop below that.

in laptops, some of them have a battery-retest function of some kind, like an old gateway i have arund here somewhere. if they do, it may be able to test with the new cells and see what capacity they have, and then afterward be able to use that new greater capacity.


i don't expect this would be a problem in most packs, prbably none of the ebike packs i've heard of so far, but i don't have direct expereince doing any of this, i'm just interpolating data others have posted here and elsewhere. ;)
 
Wow! Valuable information for future laptop battery users, wish I'd of known this a couple of days before I stripped my pack.

Thank you very much sir :thumb:

I have ordered a new genuine Dell pack NOS 60w instead of standard 48w so hopefully will last.

Absolute shame about this BMS and others using chemical fuse, and count technology almost renders them impossible or near on useless to replace the cells.

Hopefully others will read and solutions will be posted, as I can't believe there's not many threads about this.

Thanks again Amberwolf for your quality time in posting, and clearing up what seams to be a little allusive subject, cheers
 
mxer said:
Absolute shame about this BMS and others using chemical fuse,
not sure what you mean by "chemical fuse", as i'm not aware of anything like that in there. i'm sure they do have a normal fuse in them someplace in case of shorts, etc., to prevent fires.

there are fuses that are auto-resetting instead of just wires that melt but they make a fair bit of heat at useful currents, so have to be quite large and thus are rarely used in compact power sources like laptop batteries (i have seen one in a picture of a bms for ebike batteries, probably on its charging port); theyr'e usually used for things like usb port power protection, since that's only 500ma.


if you're referring to the bmses taht brick themselves once cells drop too low for safe reuse, that's not a fuse, just a software/firmware thing. its' actually a safety feature, since it is possible for an overdischarged cell to eventually go back up in voltage to above the safe level over a long enough time with no load on it, or if temperature rises enough, but it would still have any damage that occured from the overdischarge, and still be unsafe to recharge if so. so the bms is designed to prevent that from happening, to keep a fire from occuring.

almost no one using such packs would ever do what we do here, and recell them for various reasons, so it's a perfectly valid and reasonable thing to disable the pack completely under those conditions. i'd guess that less than one in a hundred thousand, maybe even one in a million, laptop users would ever even open their packs for any reason at all, much less recell them or repair them.


same thing with the oem ebike packs taht do this, though there is probably a much higher percentage of users (though likely still far less than one in ten thousand) that repair them on their own, rather than just replacing them, mostly simply because of cost (hundreds or thousands of dollars, vs a few dozen dollars for a laptop pack).




the coloumb-count technology has its' reasons, too, as it's a more accurate way to keep track of cell health, especially when cells are only 1 or 2p. this means a bms can monitor voltage of the lowest cell during discharge, and also count the total pack ah as it is discharged, so that it can report this capacity to the device it's part of, so that device much more accurately knows how much runtime is available for various discharge rates, and can accurately tell the user how long they can keep using the device before they have to plug it back in.

since generally cells don't acquire more capacity, it's also perfectly reasonable taht the bms doesn't recheck for greater capacity, and only assumes taht once a previous discharge-to-empty amount has been reached, it should shut off the output and tell the device it has to be recharged to continue. as previously noted, some devices have the ability to recalibrate this number in the bms, by doing a complete charge/discahrge/recharge cycle, which would then increase the capacity of the pack if it's been recelled with new cells. but not all devices have this ability.

some bmss may even do it automatically, but we'd never know about those, unless it's only done on every x number of cycles, and the recelling was done between those, such that the first times it was used after recelling it only had the old capacity, and then suddenly after a few cycles it has the new capacity, with no gradation in between.



anyway, jsut saying that it is perfectly reasonable for designers to make packs this way, for both safety and for usability (knowing your runtime).


it's only us, an extreme minority, who find this annoying. ;)
 
A long time I did mine, but this should be helpful to others

In short, use the same capacity cells

I used NCR18650B, 3400 mAh vs the 2200 mAh original, at the first charge and discharge it worked, the BMS was calibrating. It lasted a long time compared to the old battery, nearly doubled

