Popular Post Chriull Posted February 6, 2019 Popular Post Share Posted February 6, 2019 (edited) As i just dig a bit more into li ion cells and BMS in https://forum.electricunicycle.org/topic/12618-ks16s-damaged-battery/, here my gained insights and some conclusions. I'm happily awaiting corrections, additions and whatever else others find to improve this! Unfortionately we need the Li Ion cells in series to get usable voltages to drive our EUCs, which leads to a couple of problems. Mainly by different "aging" of the cells the weakest one gets "hammered" by the others and will be the first to quit working. From my readings the only real solution to have long lasting packs is perfect cell matching when producing battery packs! What provides the BMS to prevent premature cell degradation and keeps us safe (fire hazard): - overvoltage protection on cell basis: As in the above linked topic was stated for example the KS18L/XL BMS shuts of battery the charging input if any cell reaches 4.24V! Afaik this is implemented in all (brand) wheels BMS! This is very important for our safety as Li Ion cells react very badly on overcharging (venting, fire). - balancing: Afaik from different discussions/BMS fotos/datasheets normally used BMS provide passive balancing. From (2) "Passive balancing bleeds high-voltage cells on a resistor during charge in the 70–80 percent SoC curve; active balancing shuttles the extra charge from higher-voltage cells during discharge to those with a lower voltage." But (2) "A battery expert once said: “I have not seen a cell balancing circuit that works.” For multi-cell packs, he suggested using quality Li-ion cells that have been factory-sorted on capacity and voltage. This works well for Li-ion packs up to 24V; packs above 24V should have balancing. Most balancing is passive; active balancing is complex and is only used in very large systems." Some calculations and thoughts on this passive balancing from me can be found in https://forum.electricunicycle.org/topic/12618-ks16s-damaged-battery/?do=findComment&comment=213962 and the following post. Sum up: Passive balancing seems to be very ineffective and can (maybe) equalize just some very small inbalances ... What a BMS does not provide: - cell undervoltage detection: (except the new ninebot BMS - they imo report cell undervoltage!) Normal minimum manufacturer specified voltage for li ion cells is 2.5V. Cells that were discharged under 2V(1.5V) for longer times (longer than some days, ~1 week) are likely to be dead and dangerous (3)! BMS of earlier wheels (maybe) had this check and shut off the output and caused faceplants in the middle of riding! So not shutting off by cell undervoltage is an improvement for rider safety, but one has take care onself once such dead cells exist in ones battery pack - it's unsafe and a possible fire hazard! So how to check ones battery state? First thing to check is always the charger! Most of them are cheap, not too "reliable" products and could over time misadjust to deliver less voltage. Output voltage can easily be checked with a Voltmeter, so one can exlude one source of error. - Easiest way: Unfortionately just for (new) ninebot owners : cell voltages are shown in the app (Did i forget some wheel brands with this capability?) For others: Driving range getting less, acceleration capabilities diminish, pedals feel "weaker"/not so responsive are signs of weaker (degraded, old) batteries. But this same symptoms are also shown when driving at low temperatures and not fully charged batteries (roughly <50%)! - The batteries cannot be charged fully: One has to look at the voltage reported by the app (manufacturer app, wheellog for android or darknessbot for ios). Reported charge percent are inaccurate and show 100% for a quite big range (6)! Full charged li ion battery cells have 4.2V and by this 67.2V/84V for 16/20 cells in series configurations. (4) So once one cell gets "really dead" (staying under 2V all the time and cannot be charged higher) and the rest of the cells will only be charged to 4.24V the maximum voltage that can be reached for such a pack is <65.6V for a 16s configuration instead of the 67.2V! (<82.56V instead of 84V for a 20s configuration). So by the under (4) and above mentioned limitation/inaccuracies it's advisable to note the reported voltage of the fully charged new wheel and then compare it lateron! Same for the charger - measure the output voltage at this starting point and make sure the output voltage did not get less. So once this "lower charging capabilities" show for a battery pack it's unsafe and potentially dangerous to use them! Repairing such a pack is in many cases unfortionately not worth the work (7). - Premature current "cut-off" while charging: The second stage of charging a li ion cell (constant voltage) (5) is ended once the charging current gets lower than some threshold depending on the charger (?~100 mA?). If charging stops before (because one cell reached this 4.24V) this is a sign of beginning imbalance of the pack. For people having a current meters (for example the charge doctor) installed can check this. Some sources and comments: (1) https://batteryuniversity.com/learn/article/serial_and_parallel_battery_configurations (2) https://batteryuniversity.com/index.php/learn/article/bu_803a_cell_mismatch_balancing (3) "Do not boost lithium-based batteries back to life that have dwelled below 1.5V/cell for a week or longer. Copper shunts may have formed inside the cells that can lead to a partial or total electrical short. When recharging, such a cell might become unstable, causing excessive heat or show other anomalies." from https://batteryuniversity.com/index.php/learn/article/low_voltage_cut_off (4) Also voltage measurement of our wheels can be quite inaccurate - some +/- 1V were reported... So checking this already with the new wheel and good batteries can give one a starting point... As mentioned above chargers are not the most accurate, so one has to start from the maximum voltage (no load) they can deliver and subtract something like 0.3-0.7V (for the voltage drop at the BMS input protection circuitry) to get the maximum voltage the battery can be charged! So for people, who are ready to open the wheel the battery voltage can be measured at the ouput of the battery pack with a volt meter. !!!! But one has to be very careful to not short this battery pack output pins accidently while measuring - they have enourmos energy and this will lead to dangerous situations! From molten metal parts from the ?test prod?/pins itself flying around to gas from the molten cable insulation till the whole battery pack going up in fire can happen !!!! (5) https://batteryuniversity.com/learn/article/charging_lithium_ion_batteries (6) For some of the 16s wheels 66V/65.8V are still reported as 100% battery charge! (7) https://batteryuniversity.com/learn/article/how_to_repair_a_battery_pack Further readings: A great paper in regard to li ion cell matching and balancing: https://www.ti.com/download/trng/docs/seminar/Topic 2 - Battery Cell Balancing - What to Balance and How.pdf Edited February 6, 2019 by Chriull 4 2 Quote Link to comment Share on other sites More sharing options...
