I will not lie. I was slightly (more than slightly) hoping that gang of angry-as-hell envelope cells would have gone thermonuclear. I sort of chanted in my head ala Andy (photonic) style "I'm not avin' it!! Flames!! I want FLAMES!!! Crank it RIGHT UP!!" Side note, I miss Andy and his old self.
No point making such a circuit. The parts you'd need to get (from China) to built one cost more than a ready made module you already can get (from China).
The design for a balance circuit for 2 to 4 cells is usually in the datasheet for the part you are thinking of using. They are used in just about everything these days.
Actually the one that reaches the high cutoff voltage first is likely the lowest capacity one, and they will all reach the low cutoff voltage at roughly the same time because even though one might only be charged to a lower voltage it would still contain the same capacity.
You said that balance charging would bypass the cell when it got to it's full voltage. That's interesting - I've never seen a circuit like that. All balancing circuits I've seen actively discharge the highest cell to bring it's charge down in line with the rest. Wouldn't it be a very complicated circuit if it were to bypass a battery and still maintain the correct voltage across the remaining cells?
That would greatly amplify any real explosion. Leaving it without a lid directs any explosion upwards in a relatively "safe" direction. A fast flare-up and a fire is more likely than a BOOM type explosion in the first place, though. Have a look at detonation velocity on wikipedia, it gives some explanation of this.
This is a fire containment dish. It holds the cell so he can carry it outside before it burns the house down. And lithium cells don't explode. They burn vigorously. Hand grenades explode. See the difference?
So batteries that have their protective board if you were to take 3 lipos like that and not run them in parallel and you ran them in series would it be okay to charge them without a balance lead or a balance board set up? because they all have their own protection board or would you still run into the same problem with them just charging at different rates and getting different voltage? I was just curious because I was looking to put a bigger battery in something, its just a 3.7 lipo but I want more mAh and I didn't know if it was okay to do that or if I needed to figure out how to make some kind of balance charger board or buy something that does that. thanks for the great videos and the awesome content! I love your stuff so much, you're so knowledgeable and I've learned so much from you! keep up the great work!
Pretty lame to not have any balancing especially the boards are so cheap, but at least each cell had protection. I'd put a 3S 4 pin JST connector on them and use a lipo charger occasionally just to knock them back into sync.
The next time you have got one of those construction lights, take a look into its battery-pack. The one in mine is constructed similar to this one: 2 cells in series, each with its own protection circuit on a common board but no balancing.
I have actually noticed a slight trend with DW-01 and dual mosfet protection modules upon cutoff in series with other cells on discharge they tend to act like a reverse bias diode at a few percent of their max current allowance so if a pack like this was, say, used on a short LED strip(the likely intended use case) it would dim down to pack-1 cell voltage(6-8.4v) and keep chugging along as long as the current remained under about 10% of the module limit still not ideal(I would slap a cheap Schottky 10-15a reverse bias in parallel with each cell ) then use this on a voltage variant tolerant system like the USB buck port actually with the diodes this pack would be fairly ideal for an "always on" usb port possibly even just using a 13v input with a diode and letting it stay plugged in forever and using the usb to power something like a small network switch that takes 5v power goes out and you keep your basic access point/switch going without a fully AC UPS device heck even the ~12v port would do that well as most "12v" network gear takes in 12v but then instantly bucks down to 5v or less and fuck my worn laptop keyboard the K key needs to be jabbed so hard to register
Even if the batteries are made to be fully matched at the factory, they will inevitably vary with use due to external factors. That's why I think that balancing is major important. Laptop batteries do use this technique but they connect each two cells in parallel without any means to disconnect any of them in case of failure. Sad but it happens.
The best I’ve found for a 6# pack I made was a BMS that would monitor the cells and disconnect them whenever any was outside it’s operating range, and a balancing module which would stick a 40 (I think?) ohms resistor across the cells that went over 4.2 Is there any module to be bought that does what clive showed, with diverting all the current and balancing the cells in a non-wasteful way ?
Since we grow ever closer to the permanent move from our sticks and bricks home in South Dakota, where the coyote's howel, the snow piles up so high that cars are found by the radio aerials on their roofs, and the winters so harsh to freeze animals where they stand on the prairies, into our 36 foot Class A Motor Home, which we have used for the past 3 winters to escape to Sunny Arizona, where it is always warm and sunny, well except when it isn't. I have been planning on more 12 volt supplies for the toys that I build using Arduino's and ESP8266's I have all the parts necessary, as well as many recovered Lipo cells that I have torn from old laptop batteries, old tablets, and so forth. I have some circuit boards that supposedly regulate the cells so as to charge them properly. I currently have two built, one with a step up power supply, and another with a step down. Both work wonderfully, and I have equipped them with those cheap power displays so I know pretty much what they are putting out at the terminals, but they are rather large and bulky, I have them housed in white plastic boxes with clear lids so I can see the display through the top, and drilled holes to access the rheostats to regulate the power. But I really want one like you have covered here. I believe I have plenty of parts to build one,however I never did build my 3d printer, so I will have to mount it all in wood, or worse yet steel cut from old power supplies out of ancient computers. So at long last, I must end my run on sentence with an expression of thanks for covering this device, you see in order to be able to make our transition from house to RV, I must MUST absolutely MUST stop buying crap from China, don't believe me, just ask my wife of 48 years and she will let you know in no uncertain terms that I am NOT allowed to even THINK of EBAY this summer, lest I find myself grounded in our old house here in South Dakota and suffer through another year of cold, wind and excessive snowfall, and man I just cannot take that shit an more. After 67 years of snow in the winters, only interrupted by a few years living in Texas, a year in Nam I find the very thought of shoveling one more scoop of snow revolts me, and makes my heart beat like it may be the last of such abuse. So thanks a million for showing me the make up of this one, I fully intend on building one, what I have in mind is to use wood to build a box that will appear to be a miniature 12 volt battery with the caps and all, perhaps even paint it black, where the posts should be will be the banana plugs, and perhaps I shall have the caps light up when power is turned on. I will use a power brick that used to power my old router before it gave up the ghosts this spring for charging it, and will have a flip door where the display hides so I can check the condition of my home built 12 volt cell. I am even considering making it with more flat cells so I can have enough amps there to power my Ham Radio.
