The bloating is gas caused by electrolyte decomposition. Most common cause is over charging but can also be caused by overheating or extremely low cell voltage. Electrolyte becomes unstable above a cell voltage of about 4.3v or below 0.1 vdc, or high temp from over current discharging or charging currents. The gas is not the main problem. The electrolyte decomposition also creates tar that clogs up the LFP and graphite electrode pores increasing cell impedance. The brownish discoloration is the tars left behind from electrolyte decomposition.
I had this happen to a cell when putting a fully charged cell parallel with cells that were not charged. The fully charged cell puffed up during the charge cycle.
A great video and I'm just getting around to watching it 7 years later! I wonder if the fault was caused by the same thing that happens to Lithium Ion cells? With LiIon a crystaline structure develops between the positive and negative terminals of the cells and shorts it out. LiIon will also swell up and may even burst into flames.
Li-ion plate out metallic lithium and generate hydrogen fully discharged. If left that way the process continues and if dendrite growth pierces the internal menbrane/shorts the electrodes then "bad things happen" - especially once the cell bursts and hot lithium + hydrogen are exposed to oxygen LiFePo4 are a lot more stable but I wouldn't leave them discharged for prolonged periods Always store your Lithiums at 1/3-2/3 charge, never leave them dead, even if the pack is "dead" Never keep dead lithiums around - they're a fire hazard. Send them for recycling
I have the same situation of an expanding cell in my Lifep04 Ping bike battery that is only 3 years old and without a lot of usage. I had a buddy of mine that has been e-biking for a long time told me it can be caused by charging when it's freezing temps or because I didn't keep it charged like I should have. You can take that for what it's worth... Hopefully this will save someone from this problem...
add coroplast/coreflute between cells in next pac with thermo fan to push air between cells through coroplast/corflute for heat dissipation should be able to sustain heaps more amps without overheating pack and clamp pack to keep laminates tight for longer life
Hi ebinary - Same place as I got the pack from... pingbattery.com ... You'll have to mail the guy as he doesn't have individual cells on his website, but he's always been helpful to me for doing custom packs, controllers cells etc. Hope that helps. Cheers!!
The tabs were soldered onto little PCBs that connected the cells together. I just used my soldering iron to melt the solder and a small screwdriver to pry up the tabs whist the solder was still molten. Cheers!!
very interesting video of the inner assembly of a LiFePo4.............as someone who uses the smaller LiFe packs ( 1700mah - 2200mah 2 cell packs ) for my my rc airplane receiver, do you know of a simple and safe charging rate rule for these batteries that I can follow to prolong the life of my pack, I charge at 1/2 C of the battery rating .....example: a 2000 mah pack gets charged at .5 C or 1 amp.....thanks for any help you can offer .....
Hi Gary. 1/2C is the recommended rate for most LiFEPo4s (unless the manufacturer tells you otherwise), and 3.65V per cell for the cut-off voltage. LiFEPo4s are supposed to be tolerant to overcharging, but the cell I pulled apart went puffy because the BMS wasn't terminating at the correct voltage. After a few hours sitting at full charge it was enough to damage the cell. Hope that helps. Cheers!!
@@ThingsWhichArentWork yes, it helps and I think I will stick with the .5C charging rate as a " rule of thumb" for my LiFePo4's...the only drawback in using these in the rc world that I have found is the voltage seems to drop off very quickly , and my LiPo low voltage alarm will not work for the LiFe packs....thanks much.
Jim Conner I was searching yesterday about some Lithium Titanate batteries, they still seem to be very pricey. Here's an example: www alibaba com/product-detail/Lithium-Titanate-Battery-2-4V-15AH_60280829686 HTML EDIT: TH-cam seems to edit the URL and exchanging the Filename HTML into lower case to so some awkward Unicode chars. If the link is not working, first copy and insert it in notepad, then add the Periods and change everything in ASCII which is not ASCII. HTML needs to be changed into lower case.
Charging while at or below freezing destroys these batteries every time. That is likely what caused the cells to bloat. When charging always ensure they are above freezing. Higher end BMS have temperature sensors to avoid damage because they will not charge below 0c. It is sad when they bloat and quit working like that....
I can promise you that it never gets below freezing inside my house where this battery lives. Personally I think it was a bad BMS that killed these cells by failing to prevent overvoltage conditions properly. Thanks for taking the time to comment. Cheers!!