Tadas Posted July 30, 2019 Share Posted July 30, 2019 Very valuable post, thank you! On 2/6/2019 at 1:36 PM, Chriull said: Sum up: Passive balancing seems to be very ineffective and can (maybe) equalize just some very small inbalances ... For example in KS18XL, the balancing is only done via 100 ohm resistor, 42 mA current. It would take 2.5 hours to rebalance a 1% imbalance. 25 hours for 10% imbalance! Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 1, 2019 Share Posted August 1, 2019 On 2/6/2019 at 12:36 PM, Chriull said: The batteries cannot be charged fully: 1) One has to look at the voltage reported by the app (manufacturer app, wheellog for android or darknessbot for ios). Reported charge percent are inaccurate and show 100% for a quite big range (6)! Full charged li ion battery cells have 4.2V and by this 67.2V/84V for 16/20 cells in series configurations. (4) 2) So once one cell gets "really dead" (staying under 2V all the time and cannot be charged higher) and the rest of the cells will only be charged to 4.24V the maximum voltage that can be reached for such a pack is <65.6V for a 16s configuration instead of the 67.2V! (<82.56V instead of 84V for a 20s configuration). 3) So by the under (4) and above mentioned limitation/inaccuracies it's advisable to note the reported voltage of the fully charged new wheel and then compare it lateron! Same for the charger - measure the output voltage at this starting point and make sure the output voltage did not get less. 4)So once this "lower charging capabilities" show for a battery pack it's unsafe and potentially dangerous to use them! Repairing such a pack is in many cases unfortionately not worth the work (7). 1) My Tesla reports 82.6V in wheellog when fuly charged & 100% it says "on the bars" by then. 2) So basically that means one battery is dead and i should replace one ? both ? packs (< 82.65V coment). Cant the entire pack degrade to make it charge less? I mean 0.1V less per battery would decrease voltage by 2V right? 3) Should I measure charging voltage as it is charging (where would it be easiest to messure in that case), or just measure right in the probe when not charging? 4) One side of my battery pack feels a bit warmer (I can feel it thorugh the shell) than the other side battery pack when i charge full - LED goes green. I normally try to stop a bit earlier, but sometimes i forget and its at green a few hours. This warm side got me worried, but I dont know "how warm" the packs normally get without a charge doctor. Ive ridden about 3500 km in a bit more than a year. Dont think I have more than maybe 80-100 charges on the pack, so it seems really early to fail now? Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 2, 2019 Author Share Posted August 2, 2019 8 hours ago, Boogieman said: 1) My Tesla reports 82.6V in wheellog when fuly charged & 100% it says "on the bars" by then. As written charge % have a huge range for 100% - that value tells about nothing. How many volts showed the Tesla before (in the beginning) once fully charged? Can you measure the charger output voltage? Depending on your miltimeter the measured values can be quite off (as the voltage reported by the wheel): 8 hours ago, Boogieman said: 2) So basically that means one battery is dead and i should replace one ? both ? packs (< 82.65V coment). Here were some reports that often just one pack is "dead" and the other one still fine. So once you open the wheel and always detach one pack and try charging always only one you'll know. But then (if one is dead) there is no "real" way to ride again until you got a new pack... 8 hours ago, Boogieman said: Cant the entire pack degrade to make it charge less? I mean 0.1V less per battery would decrease voltage by 2V right? Imho not. Afaik packs with all cells degraded more or less euqaly will still charge to 4.2V - just "loose/settle" the voltage after charging faster. Just the charger could have gotten miscalibrated and charging to just too low max voltage by now. 8 hours ago, Boogieman said: 3) Should I measure charging voltage as it is charging (where would it be easiest to messure in that case), or just measure right in the probe when not charging? Measuring while charging is not simple, as there are no free reachable contact points... If you can make some additional connector with "measurement points" one could analyze the charging (mainly max charging voltage vs. max charger voltage without load) without opening the wheel. Once you opened the Tesla you can measure at the battery output (were it normally connects to the motherboard). You'd have to disconnect this point anyway once you seperate the two packs. But be carefull - at this ouput connector is no fuse or current limiting! To no short this with the probes or the multimeter accidently beeing in current measuring mode! 8 hours ago, Boogieman said: 4) One side of my battery pack feels a bit warmer (I can feel it thorugh the shell) than the other side battery pack when i charge full - LED goes green. I normally try to stop a bit earlier, but sometimes i forget and its at green a few hours. This warm side got me worried, but I dont know "how warm" the packs normally get without a charge doctor. Ive ridden about 3500 km in a bit more than