dirkclod
Premium Pilot
Haven't really checked but don't think so . I will charge one tomorrow and see though !!
Battery problems
- Thread starter Squiddley
- Start date
dirkclod
Premium Pilot
My battery's are full right now or I would check but don't remember mine getting that way.
I usually just mess with the plug though .
okay let me know what ya find out next time you charge, that would be great.....maybe my charger is screwing up the batteries, charged my new battery tonight seemed to work okay but who knows, maybe its not supposed to get warm
dirkclod
Premium Pilot
Hi squid,
I've just finished charging my batteries up & yes the psu (charger) does get very warm, this is normal.
Hope this helps..
well guess my one battery in finished went out tonight to give it a try had it fully charged got down to about 60% and then she just started dropping real quick barely had time to get it on the ground
That thing with pressing the buttons to see how much battery life is only telling you have many charges its had out of the 300 charges the batteys can get before they brick them selves when the get down to 0% of there life. (not that any one has ever gotten even close to that many before there smart battery was already toasted any way)
The remaining health in no way shape or form tells you any thing about the real health of a battery and is just another way that they count the charges every 3 charges ticks off 1% of the remaining life number. You could have a battery that is in perfect condition but had 99 charges cycles on it and it will say its only got 70% remaining life even if there is not a thing wrong with it. OR you could have a battery that crapped the bed after 3 charges and it will tell you that the battery has 99% remaining health even if its toast and completely unusable. So its sort or silly and also completely useless for telling you any thing at all about a batteys health Being its just telling you how many charges its had out of the max limit of 300 charge cycles That the smart battery's firm ware will allow.
also when a cell gets locked out from the not all that smart smarty pants battery the leds that show you the % of charge only read the cells that are still not locked out and still working. So you could have a battey with 2 cells that are locked out and the one cell thats working will charge up fully and in about 1/3 of the time as it would take to charge a good battery and then it will tell you that you are 100% charged even tho its full of crap and is just the one cell thats charged fully.
So if I ever have a battery get done charging way sooner then it should thats when its time to make sure to plug in to the naza assistant and actually look at the cells voltage and the total volts as well as the ma's to know if its toast or not. being thats the only way to really know.
The remaining health in no way shape or form tells you any thing about the real health of a battery and is just another way that they count the charges every 3 charges ticks off 1% of the remaining life number. You could have a battery that is in perfect condition but had 99 charges cycles on it and it will say its only got 70% remaining life even if there is not a thing wrong with it. OR you could have a battery that crapped the bed after 3 charges and it will tell you that the battery has 99% remaining health even if its toast and completely unusable. So its sort or silly and also completely useless for telling you any thing at all about a batteys health Being its just telling you how many charges its had out of the max limit of 300 charge cycles That the smart battery's firm ware will allow.
also when a cell gets locked out from the not all that smart smarty pants battery the leds that show you the % of charge only read the cells that are still not locked out and still working. So you could have a battey with 2 cells that are locked out and the one cell thats working will charge up fully and in about 1/3 of the time as it would take to charge a good battery and then it will tell you that you are 100% charged even tho its full of crap and is just the one cell thats charged fully.
So if I ever have a battery get done charging way sooner then it should thats when its time to make sure to plug in to the naza assistant and actually look at the cells voltage and the total volts as well as the ma's to know if its toast or not. being thats the only way to really know.
Very interesting; reducing batt life indicator by 1% every 3 charges is not a normal way to provide life remaining info.... How do you know this is what they do?
What do you mean by cells getting locked out? Are you saying there is a way inside our smart batteries to bypass or shut off individual cells?
For I know it cause it just is all it is is redundant count of the charge cycles. But any one can verify it just by taking any of there smart battery's and seeing plugging it in to the naza assistant and seeing the number of charge cycles its had and them look at the life or health number and see that it will always be a % thats reduced by 1 % for every 3 charges. DJI even says that when one ever were to get to 0% after the 300th charge cycle its supposed to pop up a window telling you the battery can no longer be used and that its time to buy a new one. But thats not some thing i ever seen yet or ever even heard of any one seeing cause if there is any one out there thats ever gotten a dji battery to go 300 charges before crapping the bed they havent come to any site and posted the results of being the only person to have one last that long.
Tho there are thos that have hacked the smart battery's to use other lipo battery's that have gotten a total of 300 charges out of the smart circuit board out of the dji battery's and then the board is not useable. So now there are hacks were people have figured out a way to put a resister in the part that does the count so that it only counts every other charge as I charge and then it takes 6 charges to tic of 1% of the life counter. So they can get them to go for 600 cycles before the board bricks its self.