That’s not really correct, the cathode is where the reduction occurs so in charge the cathode is the negative electrode and in discharge, the cathode is the positive electrode
Wow, thanks mate! I'm on the exact same track just now. One of four of my cells bloated up just like yours and it deformed all the other 3 cells next to it. It's been the 3rd cell in the pack. So you think it's not the BMS/Balancer that overcharged the cell, but a failure within the cell itself? I'm really curious, because naturally I'd like to know if it's safe to reuse my balancer module with new cells, or if the whole thing should be disposed of. I'd just like to know my chances before reusing the module just to watch it fry the next pack... Maybe I find a way to test the module health before reassembly...
Hi ***** - That's what I thought when I made the video, but if you have a look through the comments below, there's a good conversation between myself and ***** about the SignalLab BMS that was on my pack - Apparently the chip used on the SignalLab BMS does not cut off the charge until the pack voltage hits 3.95V (rather than 3.65 which it should be) - They've got away with this because LiFePO4 is generally pretty tolerant of overcharging conditions. In the case of my pack, the failure followed a very long charge cycle (I forgot about it for half a day) so the cell voltages will have risen very high on my pack. Having said all of that, after replacing the failed cells, I have continued to use the SignalLab BMS, but I've been a lot more careful about ensuring that I terminate the charge in a timely fashion. I only use that pack a few times a year, so having to keep a close eye on it when charging isn't a huge hardship for me - for other use cases it may be more problematic. Hope that helps. Cheers!!
Jim Conner Yeah, that helps, thanks! I watched this video too pxP0Cu00sZs - seems like heat + high voltage + slow charge/discharge = overkill for lithium cells. I'm using them as a motorcycle starter battery, so it's pretty hard to tell what exactly is going on between the generator, the lead-acid-battery-designed loading unit and the cells. I've got a balancer in between, but since a cell bloated up, that doesn't seem to be very good. It's a SHIDO YTZ-10S, but claiming warranty is pretty hard - they just didn't respond - and now I'm going to replace the cells myself, only keeping that balancer module.
***** I don't know what cells you have in your pack, but in general the square, flat cells (aka prismatic cells) don't have a great current delivery capacity, whereas the cylindrical cells have a lot more oompf available. My 20Ah pack is only rated for 40A max current delivery. Your Shido pack seems to be designed specifically for use on motorbikes though. I found some specs on your pack and they seem to suggest that it can deliver 190Amps for cranking, and can take a charge at up to 20Amps rate (6 mins to charge the pack). I'd be tempted to have a go at claiming the warranty if I were you, as it seems that you've been using the battery for exactly what it was designed for and it's failed anyway. Hope that helps. Cheers!!
Jim Conner in theory, yes. It's just that *now* I've taken a look inside. Didn't cut the bloated packs like you did tho, but I broke the warranty seal anyways. So even if they'd reply to me now, they surely would just laugh at me trying to claim warranty for the opened box.
yes you should check it. I've found the best way to check batteries and or chargers is to sit them on a table in front of you and look at it look at it for 2 minutes scratch your head and move to another spot and continue to watch it for another 10 minutes scratch your bumole using two fingers making sure that your fingers do not exceed the second knuckle In your bumole. after scratching you test the batteries.using the finger u scratched your bumole. smell your fingers to check for appropriate scent .
***** Hi. I started out thinking that the BMS was likely to blame, but came to the conclusion that it probably wasn't, because a.) none of the cells in that group indicated an overvoltage condition, b.) not all the cells in that group were damaged, and c.) LiFePO4 batteries reportedly have a good overcharge tolerance (I've read of cells being hammered up to 30V and still appearing fine afterwards).... I ultimately decided that it was most likely that dendrite growth from the edges of the plates had pierced the polymer seperator and caused the cell failure. Yes - the group of cells were in parallel - 4x 5Ah cells to make a 20Ah pack. Where did you find that 2.8V minimum figure? SignaLab (the cells I have) and A123 both specify 2.0V as the absolute minimum - I personally would have stopped the discharge at 2.5V, but the 2.0V minimum was pre-programmed into the charger and I couldn't work out how to change it. The BMS I am using (SignaLab again) has the low voltage cutoff at 2.1V per cell. As it's unlikely that you spotted the comment I added yesterday - I highly recommend this fascinating talk I stumbled across last night which is all about the reasons why Lithium batteries fail - th-cam.com/video/pxP0Cu00sZs/w-d-xo.html - It's a great rundown of the capabilities of a variety of different lithium battery types too. Having watched that, I'm more convinced that the cell finally died because I'd always stored it at 100% State of charge. Thanks for watching and thanks especially for taking the time to comment. Cheers!!