a year. Dont think I have more than maybe 80-100 charges on the pack, so it seems really early to fail now? The packs should last longer - but with some bad luck one got a pack with badly matched cells and such things happen as early as in your case. Afair here were some reports of such cases already. So the steps for checking would be: - Measure the charger output. Look at the specs of your multimeter for the reachable accuracy. Maybe just your charger is miscalibrated so it does not deliver full voltage anymore! And the warming up of the one pack is "just normal" - although i would follow this further and look at least once with the wheel opened how warm warm is really directly at the battery pack! - Open the Tesla and disconnect the batteries from the mainboard. At this connector the output voltage of the battery cells can be measured. As written above - to this careful and diligently, there is no fuse or overcurrent protection! The difference should be maximum some 0.3-0.5V, if a protection diode is at the charging input. - Split the two packs and measure their voltages. Charge both seperately and look if one can be charged fully while the other one cannot be charged. If one is charging less then the other you can cut open the pack to get access to the single cells and measure the single cell voltages. If some blocks are below 1.5 - 2V i'd recommend to replace the whole pack. If there are just some cells at lower voltages one could try to recharge them one by one with some "normal single li ion charger" to get the whole pack balanced again. @Marty Backe - you could think of some more tipps/recommendation with opened GW wheels. You have more experience with them?! Or other knowingly of GW's! 1 1 Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 2, 2019 Author Share Posted August 2, 2019 @Boogieman - after rethinking this procedure, maybe the best/easiest and safest way could be: Since one pack gets warm while charging it has to be examined! So you need to open the wheel. Afaik the batteries are "just" in parallel - so one pack can be disconnected (at borh sides). @Marty Backe or some Tesla owner can confirm this? Best is to start with dusconnecting the pack that warms up. Then power up the Tesla. If it now reports more voltage you're done - the other pack is dead... If you want, you can examine this bad pack in detail as written in the above post. If the voltage stays about the same charge the wheel with just this one pack. If it charges up to normal, higher voltages reported by the Tesla one knows again that the other pack is dead. If not, repeat this with just the other pack attached. The procedure of attaching packs to the mainboard is best explained by someone who is experienced with this (not hesitant, maybe best with an anti spark plug - @Marty Backe again?! ) 2 1 Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 2, 2019 Author Share Posted August 2, 2019 @Boogieman: second addentum - i should gather my thoughts better before posting Watch the temperature of the packs while charging. The presumably good one should not really noticable warm up while charging with a 1.5-2.5A stock charger. If the other pack warms up considerably more or especially has some "hot spots" stop the charging and dispose it safely! 1 1 Quote Link to comment Share on other sites More sharing options...
Marty Backe Posted August 2, 2019 Share Posted August 2, 2019 22 hours ago, Boogieman said: 1) My Tesla reports 82.6V in wheellog when fuly charged & 100% it says "on the bars" by then. 2) So basically that means one battery is dead and i should replace one ? both ? packs (< 82.65V coment). Cant the entire pack degrade to make it charge less? I mean 0.1V less per battery would decrease voltage by 2V right? 3) Should I measure charging voltage as it is charging (where would it be easiest to messure in that case), or just measure right in the probe when not charging? 4) One side of my battery pack feels a bit warmer (I can feel it thorugh the shell) than the other side battery pack when i charge full - LED goes green. I normally try to stop a bit earlier, but sometimes i forget and its at green a few hours. This warm side got me worried, but I dont know "how warm" the packs normally get without a charge doctor. Ive ridden about 3500 km in a bit more than a year. Dont think I have more than maybe 80-100 charges on the pack, so it seems really early to fail now? FYI, I just measured (via WheelLog) my Monster (2-1/2 years old) and MSX. Both read 82.4-volts @ 100-percent. One of my ACM2's reads 83.2-volts. They would ride just as far as they always have. I wouldn't worry about it. If you really had a problem you would need to directly measure the battery packs relative to the actually charger voltage. I just don't know how much you can depend on the voltage values being generated by the control board circuitry. 1 Quote Link to comment Share on other sites More sharing options...
LanghamP Posted August 3, 2019 Share Posted August 3, 2019 Are lithium ion batteries recycled into new batteries, or are they simply disposed of in landfills? Considering the resources to make lithium ion batteries, I hope they can be reused. 1 Quote Link to comment Share on other sites More sharing options...