Frankly, a pack with no balancing and with the battery manufacturer's protection boards is likely much safer than a pack with a chinese balancing circuit. With protection boards, balancing is just a matter of long-term performance, not safety.
Huh? It won't be "the battery manufacturers protection board", but just a single cell protection board that was added later down the production line, and then the battery + protection PCB (also made in china) repurposed for this product. Whilst I agree that balancing is more about performance than safety, since all this stuff is made in china anyway, there is no basis for a claim that it's safer than a proper PCB that would have cost 50c more which had individual cell balancing and whole pack protection.
With the psu since the light lit when it was plugged into the power bank and not the wall indicates to me that you would get a nasty shock if you bridged the two bottom prongs
I nearly wrote a comment complaining about the way this pack is wired, then I suddenly remembered that about 95% of all of the sub-$60 Li-ion gadgets I've dissected to date have been set up exactly the same way -- cells in series, no balancing, and basically zero f*cks given by the manufacturer. Re-purposing 18650s, I've come across my fair share of dead or reversed-polarity cells.
In some cases it is possible to have Li-Ion packs without balancing, if you use cells with LiMn2O4 cathode material. They are often used for older ebikes or powertools, for 18650s they are mostly 1,3 or 1,6Ah cells. Best examples are early Makita packs, Dyson 21,6V or Bionx ebike batteries. Those cells have a self balancing behavior if they are closely matched (which you need for any pack with cells in series) little bit like NiMH did. If you have a Makita battery with no balancing and dead cells, it‘s because one cell died. Balancing wouldn‘t have done a thing in this case.
I will not lie. I was slightly (more than slightly) hoping that gang of angry-as-hell envelope cells would have gone thermonuclear. I sort of chanted in my head ala Andy (photonic) style "I'm not avin' it!! Flames!! I want FLAMES!!! Crank it RIGHT UP!!"
Yeah that's not ideal. Unlike NiMH, Li will not find an equilibrium between cells so they will "run away" from each other so the voltage differential will get larger and larger. You have protection so there is little danger but the effective capacity will plummet. Shove a 4V2 zener between cells, will be reet haha
Actually, assuming the charger is fairly dumb, putting a zener as high as 5v would fix the problem. The higher voltage would drop the charge current down low, allowing a small amount of charge to bypass the fully charged cell to top up the other ones. When all are charged, almost no current should flow across the 15v of all three.
True but NiMH is not impervious. The balancing needs to be done at a fairly low current, otherwise the charged cell with get very hot (and potentially vent/explode). But, balancing them this way also uses up the chemicals inside the cell faster, which causes the cell to die sooner.
Interesting. My brother bought one of those, hoping they did something better than this, as he paid a fair price. But no. Also, one of the cut-off circuit chips was broken and (also) gave a terrible ripple, which worked horribly bad on what he wanted to use it for. Nice fix for it: on ebay or aliexpress you van buy charge balance circuits for around $2.
In the early days of RC lipos a very good make didn't have balance leads. So after a few years of use i fitted balance leads and all 3 cells were the same voltage
I found this exact same arrangement inside the first generation Mini-rig portable speaker. Interestingly the mini-rig charges from 5v and uses a boost converter to charge the 3 cell LiPo pack.
Most of the "balancing" lithium-ion battery pack chargers don't do that "bypassing" trick you mentioned. I wish they did... They tend to charge the serially connected cells directly from both ends and check the individual cell voltages while doing so. When one of the cells reach above the set voltage, it is then slowly discharged by applying a resistance on to it thru a mosfet, keeping the cell voltage at the set level while waiting for the other cells to keep up. I never understood this approach, it is a waste of energy and it reduces the battery life significantly since it constantly cycles the cells. Not to mention it takes more time to charge the pack this way since some of the energy that put in is converted to heat by the discharging mosfets.