Yes infact all smart batterys have a protection circuit in them that shuts off the flow of power to any cell that ever goes above or below a certain cut off voltage that the protection circuit has to cut off the flow of power in or out of that cell.
Even single cell lipos that have the smart circuits in them the only way to fix a cell or unlock it is to either put some power in if its locked from being to low of a voltage. Or to drain some out directly intill it falls below the high voltage lockout voltage. Tho in the dji battery's it would involve busting the battery's cage open and getting to top off it. Then in side its just a regular lipo 3x2 cell* battery with some circuit boards in between it and the plug. The board has all the so called smart circuity on it. Which is basically a charger and balancing circuit it on it. and some stuff for the data connection and the protection circuit. * 3x2 cell is basically 6 cells but wired so there are 3 cells that each have two battery's wired together in parallel. and then each group of 2 are wired in series to the other groups of 2.
J James, thanks for the reply. I had not correlated the "number of charges" to 1/3% less life - very interesting.... And as we all know now, this % life left has nothing to do with killing the batteries - most 'bad' batteries still show in the 90% life area. So this is interesting but not of any value to us trying to find out why the batteries actual appear to fail early.....
this mystery is why I propose someone investigate the short life.
Your answer to my follow question has me still intrigued. I have worked with lithium high power batteries now for some time. I was able to correct charging and algorithm mistakes in some high profile mfgrs and BMS mfgrs systems by simply following engineering lab 101 note taking procedures and methodically documenting what was happening and pointing out how it did not match the beliefs of the mfgr.... Since no one has proposed to do THIS yet, that is why I asked for 'bad' batteries - to do the same methodical testing and documenting results to see if they pop out something so far unseen.
Anyway, on 2nd question, you suggest that our dji batteries have a way to shut off INDIVIDUAL CELLS. I disagree.
Assuming I was behind the times in latest battery technology designs, I researched battery lockout functions. I think I can fairly say they have not changed in the last couple years, and there is NO INDIVIDUAL CELL LOCKOUT CAPABILITY IN EXISTENSE.
Please prove me wrong! Show me what I am missing! I would love the day to come that technology can make a FET switch with low enough ON resistance to be put in series with a good lipo cell! I did not research FET on resistance, so I again may be outdated, but I think you will find the minimum ON resistance possible today is still in the hi single mohm digits, typically 10+mohm.... this make them totally impossible to put in a batteries series path.
I believe this is why "lockout" is actually only a SINGLE or TWO FETS in series with the WHOLE pack: one for opening if voltage of a SINGLE cell goes too low, or the other open if a single cell goes too high. No individual lockout possible.
Said another way, if you put these 10mohm series resistance FETS in series with each cell, you will end up with a battery resistance so high it cannot produce the 3-20x peak current required to operate our quads....
An example is this smart battery device that shows this in detail:
http://www.ti.com/lit/ds/symlink/bq77910a.pdf
This is why the technology today is still whimpy trying to balance cells against each other : we cannot put a FET in series with each cell so we put one in PARALLEL around it with a series 5-10ohm resistor - and turn it on to SHORT out the cell during charging, thereby charging it a tad LESS than then other cells. So to balance, the HI voltage cells get slightly bypassed by this small 10-100ma current, letting them charge slower than the other cells....
So it is my thesis than the low cycle count life of our dji batteries MAY be due to a cell falling behind in charge and therefore, with this whimpy bypass charge scheme used today, never charging fully again before the others do and shut off the charger. Premature dji battery failure caused by ONE cell simply notbeing able to fully charge ever again!
THIS is my idea, and why I want 'bad' batteries sent to me to test! If I am wrong, no damage to anyone! The bad batteries sent to me to test were of no use to their previous owner anyway! But it may show that we can get those 300 cycles -as we should - by simply putting each cell on a multi cell lipo charger....
ps: sorry if my thoughts are too random - I have been in engineering meetings all day, came home, and wife wants my attention while I want to write this
this mystery is why I propose someone investigate the short life.
Your answer to my follow question has me still intrigued. I have worked with lithium high power batteries now for some time. I was able to correct charging and algorithm mistakes in some high profile mfgrs and BMS mfgrs systems by simply following engineering lab 101 note taking procedures and methodically documenting what was happening and pointing out how it did not match the beliefs of the mfgr.... Since no one has proposed to do THIS yet, that is why I asked for 'bad' batteries - to do the same methodical testing and documenting results to see if they pop out something so far unseen.
Anyway, on 2nd question, you suggest that our dji batteries have a way to shut off INDIVIDUAL CELLS. I disagree.