wheelr Posted August 3, 2019 Share Posted August 3, 2019 34 minutes ago, LanghamP said: Are lithium ion batteries recycled into new batteries, or are they simply disposed of in landfills? Considering the resources to make lithium ion batteries, I hope they can be reused. Some power companies repurpose used batteries for storage on the grid. I'm not aware of any recycling beyond that. Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 4, 2019 Share Posted August 4, 2019 (edited) On 8/2/2019 at 10:21 AM, Chriull said: 1) How many volts showed the Tesla before (in the beginning) once fully charged? 2) Can you measure the charger output voltage? 3) Depending on your miltimeter the measured values can be quite off (as the voltage reported by the wheel): 4) Here were some reports that often just one pack is "dead" and the other one still fine. So once you open the wheel and always detach one pack and try charging always only one you'll know. 5) But then (if one is dead) there is no "real" way to ride again until you got a new pack... 6) Imho not. Afaik packs with all cells degraded more or less euqaly will still charge to 4.2V - just "loose/settle" the voltage after charging faster. 7) Just the charger could have gotten miscalibrated and charging to just too low max voltage by now. 8) Measuring while charging is not simple, as there are no free reachable contact points... If you can make some additional connector with "measurement points" one could analyze the charging (mainly max charging voltage vs. max charger voltage without load) without opening the wheel. 9) Once you opened the Tesla you can measure at the battery output (were it normally connects to the motherboard). You'd have to disconnect this point anyway once you seperate the two packs. But be carefull - at this ouput connector is no fuse or current limiting! To no short this with the probes or the multimeter accidently beeing in current measuring mode! The packs should last longer - but with some bad luck one got a pack with badly matched cells and such things happen as early as in your case. Afair here were some reports of such cases already. So the steps for checking would be: - Measure the charger output. Look at the specs of your multimeter for the reachable accuracy. Maybe just your charger is miscalibrated so it does not deliver full voltage anymore! 10) And the warming up of the one pack is "just normal" - although i would follow this further and look at least once with the wheel opened how warm warm is really directly at the battery pack! - Open the Tesla and disconnect the batteries from the mainboard. At this connector the output voltage of the battery cells can be measured. As written above - to this careful and diligently, there is no fuse or overcurrent protection! 11) The difference should be maximum some 0.3-0.5V, if a protection diode is at the charging input. - Split the two packs and measure their voltages. Charge both seperately and look if one can be charged fully while the other one cannot be charged. 12) If one is charging less then the other you can cut open the pack to get access to the single cells and measure the single cell voltages. If some blocks are below 1.5 - 2V i'd recommend to replace the whole pack. If there are just some cells at lower voltages one could try to recharge them one by one with some "normal single li ion charger" to get the whole pack balanced again. @Marty Backe - you could think of some more tipps/recommendation with opened GW wheels. You have more experience with them?! Or other knowingly of GW's! 1) don't know, never really looked on voltage in wheelog, but recently discovered it said 67V due to me not changing to 84V in settings (=scaling wrong) and i can't say i trust apps to be correct, as they are based on the original Chinese app i guess. Noone really knows what happens inside the circuitry and then sent via bt to an app with 100's of versions :-) some of which @Marty Backe allready proven to show wrong percentage reading on nikola (and thus likely also wrong voltage) 2) yes, got a fairly accurate multimeter (or?). Just need to find which pin is +/- on the 4 pin connector 3) its a good allround multimeter (measures transistors, temp probes, capacitors and whatnot). Not a fluke but....yeah lets check dc accuracy in leaflet Range: 40-400 V DC, resolution: 0.1V, accuracy +/-0.5% of reading +/- 3 digits. So at 80V - accuracy is +/-0.4V ???+/- 3 digits??? = Total +/- 0.7V if i understand it right on the digit part 4) You mean disconnect one pack and charge the other CONNECTED still to wheel = charging wheel with only one pack? So they are in parallell allways? I didnt know. 5)If they are in parallel, i could ride but ill have less range right :-) just have to tape up the free connector to prevent accidental shorts 6) when i checked the app, it was actually 24h after full charge.maybe they settled? 7) I will meassure the idle output voltage (no load attached) 8) Yes, that was my thought. I need a charge doctor allready ;-) 9) I can connect the packs one by one, simple quick connectors, and measure 10) I just found it strange its allways the same side that gets warm. Ill put a xiaomi temp probe on next time i charge, covers off. And try to feel which battery is the hot one. It is for sure a hot spot, which made me worried in the first place that one battery is heating up. 11) accuracy of meter is +/-0.4v as i see it at point 3) but it should be in the same "range" ao i can still see diff between packs, albeit not the defintive voltage (+/-0.4 off) 12) oh, you can measure one cell at a time, i thought that there were a few in parallel that were put in series. Seems like a heck lot more than 20cells. But maybe 20/pack *2 in parallell then? HUGE THANKS for your time to reply guys. I am a bit worried as i charge in my apartment, often night time. Sometimes charger is green for a few hours (slerpy) before i disconnect (which seba says is good as charging is still ongoing) and sometimes i try to disconnect around 90% (disconnect and power on wheel for reading, charge more if still at <90%) to increase charge cycles but that i started with just recently (1-2months ago). I will measure the charger first. Then i will charge full (really fulk, green led for like 4 hours) and check again and if still low voltage measure each battery on its own. At least i will se if there is a difference as error should be the same in multimeter regardless of pack Thanks Boogie Edited August 4, 2019 by Boogieman Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 4, 2019 Share Posted August 4, 2019 (edited) On 8/2/2019 at 11:39 AM, Chriull said: @Boogieman: second addentum - i should gather my thoughts better before posting Watch the temperature of the packs while charging. The presumably good one should not really noticable warm up while charging with a 1.5-2.5A stock charger. If the other pack warms up considerably more or especially has some "hot spots" stop the charging and dispose it safely! I removed the cover and started charging. Charger started on green = fully charged. The pack only heats up when fukly charged and more. That made me thinking, could it be the BMS that gets hot as its close to full? Also i can feel already (5min charge) on the pack that its a bit warmer in one spot, but its just where the wires enter the pack, thats where i would put a BMS. Also the pack is a bit bulgier right there (either BMS or battery about to go poof :-S ) Its slightly warmer right now where the white temp puck is located, also where wires enter, where the BMS likely is??? And i guess there are a couple of resistors on the BMS that could get warm? Slightly warmer = body temp, while rest.of pack feels cool like room temp. Could be that simple, but ill let it charge a few hours and keep track. Then check app, then check charger, then well i guess disconnect one battery, but is there one bms on each batt pack? Can i charge the one without BMS (with bms pack disconnected) if there is only a bms on one pack? I mean it doesnt fell like i know what im doing DISCONNECTING one battery and charge only ONE like nothing has happened :-) Edited August 4, 2019 by Boogieman 1 Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 4, 2019 Share Posted August 4, 2019 Definitely seems like BMS is at the hot spot, or what are those sauare things if not components of some kind? Quote Link to comment Share on other sites More sharing options...