Jim Conner Those Signalab BMS use fixed treshold IC designed for LiCoO2 so the tresholds are not good for LiFePO4. They only stop the charging when highest cell is at 3.95V. And yes LiFePO4 is tolerant to higher voltage in the sens that it will not catch on fire but it dose not like beeing charged at more than 3.6V and is even worse if they need to stay at that level for long period of time. I will take a look at that video after this comment thanks for the link. I use LiFePO4 for a few years now offgrid with no problems. I designed a Solar BMS charger for offgrid application (is less ideal for electric vehicle or eBikes) And I have some videos on my channel about 100Ah GBS cells and A123 20Ah cells that may be of interest to you. Quality of the cells is also important. There can be manufacturing defects that will lead to failure even with a proper BMS.
***** Perhaps Signalab design for both chemistries. I'm 100% certain that it cuts off at 3.65V as I've checked it out with my multimeter in the past. LiCoO2 cells look like the way of the future though - amazing longevity on those things.... and they happen to be the ones that Tesla are using too, so looks good for any Tesla owners out there. Have just checked out your channel and subscribed... excellent - that'll give me something to watch over my christmas holidays :). Cheers!!
Jim Conner Thanks Jim for that link. I just watched that video and there are good information there. I have doubt that your Singnalab BMS stops at 3.65V. The reason for that is that even if recently there are IC that have LiFePO4 thresholds they are way more expensive than this old ones for LiCoO2. LiCoO2 is the most common Li-ion battery is what you have in your phone and tablet and billions of them are build every year. They are normally charged at 4.2V so they have the best energy density since for portable applications like phones and laptops the weight is way more important than cycle life (they will do fine for two or three years before you upgrade to a new model) Charging those same cells (that normally have a 500 cycles life at 4.2V) at 4.1V will decrees the capacity to about 80% but last 2x as long for so 1000 cycles in this example going down to 4V will double again to 2000 cycles and at 3.9V it doubles again to 4000 cycles but capacity will be just 65% of the one at 4.2V. At lower than 3.9V the advantage starts to disappear since the capacity will go down to quickly. So in special application like military and space they use Li-ion charged at a 3.95V that offers the best life and those are the IC used in almost all low cost BMS on the market that use that 5 pin or 6 pin IC for each cell and has fixed thresholds. If you try to charge the LiCoO2 at 4.35V or higher the cycle life dramatically goes down (you can see graphs even in the video you provided link to I think at about minute 37 or so) LiFePO4 is more tolerant to overvoltage but it will still damage the cell and reduce dramatically the life of the cell. Form 3.6V (that is already a stress) to 3.9V is a huge difference. What you have measured as 3.65V is do to the fact that your cells where balanced at the time you did the measurement but got unbalanced over time (that passive unintelligent balancer will do nothing and start only above 3.65V to late). So the reason you measured 3.65 is that the charger not the BMS stopped the charging (and at not sure how many cells you had say for 8 series cells) then charging ended at 28.8V and since cells where equal it stopped there but over not to many cycles the cells got unbalanced and that group of 4 parallel cells where overcharged at 3.9V or close and since they spend a lot of time at that level that degradation occurred and the weakest cell or most affected cells out of those 4 ended dead first. You know your exact model of BMS just go and check all the spec and they usually do not hide that Over voltage disconnect is at 3.9V but they say that is fine :) since you battery will die after quite some time (max hundred or so cycles usually a few years of casual use) and they do not care about that.
***** I hear what you're saying there, but I think that I have dodged that particular bullet. My BMS has two 4-pin chips (plus 5 resistors, an LED and a SOT23 package) per channel. The chips are Fairchild branded and say V831Y 817B on the top, but I've been unable to find a datasheet for them, so am uncertain as to exactly what their function is (They look just like opto isolaters, but could well be transistors in a strange package) . In the past I have tried charging the pack using my Bantam eStation 902 charger, but that was unable to complete a charge because the BMS disconnects the battery charger when any individual cell gets too high, causing the Bantam charger to error with 'Connection Broken' during the charge cycle - this appears to happen whenever a single cell reaches the cutoff voltage (which is higher than the voltage where balancer bleeding starts to occur). My pack is a custom 9S 29.7V 20Ah pack (it started life as an 8S) which is running with a tweaked 12S BMS.The power supply that came with the pack used to be a constant voltage 30V, 3A switchmode, but I tweaked that up to 33V to work with my 9S pack. I found a picture of my BMS online 4.bp.blogspot.com/-NxOQejvuXy4/TlPeQRif0qI/AAAAAAAAQ2w/ZmrIG26kT18/s1600/IMG_6690-1.JPGBecause the bleed rate on that BMS is pretty limited, I actually balance the packs using Bantam pack balancers (with 500mA bleed rate) whilst they're charing too - it speeds up the charging quite a lot as there's less time waiting with the PSU disconnected as cells bleed down from their max voltage. After a charge cycle (with balancing at the same time) the pack measures in at a little under 33V, I wish I could find an exact model number for that BMS, but info from SignaLab seems to be nonexistent. I guess I could just mail Li Ping and ask him though - he's been good at replying to my questions in the past.Anyway.. this reply is long enough now and I have a new video to upload :-)Cheers!!