As soon the BMS realized something wrong was going on, by going from 48 Wh to 75 Wh, it blocked the battery, it won’t charge anymore or even power on the pc, I’ve read a few mods to the board that would bypass that but never pursued it

Some BMSs can be reprogramed, but mine was one of the few that hasn’t been cracked yet

In the end I bought a 9 cells battery that lasts me between 4 and 6 hours
 
amberwolf said:
mxer said:
Absolute shame about this BMS and others using chemical fuse,
not sure what you mean by "chemical fuse", as i'm not aware of anything like that in there. i'm sure they do have a normal fuse in them someplace in case of shorts, etc., to prevent fires.

there are fuses that are auto-resetting instead of just wires that melt but they make a fair bit of heat at useful currents, so have to be quite large and thus are rarely used in compact power sources like laptop batteries (i have seen one in a picture of a bms for ebike batteries, probably on its charging port); theyr'e usually used for things like usb port power protection, since that's only 500ma.


if you're referring to the bmses taht brick themselves once cells drop too low for safe reuse, that's not a fuse, just a software/firmware thing. its' actually a safety feature, since it is possible for an overdischarged cell to eventually go back up in voltage to above the safe level over a long enough time with no load on it, or if temperature rises enough, but it would still have any damage that occured from the overdischarge, and still be unsafe to recharge if so. so the bms is designed to prevent that from happening, to keep a fire from occuring.

almost no one using such packs would ever do what we do here, and recell them for various reasons, so it's a perfectly valid and reasonable thing to disable the pack completely under those conditions. i'd guess that less than one in a hundred thousand, maybe even one in a million, laptop users would ever even open their packs for any reason at all, much less recell them or repair them.


same thing with the oem ebike packs taht do this, though there is probably a much higher percentage of users (though likely still far less than one in ten thousand) that repair them on their own, rather than just replacing them, mostly simply because of cost (hundreds or thousands of dollars, vs a few dozen dollars for a laptop pack).




the coloumb-count technology has its' reasons, too, as it's a more accurate way to keep track of cell health, especially when cells are only 1 or 2p. this means a bms can monitor voltage of the lowest cell during discharge, and also count the total pack ah as it is discharged, so that it can report this capacity to the device it's part of, so that device much more accurately knows how much runtime is available for various discharge rates, and can accurately tell the user how long they can keep using the device before they have to plug it back in.

since generally cells don't acquire more capacity, it's also perfectly reasonable taht the bms doesn't recheck for greater capacity, and only assumes taht once a previous discharge-to-empty amount has been reached, it should shut off the output and tell the device it has to be recharged to continue. as previously noted, some devices have the ability to recalibrate this number in the bms, by doing a complete charge/discahrge/recharge cycle, which would then increase the capacity of the pack if it's been recelled with new cells. but not all devices have this ability.

some bmss may even do it automatically, but we'd never know about those, unless it's only done on every x number of cycles, and the recelling was done between those, such that the first times it was used after recelling it only had the old capacity, and then suddenly after a few cycles it has the new capacity, with no gradation in between.



anyway, jsut saying that it is perfectly reasonable for designers to make packs this way, for both safety and for usability (knowing your runtime).


it's only us, an extreme minority, who find this annoying. ;)

:thumb:
 
I did the battery cell swap on my Toshiba Satellite laptop and it won't do anything now. Just shows no battery detected.

My Sony Vaio shows no battery detected but runs fine off the charger connection. Battery is there just pitifully low voltage. Laptop is used as desktop only.
 
Unfortunately you can't just replace the cells in laptops like most other things the bms is complicated and if anything goes wrong including disconnecting cells it will shut down for life unless you have a special programmer just for laptop batteries that connects to the contacts. From there you can read the bms info such as charge cycles and reset the bms to work.

I worked at batteries plus for 5 years and rebuilt hundreds. If you have a batteries plus near you maybe they can help. Don't know if they still exist.
 