Assuming I was behind the times in latest battery technology designs, I researched battery lockout functions. I think I can fairly say they have not changed in the last couple years, and there is NO INDIVIDUAL CELL LOCKOUT CAPABILITY IN EXISTENSE.
Please prove me wrong! Show me what I am missing! I would love the day to come that technology can make a FET switch with low enough ON resistance to be put in series with a good lipo cell! I did not research FET on resistance, so I again may be outdated, but I think you will find the minimum ON resistance possible today is still in the hi single mohm digits, typically 10+mohm.... this make them totally impossible to put in a batteries series path.
I believe this is why "lockout" is actually only a SINGLE or TWO FETS in series with the WHOLE pack: one for opening if voltage of a SINGLE cell goes too low, or the other open if a single cell goes too high. No individual lockout possible.
Said another way, if you put these 10mohm series resistance FETS in series with each cell, you will end up with a battery resistance so high it cannot produce the 3-20x peak current required to operate our quads....
An example is this smart battery device that shows this in detail:
http://www.ti.com/lit/ds/symlink/bq77910a.pdf
This is why the technology today is still whimpy trying to balance cells against each other : we cannot put a FET in series with each cell so we put one in PARALLEL around it with a series 5-10ohm resistor - and turn it on to SHORT out the cell during charging, thereby charging it a tad LESS than then other cells. So to balance, the HI voltage cells get slightly bypassed by this small 10-100ma current, letting them charge slower than the other cells....
So it is my thesis than the low cycle count life of our dji batteries MAY be due to a cell falling behind in charge and therefore, with this whimpy bypass charge scheme used today, never charging fully again before the others do and shut off the charger. Premature dji battery failure caused by ONE cell simply notbeing able to fully charge ever again!
THIS is my idea, and why I want 'bad' batteries sent to me to test! If I am wrong, no damage to anyone! The bad batteries sent to me to test were of no use to their previous owner anyway! But it may show that we can get those 300 cycles -as we should - by simply putting each cell on a multi cell lipo charger....
ps: sorry if my thoughts are too random - I have been in engineering meetings all day, came home, and wife wants my attention while I want to write this
I dont know exactly what circuitry is used in any lipos protection circuit but I do know its very common to have to jump start many lipos to be able to get them to work again. A very good example of one that have protection circuits in them are cree flash light battery's. Which are 18650 3.7v battery's with protection circuit. and there are some with out the protection circuit.
The ones with a protection circuit if they ever get over discharged or over charged even a tiny bit the protection circuit cuts the power flow to the battery and you cant ever get the to charge or ever work again with out having to strip the battery and either take out the protection circuit or jump start them if they are under charged or drain some out if they are over charged. and it happens quite a lot esp if some one uses a cheap charger to charge them. Being they dont have there own charge circuit in them. Tho they are so cheap when one locks out its just as easily to just toss them and get a new one and one with out the protection circuit in them. Esp being a flash light really dont need a protected battery and even if a charger charges them up a few fractions of a volt they are not going to explode and the worse thats going to happen is it may shorten the life of the battery some. But in reality They will last longer then a perfectly good almost new protected one that decides to have the protection circuit kick in and cut power to the battery. Even when they are working and all good even in use in a flash light or any thing else. They will do what a regular battery or one of the non protected 18650 battery's will do in some thing like a flash light. When you are using them and they get down to the low cut off voltage they just shut off. Tho you usually can just let them sit for a bit and the resting volts will be high enough again that you can throw them on the charger and they will still charge being the lock out will let the power flow again if the resting volts goes up enough to not have the protection circuit kicked in any more. Or you can let the flash light sit for a bit and then turn it on and get a few more seconds of light but if you do it a few times then you end up with it being so low the protection circuit will not open again and the battery is basically not useable any more. Yet the very same battery's that are not protected battery's. If you keep the flash light on it will just keep getting dimmer and dimmer till the battery is flat. But will take a charge again being there is no protection circuit in them at all. and most chargers for them also are not smart chargers so they will even start charging even if the battery's are completely drained. Tho it might take some of the battery's useful life and capacity away being lithium battery's dont like being drained fully They will still work but just not last as long per charge. But it beats having to toss a battery that the protection circuit kicks in.
I have gone threw about 18 of my original 20 18650 battery's that the protection circuit bricked the battery's on all within the first year and many of them in just a few uses in the first few months I had them and I have about a dozen non protected 18650 battery's I use in my cree flashlights and in a 3 watt hand held burning Lazar that I have had for a few years now and have never had to toss one of them yet.