Depends what cells, don't LiPo's have a charge controller built in, and need to be paralleled in a special way ? Also, looking at your cells, if both sides of the plates are coated with carbon, then both the 'front' and 'back' are against a plate of the opposite polarity, so making a much larger surface area, and that means more current
***** Hi Sparky. You can get LiPo and LiFePO4 cells which do have a BMS onboard, but these cells don't - they rely on an external BMS to keep everything in check. If the cells did have onboard BMSes then it would not be viable to group the cells together as they are in my pack. You are totally right about both sides of the cells being coated in material for increased capacity and current capabilities - I added the badly-drawn back-of-an-enevelope diagram was an attempt to highlight the role of the electrolyte in the cell because I didn't feel that my rambling did the job of explaining well enough. Thanks for watching, and taking the time to comment. Cheers!!
D'you know... that's not as easy a question as it might first seem :-)... 7 out of the 9 cells were fully charged, but the two that I autopsied had shorted themselves internally and would be dead. Cheers!!
I just stumbled across a superb talk about the different types of Lithium cells and why they all ultimately fail. I thought it was such a good talk I'd put a link here for anyone who fancies it. th-cam.com/video/pxP0Cu00sZs/w-d-xo.html Cheers!!
PATRICK MCKOWEN I'll admit that some of it went over my head too... I'm like a moth to a flame when it comes to clever science... always attracted to it, but sometimes get my wings a bit burnt :).... Cheers!!
@ 12:29 the join in the plastic membrane might have caused a higher internal resistance for that cell partition relative to the other cell partitions causing an imbalance of internal charges. over a 5 year period the imbalance would most probably have grown substantially probably being one of the causes of the over heating problems.
Looking at the video again @ 12:48 the plate you just flipped up along the left hand side I notice there seems to be a blackened section almost the same size as the membrane over lap, id say this is evidence the membrane join was the cause.
Good video...First i would be concerned NOT to inhale any of the gasses...not good...Second, adding new batteries, to a 5 year old pack can sometimes have issues(ie. charge/discharge?)...Has your new set worked into this pack ok?
Hi Alexandre Rouma - I didn't think that LiFePO4 batteries could produce hyrdogen - I know that Lead Acid batties produce it, but I thought that LiFePO4 were supposed to be intrinsically safe and not produce any explosive gasses. I'll do some googling and see if I can work out the relevant chemistry that would result in hyrdogen production... Got any pointers for me? Cheers!!
The bloating is gas caused by electrolyte decomposition. Most common cause is over charging but can also be caused by overheating or extremely low cell voltage. Electrolyte becomes unstable above a cell voltage of about 4.3v or below 0.1 vdc, or high temp from over current discharging or charging currents.
The gas is not the main problem. The electrolyte decomposition also creates tar that clogs up the LFP and graphite electrode pores increasing cell impedance. The brownish discoloration is the tars left behind from electrolyte decomposition.
I had this happen to a cell when putting a fully charged cell parallel with cells that were not charged. The fully charged cell puffed up during the charge cycle.
A great video and I'm just getting around to watching it 7 years later! I wonder if the fault was caused by the same thing that happens to Lithium Ion cells? With LiIon a crystaline structure develops between the positive and negative terminals of the cells and shorts it out. LiIon will also swell up and may even burst into flames.
Li-ion plate out metallic lithium and generate hydrogen fully discharged. If left that way the process continues and if dendrite growth pierces the internal menbrane/shorts the electrodes then "bad things happen" - especially once the cell bursts and hot lithium + hydrogen are exposed to oxygen
LiFePo4 are a lot more stable but I wouldn't leave them discharged for prolonged periods
Always store your Lithiums at 1/3-2/3 charge, never leave them dead, even if the pack is "dead"
Never keep dead lithiums around - they're a fire hazard. Send them for recycling
Nice disassemblying, you got to the core !
Nice demonstration of the internals of a lifepo4 battery
Glad you enjoed the video Victor - Thanks for taking the time to comment and let me know. Cheers!!
