The Consequence Of Not Using BMS With LiPo Batteries

Lyen

Lyen

10 kW
Joined
Aug 16, 2007
Messages
621
Location
San Francisco, California, USA
Here is the consequence of not using a BMS for the LiPo batteries. Here is the story, one night I was riding my ebike to work and while I was on my way I have noticed something smell sour from inside the rack bag. I then opened up the bag and noticed the batteries were cooking due to over discharge. The setup is 3 packs of 14.8v 10aH LiPo RC battery with 20C rating (Total = 44V with 3.7v of 12S1P) and my /controller motor setup is 48V 35A controller + 750W Super BMC planetary drive motor. My guess was that the RC charger did not get all the cells charged up and a number of cells were over-discharged and started to swell. The image were taken after the batteries were cool down. The size at the time I discovered when hot to touch was about 30% bigger. I think the temperature was over 212F or 100C because it melted the things surrounded it in the bag and burned my fingers when I touched it. They were swelling like a bread place in water when they were about to explode. Three cells of the two batteries actually did some how exploded and leaked out the sourly smelling gas with holes on the cells. Do you guys have any ideas how I can avoid it from happening again?

See my damaged cells below:





 
Yikes! You're lucky it didn't catch on fire.

Did you notice the bike was slowing down right before this happened?

Other than a low charge, the other possibility is the discharge current is too high for the cells and they overheated due to internal resistance. Don't always trust advertised "C" ratings.

I guess it's nice to know those cells can be seriously abused like that without a catastrophic fireball.
 
Just use two of these while discharging.
They will monitor your cell voltage, and start beeping if one of the cells goes bellow 3.0v
http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=8927&Product_Name=BM-6_Cell_Voltage_Monitor_2-6S_Lipo_%28New_Version%29
 
fechter said:
Yikes! You're lucky it didn't catch on fire.

Did you notice the bike was slowing down right before this happened?

Other than a low charge, the other possibility is the discharge current is too high for the cells and they overheated due to internal resistance. Don't always trust advertised "C" ratings.

I guess it's nice to know those cells can be seriously abused like that without a catastrophic fireball.
Click to expand...

Hello Fechter! Yes, I did notice a slow down while riding to work. However, it was too chilly on the road especially at over 27mph and my only thinking was to get to the destination ASAP & nothing else. The discharge current should be about 35A as the analog brushless controller was rated. If the battery is rated for 20C @ 10Ah a cell, then it should do fine at 200Ah right? I believe you are right about the internal resistance of the battery cells. Can you tell me a way or a link to measure internal resistance? I appreciate it. :)
 
Dave-s said:
Just use two of these while discharging.
They will monitor your cell voltage, and start beeping if one of the cells goes bellow 3.0v
http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=8927&Product_Name=BM-6_Cell_Voltage_Monitor_2-6S_Lipo_%28New_Version%29
Click to expand...

Dave-s, thank you very much for showing me the link. The device is very useful to alarm the user of low voltage. However, I noticed the reviews from the buyers commented device is not quite reliable. Have you been using one of those yourself? Do you have similar issues like the other buyers do?
 
pgt400 said:
Wow, what kind oi range were you getting with these?
Click to expand...

Hello pgt400! These batteries are the highest density I could find as of today. It weights a total of 2121gram or 4.68 pounds for a total of 12 cells @ 44V 10Ah. In the San Francisco hilly area with stop & go traffic. I managed to go 14 miles in as the realistic figure until the batteries reached below 3.0v. :)
 
Yep, I use 2 of them myself.
No problems like the reviewers wrote.
Very cheap and very accurate.
BTW, where did you get the 10000mah packs?
 
For what it's worth I use a pair of modified (to pull power from all 7 or 8 cells) cellLog8s (the mnod is on Junsi forum or hillrc or somthing like that) with HVC and LVC set for 4.16v and 3.25v. My CellLog8s seem more accurate than any of my other equipment (I bought 4 more) and I use the alarm output as follows:

While in discharge mode, alarm triggering (LVC) will disengage the throttle until the alarm condition passes... this isn't the greatest method, I have one setup where I am playing to toggling an Infineon EB809 Speed %1 which has been preset to 50% mode. This so far allows a reduced accelleration and top end but, the LVC condition many times can be nursed for another mile or so (think limp home mode).