The ones with a protection circuit if they ever get over discharged or over charged even a tiny bit the protection circuit cuts the power flow to the battery and you cant ever get the to charge or ever work again with out having to strip the battery and either take out the protection circuit or jump start them if they are under charged or drain some out if they are over charged. and it happens quite a lot esp if some one uses a cheap charger to charge them. Being they dont have there own charge circuit in them. Tho they are so cheap when one locks out its just as easily to just toss them and get a new one and one with out the protection circuit in them. Esp being a flash light really dont need a protected battery and even if a charger charges them up a few fractions of a volt they are not going to explode and the worse thats going to happen is it may shorten the life of the battery some. But in reality They will last longer then a perfectly good almost new protected one that decides to have the protection circuit kick in and cut power to the battery. Even when they are working and all good even in use in a flash light or any thing else. They will do what a regular battery or one of the non protected 18650 battery's will do in some thing like a flash light. When you are using them and they get down to the low cut off voltage they just shut off. Tho you usually can just let them sit for a bit and the resting volts will be high enough again that you can throw them on the charger and they will still charge being the lock out will let the power flow again if the resting volts goes up enough to not have the protection circuit kicked in any more. Or you can let the flash light sit for a bit and then turn it on and get a few more seconds of light but if you do it a few times then you end up with it being so low the protection circuit will not open again and the battery is basically not useable any more. Yet the very same battery's that are not protected battery's. If you keep the flash light on it will just keep getting dimmer and dimmer till the battery is flat. But will take a charge again being there is no protection circuit in them at all. and most chargers for them also are not smart chargers so they will even start charging even if the battery's are completely drained. Tho it might take some of the battery's useful life and capacity away being lithium battery's dont like being drained fully They will still work but just not last as long per charge. But it beats having to toss a battery that the protection circuit kicks in.
I have gone threw about 18 of my original 20 18650 battery's that the protection circuit bricked the battery's on all within the first year and many of them in just a few uses in the first few months I had them and I have about a dozen non protected 18650 battery's I use in my cree flashlights and in a 3 watt hand held burning Lazar that I have had for a few years now and have never had to toss one of them yet.
I understand the lockout feature for over or under volts. Yes, there are quite a few 'jump start' examples on hackaday.com too.
I am not questioning the fact these exist: in fact, I am agreeing that they may well be the cause of the short dji battery life!
What I questioned was your statements that INDIVIDUAL CELLS can be INDIVIDUALLY locked out - that just does not seem possible with today's technology.
Your example of the cree flashlight is a good example: IT IS A SINGLE 3.7V 'CELL'
Now it may be multiple 3.7v cells in PARALLEL, but the fact remains, it is a single lithium cell for all intnts and purposes: so yes, they can have a lockout - YOU GET ONE MAX per battery pack.
So as soon as you add another 3.7v cell in SERIES, as our batteries on dji, then you LOOSE the capability to LOCKOUT an individual cell. That is my point.
So, more details on my guess: Since we have 3 sets of cells in series, I am sure dji measures the voltage of each of thes3 3 sets. When ANY goes below a threshold compared to the other 2, those other 2 get shorted with a parallel fet to bypass a small amount of the charge: result is the low cell gets 10-100ma higher charge than the shorted 2 better ones.
But since all 3 sets are watched by the BMS, if any set goes too high or too low then it can shut off THE WHOLE PACK WITH ITS SERIES FET. Not any individual cells.
So my idea is the BMS locks out the pack due to one of these 3 cell packs - and it may simply be the whimpy charge bypass does not give enough higher charge to that one cell that went down to bring it back to full charge. subsequent flights take it down further, until it gets too low to differentially charge this way back up - may be a grand old cell still!
On the same thesis, I say since that one cell was low and they attempted to bring it up, the nearly fully charged cell may get charged too to the upper limit too many times - resulting in heat in THAT cell pack, causing the gas release and puffiness.
I propose further than this low cell that can never be charged fully again, although not bad, is the reason for batteries that run fine from 100-70%, then seem to discharge in a minute or two of flight time - that one cell going down again.
All related. It may be that the dji repair places, if they were given the reasoning and proof of this, could begin a battery correction service to everyone - if this reasoning is correct, they could buy a $80 charger and individually charge and test EACH cell in the pack. I suspect with this service they could return the majority of these 20-30 cycled then bad batteries to grateful owners at a lot lower price than new ones.
But without batteries to prove or disprove this theory, I am only making educated guesses. Guess folks prefer to just buy new expensive batteries often rather than finding a potential solution to MAYBE get them to the advertised 300 cycle life. So be it.
I am not questioning the fact these exist: in fact, I am agreeing that they may well be the cause of the short dji battery life!