I have the same situation of an expanding cell in my Lifep04 Ping bike battery that is only 3 years old and without a lot of usage. I had a buddy of mine that has been e-biking for a long time told me it can be caused by charging when it's freezing temps or because I didn't keep it charged like I should have. You can take that for what it's worth... Hopefully this will save someone from this problem...
Mine seem to expand when I don't do a proper shut down and put them on the charger (cell phone batteries) but they're also cheaper brands
add coroplast/coreflute between cells in next pac with thermo fan to push air between cells through coroplast/corflute for heat dissipation should be able to sustain heaps more amps without overheating pack and clamp pack to keep laminates tight for longer life
This is fascinating. I'd love to see someone build a LifePO4 battery using off the shelf components.
createthis same
What is your source for replacement cells? I can't find anything similar online.
Hi ebinary - Same place as I got the pack from... pingbattery.com ... You'll have to mail the guy as he doesn't have individual cells on his website, but he's always been helpful to me for doing custom packs, controllers cells etc. Hope that helps. Cheers!!
Thanks!
Thanks. But how did you removed the individual cells from the bigger pack without damaging their tabs?
The tabs were soldered onto little PCBs that connected the cells together. I just used my soldering iron to melt the solder and a small screwdriver to pry up the tabs whist the solder was still molten. Cheers!!
very interesting video of the inner assembly of a LiFePo4.............as someone who uses the smaller LiFe packs ( 1700mah - 2200mah 2 cell packs ) for my my rc airplane receiver, do you know of a simple and safe charging rate rule for these batteries that I can follow to prolong the life of my pack, I charge at 1/2 C of the battery rating .....example: a 2000 mah pack gets charged at .5 C or 1 amp.....thanks for any help you can offer .....
Hi Gary. 1/2C is the recommended rate for most LiFEPo4s (unless the manufacturer tells you otherwise), and 3.65V per cell for the cut-off voltage. LiFEPo4s are supposed to be tolerant to overcharging, but the cell I pulled apart went puffy because the BMS wasn't terminating at the correct voltage. After a few hours sitting at full charge it was enough to damage the cell. Hope that helps. Cheers!!
@@ThingsWhichArentWork yes, it helps and I think I will stick with the .5C charging rate as a " rule of thumb" for my LiFePo4's...the only drawback in using these in the rc world that I have found is the voltage seems to drop off very quickly , and my LiPo low voltage alarm will not work for the LiFe packs....thanks much.
Thanks for your LiFePO4 Battery teardown.
Could you also make a Lithium-titanate battery teardown? This would be very interesting.
Hi OpenGL4ever - I'd love to make that video, but sadly I don't have any lithium-titanate batteries to take apart. Cheers!!
Jim Conner
I was searching yesterday about some Lithium Titanate batteries, they still seem to be very pricey.
Here's an example:
www alibaba com/product-detail/Lithium-Titanate-Battery-2-4V-15AH_60280829686 HTML
EDIT:
TH-cam seems to edit the URL and exchanging the Filename HTML into lower case to so some awkward Unicode chars. If the link is not working, first copy and insert it in notepad, then add the Periods and change everything in ASCII which is not ASCII. HTML needs to be changed into lower case.
Charging while at or below freezing destroys these batteries every time. That is likely what caused the cells to bloat. When charging always ensure they are above freezing. Higher end BMS have temperature sensors to avoid damage because they will not charge below 0c. It is sad when they bloat and quit working like that....
I can promise you that it never gets below freezing inside my house where this battery lives. Personally I think it was a bad BMS that killed these cells by failing to prevent overvoltage conditions properly. Thanks for taking the time to comment. Cheers!!
9:51 Al is cathode but cathode is positive in the Battery chemistry
That’s not really correct, the cathode is where the reduction occurs so in charge the cathode is the negative electrode and in discharge, the cathode is the positive electrode
Wow, thanks mate! I'm on the exact same track just now. One of four of my cells bloated up just like yours and it deformed all the other 3 cells next to it. It's been the 3rd cell in the pack. So you think it's not the BMS/Balancer that overcharged the cell, but a failure within the cell itself? I'm really curious, because naturally I'd like to know if it's safe to reuse my balancer module with new cells, or if the whole thing should be disposed of. I'd just like to know my chances before reusing the module just to watch it fry the next pack...
Maybe I find a way to test the module health before reassembly...