While in charge mode, any cell going over the 4.16v mark (the core charger, a meanwell S-350-48 configured for 62.33v which in theory should equalize all cells at 4.155v. In practice without balancers I see about 6-12mv out of balance condition fully charged and with the HobbyKing discharge style balancers attached to each pack in balance mode, once the CV phase is reached and current goes low enough the balancers actually do a nice job (even if they are not displaying accurate voltages on the LCD) of retarding the faster to charge cells to better equalize the pack... my 15s done this way fall within 5-8mv across all 15s2p (build as 2P5S3S).

Since any cell hitting 4.16v will cause a disconnect of the charger, it gives the balancers a chance to pull down that slightly high cell.

I am really waiting for Gary and Fechters balancer and cobbled this to prevent what you have experienced, though I am at a loss unless you D E E P cycled those packs how you did such damage... I have turnigy lipo packs I've run down to 1v a cell or there abouts and recovered it with full capacity (and a little bloating but nothing like yours).

Did you by chance measure the cell IRs on the packs before use? Just curious if those bloated cells were not perhaps much higher IR than expected on such rated packs.

Where did the 10AH packs come from anyway? I'd be curious to build up a bunch of those into a mega pack I have been planning...

-Mike

PS: Do your cells still work to full capacity?

EDIT:

Somthing just occoured to me... if I use a combination low current balancer system like the HobbyKing or any other available drain style balancer... the difference alarm on the cell log 8 would be perfect for acheiving very precise balancing simply by engaging and disengaging the connection to the charger when the balance gets too far out... Would this work do you think?
 
I'm using this monitor: http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=7227&Product_Name=Hobby_King_Battery_Monitor_6S
The poor man BMS :D
Balancing is done during charge.
 
What cells are these? A rough calc shows ~211 wh/kg which is pretty amazing. Wikipedia shows a high figure of 200 wh/kg for LiPo so maybe it needs to be updated. :)
 
They look like the cheap ebay packs. Akku, ecpower, etc. Nowhere near the specs they claim, but still a great value for low C rate use.
 
The basic flaw it the no bms plan is when you "think" they are fully charged and are not. Bms can be expensive and complicated, but if you do a routine ride, you would be able to know what the routine amount of kwh the charger should use. Then a cheap killawatt meter could be used to see if the cells took the charge you expect them to. better than a guess anyway... This method helps show that something went wrong with a charge cycle.
 
mwkeefer said:
For what it's worth I use a pair of modified (to pull power from all 7 or 8 cells) cellLog8s (the mnod is on Junsi forum or hillrc or somthing like that) with HVC and LVC set for 4.16v and 3.25v. My CellLog8s seem more accurate than any of my other equipment (I bought 4 more) and I use the alarm output as follows:

While in discharge mode, alarm triggering (LVC) will disengage the throttle until the alarm condition passes... this isn't the greatest method, I have one setup where I am playing to toggling an Infineon EB809 Speed %1 which has been preset to 50% mode. This so far allows a reduced accelleration and top end but, the LVC condition many times can be nursed for another mile or so (think limp home mode).

While in charge mode, any cell going over the 4.16v mark (the core charger, a meanwell S-350-48 configured for 62.33v which in theory should equalize all cells at 4.155v. In practice without balancers I see about 6-12mv out of balance condition fully charged and with the HobbyKing discharge style balancers attached to each pack in balance mode, once the CV phase is reached and current goes low enough the balancers actually do a nice job (even if they are not displaying accurate voltages on the LCD) of retarding the faster to charge cells to better equalize the pack... my 15s done this way fall within 5-8mv across all 15s2p (build as 2P5S3S).

Since any cell hitting 4.16v will cause a disconnect of the charger, it gives the balancers a chance to pull down that slightly high cell.

I am really waiting for Gary and Fechters balancer and cobbled this to prevent what you have experienced, though I am at a loss unless you D E E P cycled those packs how you did such damage... I have turnigy lipo packs I've run down to 1v a cell or there abouts and recovered it with full capacity (and a little bloating but nothing like yours).