What I questioned was your statements that INDIVIDUAL CELLS can be INDIVIDUALLY locked out - that just does not seem possible with today's technology.
Your example of the cree flashlight is a good example: IT IS A SINGLE 3.7V 'CELL'
Now it may be multiple 3.7v cells in PARALLEL, but the fact remains, it is a single lithium cell for all intnts and purposes: so yes, they can have a lockout - YOU GET ONE MAX per battery pack.
So as soon as you add another 3.7v cell in SERIES, as our batteries on dji, then you LOOSE the capability to LOCKOUT an individual cell. That is my point.
So, more details on my guess: Since we have 3 sets of cells in series, I am sure dji measures the voltage of each of thes3 3 sets. When ANY goes below a threshold compared to the other 2, those other 2 get shorted with a parallel fet to bypass a small amount of the charge: result is the low cell gets 10-100ma higher charge than the shorted 2 better ones.
But since all 3 sets are watched by the BMS, if any set goes too high or too low then it can shut off THE WHOLE PACK WITH ITS SERIES FET. Not any individual cells.
So my idea is the BMS locks out the pack due to one of these 3 cell packs - and it may simply be the whimpy charge bypass does not give enough higher charge to that one cell that went down to bring it back to full charge. subsequent flights take it down further, until it gets too low to differentially charge this way back up - may be a grand old cell still!
On the same thesis, I say since that one cell was low and they attempted to bring it up, the nearly fully charged cell may get charged too to the upper limit too many times - resulting in heat in THAT cell pack, causing the gas release and puffiness.
I propose further than this low cell that can never be charged fully again, although not bad, is the reason for batteries that run fine from 100-70%, then seem to discharge in a minute or two of flight time - that one cell going down again.
All related. It may be that the dji repair places, if they were given the reasoning and proof of this, could begin a battery correction service to everyone - if this reasoning is correct, they could buy a $80 charger and individually charge and test EACH cell in the pack. I suspect with this service they could return the majority of these 20-30 cycled then bad batteries to grateful owners at a lot lower price than new ones.
But without batteries to prove or disprove this theory, I am only making educated guesses. Guess folks prefer to just buy new expensive batteries often rather than finding a potential solution to MAYBE get them to the advertised 300 cycle life. So be it.
How'd you do THAT?
I think you may get it fully back if you turn it on and discharge it till it shuts off (crude and potentially damaging)
Better yet, if you have no precision discharge thing, turn it on in your phantom and let it run down. When I do that, I put a fan next to the phantom to keep innards cool, and watch the battery screen on the computer.... you probably will see the voltages going down and down and down with mah shown at 0, for about 5 hours.... when the batt voltages reach a certain low level, that will trigger the recalibrate routine to run... Shut it off when the lowest cell hits say 3.4 volts. Recharge and I bet your capacity comes back up to - with 45 cycles, depending on how you have treated it - about 4900maH....
LetUsKnow LetsGetStarted....
I've done that over 3 times in a row... Nothing... Still shows that. It keeps getting better overtime but then it would drop back down. Keep it mind that when i do fly with that battery (i plug it into my laptop every now and then to check the voltages when flying) It still gets about 15-18 min. The only problem is that it has is inaccurate battery percentages and it would always stay at 1Mah for hours when discharging Heres an update:
Attachments
Last edited:
It still flies for 15-18 minutes with no issues at all. Just inaccurate battery percentages.
So that proves it is just a calibration issue....
1) When you say you have discharged 3 or more times, did it ever go as low as 3.4v on a cell? If so, what were the other cell voltages when you shut it off?
2) did it ever go to 0mah and remain there for some length of time? If so, how long? and did the voltages ever get to 3.4v?
1) When you say you have discharged 3 or more times, did it ever go as low as 3.4v on a cell? If so, what were the other cell voltages when you shut it off?
2) did it ever go to 0mah and remain there for some length of time? If so, how long? and did the voltages ever get to 3.4v?
I'm not very sure of the voltages and stuff as i wasn't monitoring it. I only thought about draining the battery and looked at it every hour or so.
1. Yes, i think it did, the cells stayed quite close balanced throughout the battery drain. I think it shut off when the battery was below 3.4V Sorry, i cant confirm as i wasn't with my phantom when it shut off.
2. No, It went from 100mah and kept on getting lower and lower untill it reached 1Mah and it just stayed there. It never went to 0Mah. It stayed at 1 Mah most of the time i was draining the battery.
Let me know if you want to see some sort of time-lapse of me draining the battery. I'll drain it again and see how it goes if you want me to make the video.