Hi ***** - That's what I thought when I made the video, but if you have a look through the comments below, there's a good conversation between myself and ***** about the SignalLab BMS that was on my pack - Apparently the chip used on the SignalLab BMS does not cut off the charge until the pack voltage hits 3.95V (rather than 3.65 which it should be) - They've got away with this because LiFePO4 is generally pretty tolerant of overcharging conditions. In the case of my pack, the failure followed a very long charge cycle (I forgot about it for half a day) so the cell voltages will have risen very high on my pack. Having said all of that, after replacing the failed cells, I have continued to use the SignalLab BMS, but I've been a lot more careful about ensuring that I terminate the charge in a timely fashion. I only use that pack a few times a year, so having to keep a close eye on it when charging isn't a huge hardship for me - for other use cases it may be more problematic. Hope that helps. Cheers!!
Jim Conner Yeah, that helps, thanks!
I watched this video too pxP0Cu00sZs - seems like heat + high voltage + slow charge/discharge = overkill for lithium cells.
I'm using them as a motorcycle starter battery, so it's pretty hard to tell what exactly is going on between the generator, the lead-acid-battery-designed loading unit and the cells. I've got a balancer in between, but since a cell bloated up, that doesn't seem to be very good. It's a SHIDO YTZ-10S, but claiming warranty is pretty hard - they just didn't respond - and now I'm going to replace the cells myself, only keeping that balancer module.
***** I don't know what cells you have in your pack, but in general the square, flat cells (aka prismatic cells) don't have a great current delivery capacity, whereas the cylindrical cells have a lot more oompf available. My 20Ah pack is only rated for 40A max current delivery. Your Shido pack seems to be designed specifically for use on motorbikes though. I found some specs on your pack and they seem to suggest that it can deliver 190Amps for cranking, and can take a charge at up to 20Amps rate (6 mins to charge the pack). I'd be tempted to have a go at claiming the warranty if I were you, as it seems that you've been using the battery for exactly what it was designed for and it's failed anyway. Hope that helps. Cheers!!
Jim Conner in theory, yes. It's just that *now* I've taken a look inside. Didn't cut the bloated packs like you did tho, but I broke the warranty seal anyways. So even if they'd reply to me now, they surely would just laugh at me trying to claim warranty for the opened box.
yes you should check it. I've found the best way to check batteries and or chargers is to sit them on a table in front of you and look at it look at it for 2 minutes scratch your head and move to another spot and continue to watch it for another 10 minutes scratch your bumole using two fingers making sure that your fingers do not exceed the second knuckle In your bumole. after scratching you test the batteries.using the finger u scratched your bumole. smell your fingers to check for appropriate scent .
I want a sign for my shop when I'm out: "Gone puffy"
I was probably the BMS. Was that group of cells in parallel?
2V is way to low for LiFePO4. 2.8V is specified by most manufactures as safe minimum.
***** Hi. I started out thinking that the BMS was likely to blame, but came to the conclusion that it probably wasn't, because a.) none of the cells in that group indicated an overvoltage condition, b.) not all the cells in that group were damaged, and c.) LiFePO4 batteries reportedly have a good overcharge tolerance (I've read of cells being hammered up to 30V and still appearing fine afterwards).... I ultimately decided that it was most likely that dendrite growth from the edges of the plates had pierced the polymer seperator and caused the cell failure.
Yes - the group of cells were in parallel - 4x 5Ah cells to make a 20Ah pack.
Where did you find that 2.8V minimum figure? SignaLab (the cells I have) and A123 both specify 2.0V as the absolute minimum - I personally would have stopped the discharge at 2.5V, but the 2.0V minimum was pre-programmed into the charger and I couldn't work out how to change it. The BMS I am using (SignaLab again) has the low voltage cutoff at 2.1V per cell.
As it's unlikely that you spotted the comment I added yesterday - I highly recommend this fascinating talk I stumbled across last night which is all about the reasons why Lithium batteries fail - th-cam.com/video/pxP0Cu00sZs/w-d-xo.html - It's a great rundown of the capabilities of a variety of different lithium battery types too. Having watched that, I'm more convinced that the cell finally died because I'd always stored it at 100% State of charge.
Thanks for watching and thanks especially for taking the time to comment. Cheers!!
Jim Conner
Those Signalab BMS use fixed treshold IC designed for LiCoO2 so the tresholds are not good for LiFePO4.
They only stop the charging when highest cell is at 3.95V. And yes LiFePO4 is tolerant to higher voltage in the sens that it will not catch on fire but it dose not like beeing charged at more than 3.6V and is even worse if they need to stay at that level for long period of time.
I will take a look at that video after this comment thanks for the link.
I use LiFePO4 for a few years now offgrid with no problems.
I designed a Solar BMS charger for offgrid application (is less ideal for electric vehicle or eBikes)
And I have some videos on my channel about 100Ah GBS cells and A123 20Ah cells that may be of interest to you.