Marty Backe Posted August 4, 2019 Share Posted August 4, 2019 2 hours ago, Boogieman said: I removed the cover and started charging. Charger started on green = fully charged. The pack only heats up when fukly charged and more. That made me thinking, could it be the BMS that gets hot as its close to full? Also i can feel already (5min charge) on the pack that its a bit warmer in one spot, but its just where the wires enter the pack, thats where i would put a BMS. Also the pack is a bit bulgier right there (either BMS or battery about to go poof :-S ) Its slightly warmer right now where the white temp puck is located, also where wires enter, where the BMS likely is??? And i guess there are a couple of resistors on the BMS that could get warm? Slightly warmer = body temp, while rest.of pack feels cool like room temp. Could be that simple, but ill let it charge a few hours and keep track. Then check app, then check charger, then well i guess disconnect one battery, but is there one bms on each batt pack? Can i charge the one without BMS (with bms pack disconnected) if there is only a bms on one pack? I mean it doesnt fell like i know what im doing DISCONNECTING one battery and charge only ONE like nothing has happened :-) Maybe I'm out of touch with this thread. but all battery packs get warm when you are charging them and I'm sure there are hot spots. Each battery pack has its own BMS hardware. Strip away the blue strink wrap and you'll find the circuit board. You can always remove a parallel pack with zero impact to the wheel other than reduced range. 1 Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 5, 2019 Author Share Posted August 5, 2019 On 8/4/2019 at 3:27 AM, Boogieman said: 1) don't know, never really looked on voltage in wheelog, but recently discovered it said 67V due to me not changing to 84V in settings (=scaling wrong) and i can't say i trust apps to be correct, as they are based on the original Chinese app i guess. That's the way GW wheels report the voltage. The report as if it were a 67.2V wheel. Imho wheellog has a switch somewhere to change to 84V wheel. But one can just take the voltage and take it times *84/67.2 and has the right value. For the 100V version it would be *100,8/67.2 On 8/4/2019 at 3:27 AM, Boogieman said: 3) its a good allround multimeter (measures transistors, temp probes, capacitors and whatnot). Not a fluke but....yeah lets check dc accuracy in leaflet Range: 40-400 V DC, resolution: 0.1V, accuracy +/-0.5% of reading +/- 3 digits. So at 80V - accuracy is +/-0.4V ???+/- 3 digits??? = Total +/- 0.7V if i understand it right on the digit part Sounds nice. If at has one digit after the decimal point while measuging 80V (i.e. 80.0V) the +/- 0.7V should be right. On 8/4/2019 at 3:27 AM, Boogieman said: 6) when i checked the app, it was actually 24h after full charge.maybe they settled? Yes. Could be. Just check after fully charging again! On 8/4/2019 at 3:27 AM, Boogieman said: 7) I will meassure the idle output voltage (no load attached) Great. Maybe together with your answer to the point above everything is just fine! On 8/4/2019 at 3:27 AM, Boogieman said: 10) I just found it strange its allways the same side that gets warm. Ill put a xiaomi temp probe on next time i charge, covers off. And try to feel which battery is the hot one. It is for sure a hot spot, which made me worried in the first place that one battery is heating up. 12) oh, you can measure one cell at a time, i thought that there were a few in parallel that were put in series. Seems like a heck lot more than 20cells. But maybe 20/pack *2 in parallell then? One can measure the 20 paralleled cell packs voltages. If one of this paralleled packs (3 in parallel?) have a voltage below 1.5 - 2.0V they are all dead. But one has to open the battery pack - remove the plastic wrap. I would not recommend this as long as the pack seems to be fine! On 8/4/2019 at 3:27 AM, Boogieman said: HUGE THANKS for your time to reply guys. I am a bit worried as i charge in my apartment, often night time. Get some smoke detectors! On 8/4/2019 at 3:27 AM, Boogieman said: I will measure the charger first. Then i will charge full (really fulk, green led for like 4 hours) and check again and if still low voltage measure each battery on its own. At least i will se if there is a difference as error should be the same in multimeter regardless of pack If it charges up normally you won't need to disconnect the batteries and charge them seperately - just if you want do ... And remember to just connect them back in parallel only if they have the same voltage! (the less difference, the less current flowing...) On 8/4/2019 at 3:55 AM, Boogieman said: I removed the cover and started charging. Charger started on green = fully charged. The pack only heats up when fukly charged and more. That made me thinking, could it be the BMS that gets hot as its close to full? Also i can feel already (5min charge) on the pack that its a bit warmer in one spot, but its just where the wires enter the pack, thats where i would put a BMS. Also the pack is a bit bulgier right there (either BMS or battery about to go poof :-S ) Its slightly warmer right now where the white temp puck is located, also where wires enter, where the BMS likely is??? And i guess there are a couple of resistors on the BMS that could get warm? Slightly warmer = body temp, while rest.of pack feels cool like room temp. Sounds everything ok. Some components of the BMS get warmer, but nothing really hot. The only real concern would have been if a single/some li ion cells get hot! On 8/4/2019 at 4:16 AM, Boogieman said: Definitely seems like BMS is at the hot spot, or what are those sauare things if not components of some kind? Some components of the BMS - should be fine and nothing to worry about. 