No chance of that happening down here. Thanks for the info, though.
 
amberwolf said:
it's relatively simple, as long as none of the cell (groups) are dead or overdischarged in the old pack. if they are you can't do this.

the only non-simple part is that sometimes a bms location or or pack size / shape leaves no room for extra wire lengths, which you have to have to do this.

additonaly if there are sensors inside the old cell block that you have to leave in the pack you will have to remove them form the block without disconnecting them from the bms as you remove the old blcok in a later step below, and reinstall them in hte new block again without disconnecting them from the bms.

bascially you just need to ensure no old pack wires are ever disconnected from the bms at any time, until the new pack si completely installed.

physically build the new cell block just like it needs to be for the pack size/shape/etc.

charge (or discharge) the new cell groups to match the old cell groups.

connect the main negative of the new cell block to the main negative of the bms or pack, where the old cell block's main negative is conected to.

connect the next most positive cell group wire of the new cell block to the bms connection point of it's equivalent on the old one.

repeat until you reach the main positive of the new cell block, and connect that to the bms connection of the main positive of the old one.

now all cells are connected, and the bms has no idea that they are there.

begin cutting the old cell block away from the bms / pack connections, ensureing you do not disconnect anything from th enew pack.

phsyically remove the old cell block

test to be sure the bms is allowing charge/discharge as normal (this may not be possible on some packs iwhtout first doing the next steps)

phsyically intall the new cel block, routing the (longer) wires as needed.

reassemble the casing.

use as normal.



in some laptop packs, and perhaps fairly advanced packs in other devices, it's still possbile that the bms has been keeping track of cell capacities, and may actually shutdown pack output based on counting charge output rather than a cell-level lvc, in which case it will still only have the same capacity it last had with the old pack...if that is the case there's not much you can do about it. the good news is it won't keep dropping, it'll stay at that capacity until the new cells age enough to drop below that.

in laptops, some of them have a battery-retest function of some kind, like an old gateway i have arund here somewhere. if they do, it may be able to test with the new cells and see what capacity they have, and then afterward be able to use that new greater capacity.


i don't expect this would be a problem in most packs, prbably none of the ebike packs i've heard of so far, but i don't have direct expereince doing any of this, i'm just interpolating data others have posted here and elsewhere. ;)

Nice explanation! Do you or someone else have an example in a video (youtube/forum etc.)?

I wanna try this for a Bosch 500.
 
OEM ebike stuff is different from laptops.

A number of manufacturers use BMS boards that brick themselves if the cells are disconnected, or if the pack ever drops below the "safe recharge" LVC (different from the regular shutoff LVC).

Whether Bosch does that on your specific pack I don't know; if you don't find any results searching around on that then you would probably be the first guinea pig, potentially scrapping your pack and having to buy a new one if the BMS does brick itself, and if the Bosch system requires that specific BMS to operate (some systems do, some don't).
 
amberwolf said:
OEM ebike stuff is different from laptops.

A number of manufacturers use BMS boards that brick themselves if the cells are disconnected, or if the pack ever drops below the "safe recharge" LVC (different from the regular shutoff LVC).

Whether Bosch does that on your specific pack I don't know; if you don't find any results searching around on that then you would probably be the first guinea pig, potentially scrapping your pack and having to buy a new one if the BMS does brick itself, and if the Bosch system requires that specific BMS to operate (some systems do, some don't).

Thank you,

but there are still plenty of revision companies that only replace the cells and not the completely bms? can the bms be “hacked” by keeping the voltage on it?
 
I don't know. IIRC someone here on the forums did this for an old OEM battery (bionx, I think, don't remember for sure), and it worked, but someone else did it for some other battery and it didn't work; dont' think anyone ever found out why.
 
amberwolf said:
I don't know. IIRC someone here on the forums did this for an old OEM battery (bionx, I think, don't remember for sure), and it worked, but someone else did it for some other battery and it didn't work; dont' think anyone ever found out why.

Thnx!
 
Back
Top