Quality of the cells is also important. There can be manufacturing defects that will lead to failure even with a proper BMS.
***** Perhaps Signalab design for both chemistries. I'm 100% certain that it cuts off at 3.65V as I've checked it out with my multimeter in the past. LiCoO2 cells look like the way of the future though - amazing longevity on those things.... and they happen to be the ones that Tesla are using too, so looks good for any Tesla owners out there. Have just checked out your channel and subscribed... excellent - that'll give me something to watch over my christmas holidays :). Cheers!!
Jim Conner Thanks Jim for that link. I just watched that video and there are good information there.
I have doubt that your Singnalab BMS stops at 3.65V. The reason for that is that even if recently there are IC that have LiFePO4 thresholds they are way more expensive than this old ones for LiCoO2.
LiCoO2 is the most common Li-ion battery is what you have in your phone and tablet and billions of them are build every year.
They are normally charged at 4.2V so they have the best energy density since for portable applications like phones and laptops the weight is way more important than cycle life (they will do fine for two or three years before you upgrade to a new model)
Charging those same cells (that normally have a 500 cycles life at 4.2V) at 4.1V will decrees the capacity to about 80% but last 2x as long for so 1000 cycles in this example going down to 4V will double again to 2000 cycles and at 3.9V it doubles again to 4000 cycles but capacity will be just 65% of the one at 4.2V.
At lower than 3.9V the advantage starts to disappear since the capacity will go down to quickly.
So in special application like military and space they use Li-ion charged at a 3.95V that offers the best life and those are the IC used in almost all low cost BMS on the market that use that 5 pin or 6 pin IC for each cell and has fixed thresholds.
If you try to charge the LiCoO2 at 4.35V or higher the cycle life dramatically goes down (you can see graphs even in the video you provided link to I think at about minute 37 or so)
LiFePO4 is more tolerant to overvoltage but it will still damage the cell and reduce dramatically the life of the cell. Form 3.6V (that is already a stress) to 3.9V is a huge difference.
What you have measured as 3.65V is do to the fact that your cells where balanced at the time you did the measurement but got unbalanced over time (that passive unintelligent balancer will do nothing and start only above 3.65V to late).
So the reason you measured 3.65 is that the charger not the BMS stopped the charging (and at not sure how many cells you had say for 8 series cells) then charging ended at 28.8V and since cells where equal it stopped there but over not to many cycles the cells got unbalanced and that group of 4 parallel cells where overcharged at 3.9V or close and since they spend a lot of time at that level that degradation occurred and the weakest cell or most affected cells out of those 4 ended dead first.
You know your exact model of BMS just go and check all the spec and they usually do not hide that Over voltage disconnect is at 3.9V but they say that is fine :) since you battery will die after quite some time (max hundred or so cycles usually a few years of casual use) and they do not care about that.
***** I hear what you're saying there, but I think that I have dodged that particular bullet. My BMS has two 4-pin chips (plus 5 resistors, an LED and a SOT23 package) per channel. The chips are Fairchild branded and say V831Y 817B on the top, but I've been unable to find a datasheet for them, so am uncertain as to exactly what their function is (They look just like opto isolaters, but could well be transistors in a strange package) . In the past I have tried charging the pack using my Bantam eStation 902 charger, but that was unable to complete a charge because the BMS disconnects the battery charger when any individual cell gets too high, causing the Bantam charger to error with 'Connection Broken' during the charge cycle - this appears to happen whenever a single cell reaches the cutoff voltage (which is higher than the voltage where balancer bleeding starts to occur). My pack is a custom 9S 29.7V 20Ah pack (it started life as an 8S) which is running with a tweaked 12S BMS.The power supply that came with the pack used to be a constant voltage 30V, 3A switchmode, but I tweaked that up to 33V to work with my 9S pack. I found a picture of my BMS online 4.bp.blogspot.com/-NxOQejvuXy4/TlPeQRif0qI/AAAAAAAAQ2w/ZmrIG26kT18/s1600/IMG_6690-1.JPGBecause the bleed rate on that BMS is pretty limited, I actually balance the packs using Bantam pack balancers (with 500mA bleed rate) whilst they're charing too - it speeds up the charging quite a lot as there's less time waiting with the PSU disconnected as cells bleed down from their max voltage. After a charge cycle (with balancing at the same time) the pack measures in at a little under 33V, I wish I could find an exact model number for that BMS, but info from SignaLab seems to be nonexistent. I guess I could just mail Li Ping and ask him though - he's been good at replying to my questions in the past.Anyway.. this reply is long enough now and I have a new video to upload :-)Cheers!!