1 1 Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 6, 2019 Share Posted August 6, 2019 (edited) 1a) Measured charger output voltage to 82.5V (+/-0.7V then due to multimeter) without load. That does indeed seem to be on the low side right? But the led stays green when measuring so could it be some kind of maintenance charge that is lower? Or it should meassure 84V? Pin out was found to be GX16-4 1: V+ ; 2: 0V Info from http://hobby16.neowp.fr/2016/11/20/charger-customization/ 1b) Is it bad for the battery to charge at a lower voltage? From what i read the will be no cell balancing unless wheel is charged fully every 10th time and 82.5V is 90% charge which is good for charge cycles, but still battery needs to be fully charged every now and then for balancing. Marked in cyan a bit down at below page http://hobby16.neowp.fr/2016/11/12/autocut-adjustment/ 1c) wheellog report 82.7 V now after charging. Seems about right it only charges 82.5 (+/- 0.7 meter error). 2) Today i rode the wheel to about 20% 3) Waited an hour before starting to charge as the battery was still warm after the ride and actually still a bit warm after one hour to, also here on the right side( i read you should not charge directly after a ride, why i dont know but maybe the temp of the batteries or some normalisation?) 4) Charged it up maybe 8-10 hours and felt the battery pack was warm through covers so removed everything to see where heat came from and how hot it was. Well, it was more than my body temp from the feel, maybe 40+ hard to say, i only had a xiaomi air temp meter that i placed with the vents toward the hot spot and it reported around 35°C but it felt warmer than my skin (apartment is about 20°C and rest of battery pack is cool) but i stooped charging as i felt it was hot and started measuring, so likely a bit more than 35. In the image below i drew the hot area with a sharpie, one end hotter going towards cooler as written at the other end of the encircled area. EDIT: I stopped charging and allowed it to cool off. Then restarted and it seems like its a hot spot at the end of the battery pack and likely the rest of the heat signature is something spreading it upwards (but strangely not sideways) making it cooler the further away you get. 5) Im not sure of the orientation of the batteries or the length, but maybe someone can get something out of this? Its the right side battery. The width of the pack is 135mm, length 200mm. Both of the short edges angled (as if batteries were stacked in two layers with them packed with 50% offset to take less space) so i think the batteries are placed in the other direction than the heat signature (which would be good as that would indicate something else is heating than the batteries unless several batteries are hot just in the middle). But hey i could be wrong on the direction, it seems like there is some hard plastic under the shrink probably to protect the batteries. 6) I also noted that the glue from the self adhesive had stuck only on this hot ares when i pulled the adhesive foam off, so its likely always that area that gets hotter. 7) going to measure the packs CAREFULLY tomorrow. 8) Have fire alarms, gonna get a huge extinguisher tomorrow. I read up on the subject and you should use a ABC or BC extinguisher as li ion battery fire is classed B. https://resources.impactfireservices.com/how-do-you-put-out-lithium-ion-battery-fire They sell lithium fire extinguishers, but those are NOT for lithium batteries as the batteries contains such a tiny amount of lithium according to above article (hope they are right) Lucky for us as those extinguishers are expensive as F... (About 8 times the price of a normal extinguisher). Though should a pack catch fire i dont expect to be able to put it out, but if lucky and only one battery goes poof maybe keep it contained at that by cooling it (have a co2 extinguisher) and then use powder to contain it, maybe a fire blanket. Well I don't want to go there, i have seem an electric scooter catch fire and its nothing to toy with and those batteries are tiny in comparison. Edited August 6, 2019 by Boogieman 1 Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 6, 2019 Author Share Posted August 6, 2019 6 hours ago, Boogieman said: 1a) Measured charger output voltage to 82.5V (+/-0.7V then due to multimeter) without load. That does indeed seem to be on the low side right? Yes. Quote But the led stays green when measuring so could it be some kind of maintenance charge that is lower? No Quote Or it should meassure 84V? Yes. Between 83.3 and 84.7. Your measured 82.5V mean 81.8V to 83.2V leading to a single cell voltage between 4.09V and 4.16V. Quote 1b) Is it bad for the battery to charge at a lower voltage? From what i read the will be no cell balancing unless wheel is charged fully every 10th time and 82.5V is 90% charge which is good for charge cycles, but still battery needs to be fully charged every now and then for balancing. The BMS used in your wheel is not known in detail (as of most wheels). Common BMS have some overvoltage protection for each single cell around 4.28V and activate the balancing resistors at 4.2V +/-0.05%. So if one has 1 weak paralleled cell pack this gets discharged and charged first. So it will reach tjis 4.2V while the others still are at (82.5-4.2)/19=4.12V. the the balancing resistor at this weak cell gets activated and causes it to get charged a bit slower. Depending on the charge current/capacity difference and balancing resistor this wrak cell could now get charged up to 4.28V until the BMS cuts off the charging input. In this case the other cells would have to drop to (82.5-4.28)/19=4.11V - which is an quite impossible outcome. The other outcome could be just that the weak cell stays at 4.2V (impossible perfect balancing resistor) and the others are charged to (82.5V-4.2V)/19=4.12V. The thruth will be somewhere inbetween, and all the min max values from the measurements should be regarded to come to the real possible outcomes... So, no there is still balancing happening. And yes, with 84V max charger voltage better balancing happens, as with your 82.5 the first ever balancing happens ones the cells have alteady a difference of 4.2V-4.12V=0.08V. Quote ( i read you should not charge directly after a ride, why i dont know but maybe the temp of the batteries or some normalisation?) Because the batteries degrade faster if charged at higher temperatures. But a bit warmer does not really matter. Quote 4) Charged it up maybe 8-10 hours and felt the battery pack was warm through covers so removed everything to see where heat came from and how hot it was. Well, it was more than my body temp from the feel, maybe 40+ hard to say, i only had a xiaomi air temp meter that i placed with the vents toward the hot spot and it reported around 35°C but it felt warmer than my skin (apartment is about 20°C and rest of battery pack is cool) but i stooped charging as i felt it was hot and started measuring, so likely a bit more than 35. 35 to 40 is not "hot" - everything is ok with this. The batteries hold a huge amount of energy and charging is not lossless - so quite some energy over this 8-10hours is converted to heat. Nothing to worry about. Quote 6) I also noted that the glue from the self adhesive had stuck only on this hot ares when i pulled the adhesive foam off, so its likely always that area that gets hotter. Could for instance be the mosfets in the charge input line (which can perform the charge cut off in case of cell overvoltage). They just dissipate very little power, but cannot really get rid of it in zhe plastic wrapping so they heat up a couple of degrees. 1 2 Quote Link to comment Share on other sites More sharing options...
Chriull Posted August 6, 2019 Author Share Posted August 6, 2019 7 hours ago, Boogieman said: 5) Im not sure of the orientation of the batteries or the length, but maybe someone can get something out of this? Its the right side battery. The width of the pack is 135mm, length 200mm. Both of the short edges angled (as if batteries were stacked in two layers with them packed with 50% offset to take less space) so i think the batteries are placed in the other direction than the heat signature (which would be good as that would indicate something else is heating than the batteries unless several batteries are hot just in the middle). But hey i could be wrong on the direction, it seems like there is some hard plastic under the shrink probably to protect the batteries. The battery packs i have seen so far consist of a PCB on which the cells are fixed (the nickel strips from the cells are soldered to the PCB). This is the first mechanical fixation. The second is the plastic wrap around the cells and the PCB which prevents movement/vibration relative between the cells and the PCB so the nickel strips do not break. Here in this post some fotos can be found showing this (for a much smaller pack): If you feel that you can establish such an plastic wrapping again in the same quality, maybe the only way to go for the "sake of your mind" is to open the pack. It's more or less the only way to get the single cell voltages, inspect them and the overall "build-up" quality. Although i'd say from all i've read here that the pack is ok... 1 Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 7, 2019 Share Posted August 7, 2019 (edited) On 8/6/2019 at 9:33 AM, Chriull said: 35 to 40 is not "hot" - everything is ok with this. The batteries hold a huge amount of energy and charging is not lossless - so quite some energy over this 8-10hours is converted to heat. Nothing to worry about. Could for instance be the mosfets in the charge input line (which can perform the charge cut off in case of cell overvoltage). They just dissipate very little power, but cannot really get rid of it in zhe plastic wrapping so they heat up a couple of degrees. The battery pack doesnt heat up AT ALL until the last % before green LED turns on and then continue to heat when green LED is on. (I have been touching the packs once an hour lately to find out). And then only one single spot is heated, which seems to transfer heat upward on the pack as in the image earlier. Left side battery pack is still room temp 22°C and the same with all the surrounding areas of the right pack that heats at one point/area. That means my packs doesnt heat up at all during charge except this one spot when it is close to max charge (as per delivered by charger) So, now should i dismantle the charger and increase voltage to 83.3 (taking height for measurement error +/-0.7 on instrument? To make it start balancing or is it so "OFF" now that it will never be able to do that properly? Or maybe even better (and a lot less risky) buy one more charger that output 84V and use that one every 10th charge for cell balancing and this charger to enlongate battery pack life at 90% charge? (since I dont have a charge doctor) I really appreciate all the help here especially from @Chriull that have patience with my questions (not even all teachers had that patience in school 😊) I try to read up as much as possible but this deep understanding of batteries would have taken me the entire summer to find and read. Going shopping for an extuinguisher now and then time to measure battery pack voltages :-) Edited August 7, 2019 by Boogieman Quote Link to comment Share on other sites More sharing options...