Did you by chance measure the cell IRs on the packs before use? Just curious if those bloated cells were not perhaps much higher IR than expected on such rated packs.

Where did the 10AH packs come from anyway? I'd be curious to build up a bunch of those into a mega pack I have been planning...

-Mike

PS: Do your cells still work to full capacity?

EDIT:

Somthing just occoured to me... if I use a combination low current balancer system like the HobbyKing or any other available drain style balancer... the difference alarm on the cell log 8 would be perfect for acheiving very precise balancing simply by engaging and disengaging the connection to the charger when the balance gets too far out... Would this work do you think?
Click to expand...

mwkeefer, thanks for your ideas to prevent this from happening. The cellLog8s is really interesting as well as the BM-6 Cell Voltage Monitor from HobbyKing. Which one should I get?

Can you please tell me how you measure the internal resistance? Is there any special tools or procedure to do besides just using the voltmeter? I have a CBA III. Can I calculate the IR via discharge/capacity/timing from the CBA?

Dave-s & Mike, I have got my 10Ah packs from my uncle who lives in Hong Kong. He told me the packs were illegally imported "somewhere" from the mainland due to unapproved/licensed chemical being used in Hong Kong. If you are interested, I can hook you up on the "samples". Just PM me.

I have not touch the packs after a few cells got exploded. I will test the cells individually with me CBA III and let you know.

As for the low current balancer. I have actually tried it with my two Hyperion LBA10NET and they works okay. I think your idea might work but do use mean to use it when you are on or off the ebike?
 
Dave-s said:
Just use two of these while discharging.
They will monitor your cell voltage, and start beeping if one of the cells goes bellow 3.0v
http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=8927&Product_Name=BM-6_Cell_Voltage_Monitor_2-6S_Lipo_%28New_Version%29
Click to expand...

Dave-s, I have just looked it up from HobbyKing. This is low voltage trigger alarm circuit. I like something that is lightweight and simple like this. Thanks! :)
batt-monitor-6S.jpg
 
swbluto said:
What cells are these? A rough calc shows ~211 wh/kg which is pretty amazing. Wikipedia shows a high figure of 200 wh/kg for LiPo so maybe it needs to be updated. :)
Click to expand...

These are RC LiPo cells that my uncle bought for me after doing some research on the highest density lipos. I was lucky that it past the custom during delivery because they are unapproved to be sold in Hong Kong as my uncle told me. My have actually community to work with these packs for months. This is the only incident due to my mistakes for not check the condition of the packs after charged. I bought them because I had to take my ebike up & down the stairs to work evreyday and the lighter the ebike the more convenient for me. I have managed to shave 12 pounds from the previous Lifepo4 setup to Lipos. I can hook you up on my next order if you are interested. :)
 
dogman said:
The basic flaw it the no bms plan is when you "think" they are fully charged and are not. Bms can be expensive and complicated, but if you do a routine ride, you would be able to know what the routine amount of kwh the charger should use. Then a cheap killawatt meter could be used to see if the cells took the charge you expect them to. better than a guess anyway... This method helps show that something went wrong with a charge cycle.
Click to expand...

dogman, I regret my stupidity. I did my routine ride commuting to work via ebike for the past 2 years. I guess I was spoiled with the SLA, Nimh, and then Lifepo4 in the past. I was lucky because the pouch of the packs were heat resistance. I did not know that until after I called my uncle in Hong Kong and he told me that after the incident.

I love your awesome idea to check with the Killawatt meter. I have actually bought a bunch when they were on sale at Newegg.com. I think using the Killawatt meter is one of the best way to confirm power draw from the AC outlet. I can just press the reset button on it prior start the recharge process. Thanks! :)
 
vanilla ice said:
They look like the cheap ebay packs. Akku, ecpower, etc. Nowhere near the specs they claim, but still a great value for low C rate use.
Click to expand...