This is a great opportunity for me to learn and not have cancer doing it.. thanks for your sacrifice..
I kept myself entertained today by taking a look inside a LiFePO4 cell. Maybe it'll entertain you too. Cheers!!
You can depuff them by putting a voltage on them, in 2 weeks they will be back to normal.
Interesting
Interesting that Jim. Didn't realise there were so many individual cells in those.
Thanks John - Nor did I - I was expecting just two layers inside there. I'll never worry about running multiple cells in parallel again. Cheers!!
Depends what cells, don't LiPo's have a charge controller built in, and need to be paralleled in a special way ?
Also, looking at your cells, if both sides of the plates are coated with carbon, then both the 'front' and 'back' are against a plate of the opposite polarity, so making a much larger surface area, and that means more current
***** Hi Sparky. You can get LiPo and LiFePO4 cells which do have a BMS onboard, but these cells don't - they rely on an external BMS to keep everything in check. If the cells did have onboard BMSes then it would not be viable to group the cells together as they are in my pack. You are totally right about both sides of the cells being coated in material for increased capacity and current capabilities - I added the badly-drawn back-of-an-enevelope diagram was an attempt to highlight the role of the electrolyte in the cell because I didn't feel that my rambling did the job of explaining well enough. Thanks for watching, and taking the time to comment. Cheers!!
Wonderful video Jim ...Worth the wait .. Pitty the Cells died though !
Thanks John - The replacement cells weren't too expensive, so for entertainment value, this works out far cheaper than going to the cinema. Cheers!!
Great analysis. I wonder if you could someway of clean the cells and layers and reassemble in a better configuration spicific to your need
Now I know why people say on the Internet that it is safer to use and can be used at high temperatures..
Good teardown but i have never seen an Anode being described with a + or a Cathode as - well only in electroplating.
But well done anyhow
Thank for this! Great video! Learned a lot!
A little bit of brown paint and it's as good as new.
Full battery or dead battery?
D'you know... that's not as easy a question as it might first seem :-)... 7 out of the 9 cells were fully charged, but the two that I autopsied had shorted themselves internally and would be dead. Cheers!!
I try cut and autopsy of Lipo and get fire from bank power.
Alexsis Scott hold her to the floor Scotty!
Autopsy, Does It born died?
Good autopsy - -thanks for sharing - cheers, Patrick
Thank you PATRICK MCKOWEN . I'm very glad you enjoyed it. Hope you're having a good weekend. Cheers!!
Please activate subtitles. thanks.
This sounds like fire city to me...
I just stumbled across a superb talk about the different types of Lithium cells and why they all ultimately fail. I thought it was such a good talk I'd put a link here for anyone who fancies it. th-cam.com/video/pxP0Cu00sZs/w-d-xo.html Cheers!!
Jim Conner Yes I watched that - interesting but a bit over my head.Cheers,Patrick
PATRICK MCKOWEN I'll admit that some of it went over my head too... I'm like a moth to a flame when it comes to clever science... always attracted to it, but sometimes get my wings a bit burnt :).... Cheers!!
@ 12:29 the join in the plastic membrane might have caused a higher internal resistance for that cell partition relative to the other cell partitions causing an imbalance of internal charges. over a 5 year period the imbalance would most probably have grown substantially probably being one of the causes of the over heating problems.
Looking at the video again @ 12:48 the plate you just flipped up along the left hand side I notice there seems to be a blackened section almost the same size as the membrane over lap, id say this is evidence the membrane join was the cause.
Good video...First i would be concerned NOT to inhale any of the gasses...not good...Second, adding new batteries, to a 5 year old pack can sometimes have issues(ie. charge/discharge?)...Has your new set worked into this pack ok?
The future of battery and energy storage is organic.. how much does the human body store/use.. anyone?
thanks
Great video thx for sharing
that was cool. tkx for doing it for us
Thanks +riskinhos - Glad you enjoyed it... I certainly had fun pulling the cells to pieces. Cheers!!
It's simply the litium that releases hydrogen
Hi Alexandre Rouma - I didn't think that LiFePO4 batteries could produce hyrdogen - I know that Lead Acid batties produce it, but I thought that LiFePO4 were supposed to be intrinsically safe and not produce any explosive gasses. I'll do some googling and see if I can work out the relevant chemistry that would result in hyrdogen production... Got any pointers for me? Cheers!!
NiMH does release hydrogen.
poosibly with some hydrogen flouride mixed in... which is particularly nasty stuff and shouldn't be inhaled.
Good job
Verry good !! :)
Super !
This was vivisection.
These are not Lifepo4 !!!
Oh yes they are.
k