Tadas Posted August 7, 2019 Share Posted August 7, 2019 Perhaps it's just doing the balancing. Total BMS rebalancing might be putting out 3W of heat, that would heat up the BMS quite a lot. The battery pack stores incredible amounts of energy, let's say 800 Wh and is only being rebalanced by small bleed which heats the BMS by up to 3W power (19 cells each doing 0.16W). Keep full charge for ~50 hours, either always plugged in or plug in every 4-8 hours. It may take that long to rebalance the cell pack. Having a smoke detectors is always wise. 2 Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 7, 2019 Share Posted August 7, 2019 (edited) 9 hours ago, Tadas said: 1) Perhaps it's just doing the balancing. Total BMS rebalancing might be putting out 3W of heat, that would heat up the BMS quite a lot. The battery pack stores incredible amounts of energy, let's say 800 Wh and is only being rebalanced by small bleed which heats the BMS by up to 3W power (19 cells each doing 0.16W). 2) Keep full charge for ~50 hours, either always plugged in or plug in every 4-8 hours. It may take that long to rebalance the cell pack. 3) Having a smoke detectors is always wise. 1) Problem is charger voltage is too low and battery pack heats up ZERO when charging from 10% to around 95% (in app) then starts heating. I tested to start charging after making just a 1km run to drain battery slightly and then charge full plus a few hours with green LED after that and oddly - No heat this time? 2) You mean after I buy a charger that actually outpute 84V (not 82.5V like mine) ? 3) I do and bought a 6Kg max spec ABC extinguisher today + a fire blanket with silicon treatment (supposed to be the best and meets the current standards). 100 Euro well spent should someting go wrong, though I still dont think I would have a smidge of a chance to put out a full blown Li Ion fire - but if alarm alerts early enough I might contain it and throw the wheel out the window or at least into the stairwell that is all stone :-S Edited August 7, 2019 by Boogieman Quote Link to comment Share on other sites More sharing options...
Tadas Posted August 8, 2019 Share Posted August 8, 2019 Ideally you'd have a 84 V charger, but balancing is being done by the BMS on a cell by cell basis. Even at 82 V battery pack you can have cells that are above 4.2 V, those cells will be drained by the BMS and the BMS will heat up. Let it balance for 50 hours, especially if you have never done it before. Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 8, 2019 Share Posted August 8, 2019 (edited) 1 hour ago, Tadas said: Ideally you'd have a 84 V charger, but balancing is being done by the BMS on a cell by cell basis. Even at 82 V battery pack you can have cells that are above 4.2 V, those cells will be drained by the BMS and the BMS will heat up. Let it balance for 50 hours, especially if you have never done it before I have allowed it to balance 24-72h when not charging several times, ie charged full (green), disconnect and not ride for a couple of days. Charge balance. Normally stop charging when i notice the green LED come on (=0-10h after green led on as i often charge night time, normally 1-5 hours after though). So basically you mean that i should charge it with green LED on for 50hours (meaning it charges with a low current and BMS hopefully activates at my low charge voltage (82.5)?). It should at least cutoff if i have over voltage on a cell, meaning it could be a test for the BMS, but how can i know that? Edited August 8, 2019 by Boogieman Quote Link to comment Share on other sites More sharing options...
Tadas Posted August 8, 2019 Share Posted August 8, 2019 1 minute ago, Boogieman said: Normally stop charging when i notice the green LED come on (=0-10h after green led on as i often charge night time, normally 1-5 hours after though). There should not be an imbalance then... But still I suggest you leave it plugged in once for 50 hours. After that 1-2 hours plugged in after each charge. If that doesn't help, you might have a faulty cell Quote Link to comment Share on other sites More sharing options...
Tadas Posted August 8, 2019 Share Posted August 8, 2019 Next step is get a 84 V charger and see if the battery charges fully. Quote Link to comment Share on other sites More sharing options...
Boogieman Posted August 8, 2019 Share Posted August 8, 2019 (edited) 12 minutes ago, Tadas said: Next step is get a 84 V charger and see if the battery charges fully. I will need that 84v regardless I guess as I cant possibly have more than 82.5V on each pack if charger outputs 82.5V, right? (As there is one BMS per pack) So if i measure my packs separate and both are 82.5 packs might be ok, but i dont know until i charged them both to 84V - aka buy a new charger (or open mine and increase voltage to 83.3V)? And i guess with the 84V charger i have to charge those 50-60hours as the cells have never been charged that high before? I remeassure my charger output voltage to 82.3V....after changing the 9v battery as i noted backlight was dead but no low battery warning(it had plenty amps but voltage was down to 7.5V strange) Edited August 8, 2019 by Boogieman Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.