I was searching for it on eBay. Where can I locate it? The only 14.8V batteries I found are only up to 5300mah with links below:

1804482141088080_1.jpg

http://cgi.ebay.com/Thunder-Power-5...Control_Parts_Accessories?hash=item2a038d3c5c

2303961280248080_3.jpg

http://cgi.ebay.com/Reedy-4S-Lipo-B...Control_Parts_Accessories?hash=item35a4adeb18

2705054944148080_1.jpg

http://cgi.ebay.com/Great-Planes-El...Control_Parts_Accessories?hash=item3efb62478e
 
http://cgi.ebay.co.uk/EC-POWER-10000mAh-11.1V-20C-25C-LIPO-Battery-akku-RC-B_W0QQitemZ180440579514

There are not any on the USA ebay site right now. I have a few of their packs, but I ordered directly from their website, not from ebay. I got the product and they do work (btw 1of3 needed repairs out of the box).. but not to claimed specs. The site seems to be down right now though. Judging by V sag, Cmax is closer to 2c continuous than 20c.. haven't done any capacity testing.. Kang and a few others here also have this brand lipo, maybe one of them has tested.

http://www.ecpower.com.hk/
 
What you're looking at in the pics is a classic cell reversal from over-discharge. Cells will never all be exactly the same capacity, which means if you're going to keep running them down, some cells will run out of capacity first, and these cells no longer can no longer hold themselves above zero, and the remaining cells which still have some capacity continue to make current flow through them until some reverse cell voltage to the point they go thermal and vent/puff.


I've got a few of those 10Ah cells as well. Very high Ri and more voltage sag than any of the 5Ah cells I've got. They really did have a full 10Ah of capacity through, so for a low power, high energy application (like maybe 2P in a 20-40a controller E-bike), they would be great.
 
lyen said:
Dave-s said:
Just use two of these while discharging.
They will monitor your cell voltage, and start beeping if one of the cells goes bellow 3.0v
http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=8927&Product_Name=BM-6_Cell_Voltage_Monitor_2-6S_Lipo_%28New_Version%29
Click to expand...

Dave-s, I have just looked it up from HobbyKing. This is low voltage trigger alarm circuit. I like something that is lightweight and simple like this. Thanks! :)
batt-monitor-6S.jpg
Click to expand...
The problem with this monitor is that it only monitors the entire pack's voltage, so 1 cell can discharge below 3v and it won't set the alarm because the voltage of the good cells compensate
The monitor that I suggested monitors each cell's voltage separately, so the moment one of the cells drops below 3v it will set the alarm.
The cell-log8 is also good, but it costs twice as much as the monitor I suggested.
http://www.hobbyking.com/hobbycity/store/uh_viewItem.asp?idProduct=9282&Product_Name=Cell-Log_Cell_Voltage_Monitor_2-8S_Lipo
 
Monitoring the whole pack voltage is a killer - I've done this before with a bad (runt) pack made of konions, at 18V (6S LVC) there was a parallel pair that was sitting at 1.8V (oops!) at 3V the batteries voltage basically drops like a cliff!
 
Here is the follow up for the issues. I have attempted to repair the batteries. It turns out only 2 cells are bad with opened pouch.

Here's what it looks like after the packs extraction:




I then try to reseal the opened pouches with gluegun. I am currently in the process of recharging them. One is holding the charge, and the other one is not. See below:


When it is finished charge, I will post the remaining capable results from my CBA III. Stay tuned... :)
 
Whoa! Make sure you charge those puffed/popped cells in something that you don't mind being exposed to lots of fire and burning. Something like a ceramic flower pot sitting in a frying pan.

Once O2 is inside the cell, and once the solvents have vaporized out, you're in a VERY easy condition to cause thermal run-away and fire.

I'm all for the experimenting, and I look forward to see how they perform in that condition. Just want to make sure you're charging/testing in a place you don't mind being subjected to fire and smoke.
 
You must log in or register to reply here.

Similar threads

A
Lipo battery pack build question
Replies
4
Views
697
Atxraider
A
X
    • Like
Custom 15,000W eBike 60+ MPH 100.8V LiPo Powered
Replies
3
Views
1,319
Xform
X
K
Where can I find a BMS that can manage 2 lipo packs?
Replies
6
Views
645
karim-humanitas
K
chuyskywalker
Fiido Q1S Upgrade; 200A, VESC, Copper-Nickel 52v Molicel
Replies
23
Views
3,670
chuyskywalker
chuyskywalker
T
LiPO PSA - How Not to Bun Your House Down
Replies
15
Views
1,057
donn
D
Back
Top