In my mind a resistance parallelled to a fully charged cell in a set of hard wired cells doesn't necessarily discharge it as long as there's a charging current present. I'm more inclined to see it as a parallell path that shunts part of the charging current across the fully charged cell in order to continue charging the others at full speed, while keeping the former topped up. The parallel resistance would only discharge the cell if the voltage across it from the charging current would drop below the cell voltage it's parallelled to.
What if you stick a beefy 4.7V Zener across each protected cell (on the protection PCB)? Wouldn't that correct any imbalance exceeding 0.5V total by bypassing the cut-off fully charged cell until the other two cells exceed 3.95V average? Extreme imbalance might charge two cells to 4.25V and the last to 3.2V. A lower Zener voltage could improve balance at the cost of higher leakage.
jehugarcia said in a video, that after running he's electric converted Samba, with 18650s from old laptop batteries, and only bulk charging them (cut of at the max voltage for the series, ie: 12.6v), as long as all the cells have been tested and shown similar capacity between them and no self discharge after a charge while sitting for a few days (no internal shorts), they've shown very little imbalance between the cells.
Ultimately balancing is about performance, not safety (assuming your individual cell protection exists and works). I think realistically that especially if you only have a few cells, balancing is probably well into diminishing returns territory, assuming of course that the cells were well balanced and matched for capacity to start with. If one cell gets out of balance so far that it's going to be a significant benefit to balance then that cell is probably knocking on death's door anyway, all the balancing in the world isn't going to help a pack if one of it's cells is dying. Balancing adds a fair amount of complexity, even most of the cheap balancing boards, where they just connect a small load across high cells to bring them back down, that's a fair old accumulation of components. The only thing I would really like to see is for the DW01 to have a just slightly lower high voltage cut off 4.25 or even 4.35 (4.3+-/0.05 tolerance for the DW01+G and you never know what you'll get from China) is just a bit high for my liking, actually I'd like a DW01 (or workalike) with a 4.1v cut off.
i actually have a old style SLA battery pack has 2 12 volt ports that see in cars. a small flashlight thing on the back and has a 9/6/3 volt ports like the one in the vid. orignal cell was 4Ah but its replacement is 4.5 not that huge but its handy for certain things id be scared with at a lithium cell level. even though bit dated i love this thing for what it does aha.
It'll only be 7Ah for about 100 or so cycles, in case you forgot Lead Acid batteries shouldn't be taken below half their capacity if you want to get a reasonable amount of cycles out of them, so yeah I think I'd still rather have a fake 20,000mah power bank over an SLA.
many years ago i would agree but with how lithium cells today have better chemistry than SLA to go beyond 40% discharge without damage unlike SLA... if you got to power 12v most people will aim for 3 cells, yet at the lowest end of the cell 2.8v you get only 8.4v not 12 and generally too low for the equiptment, sooo then people go for 4 cells but their upper voltage is to high at 16.8 (4x4.25 max) so this is where you need to think, can you use a low current buck-boost converter (.>3amp) or do you need higher current and suffer with too low voltage cut off.. have a look at the size of 4s3p 2600mah 18650 cells, yes i agree its about the same size as the 7ah SLA but the 18650's can produce higher current right down to near empty! the SLA might only give you a full empty of 7ah if current is only 250ma...... then the SLA might be damaged from full discharge
I recently came across a technology I haven't seen before. It's a "powerline" internet adapter. It's basically a box that you plug into a power outlet and the box has an Ethernet cable going in. It actually creates an internet connection over the power network in your house so by just by plugging another box into an outlet you get a hard line internet connection. Would be interesting to see inside one.
Those things are fantastic and neat, I got to use them in the early 2000s and some of my family still uses them at home. It's probably quick enough for realistic 1-computer needs. You ever get a chance to use power line networking over the last few years?
Would it really have been that hard to put in a simple 3S balance charging BMS? I got 2 of those for 2 dollars and use them for low voltage cutoffs (they do need to be reset by connecting a charge pulse to the output every time they activate, but that can be also reset by connecting the output negative to input negative, after recharging and reconnecting the battery, of course. That resets the charge chip). They could even have kept the individual one cell BMS' as they wouldn't have interfered with the 3S balance charging BMS.
I made a similar thing with a slim aluminium extruded case and 3 18650s inside to allow me to use a little 12V power tool I have without having to use the mains adapter, but for the sake of an extra quid fifty I stuck a 3 cell BMS and balance board from ebay in there too. and re the voltages of the DW01, it depends which version. theres a DW01a, DW01b, c etc etc, all the way up to g I think, each with slightly different detection and release voltages for over and undercharge
That will suffer a quick loss in capacity every charge cycle. The least charged cell will die quicker and quicker each cycle since the main current support exists in the 4-4.2v range and anything below 3.7 suffers much greater voltage drop at the same amperage. Because the highest charged cycle will perform at 100% it will require less charge which will create a domino effect on the least charged cell getting ever less recharged.
In work we have some batterys that are 24v 10Ah. They contain 28 18650 cells ( 7s4p ) and they have NO Balancing or protection. The carger for the packs is a 10A charger.. Things can go really bad :)
Nice video, didn't even know these existed! They're going to be a bit dangerous though, with the lack of balancing, as soon as the cells drift (which they will do, all lithium cells in series eventually become unbalanced/one cell loses capacity first) then the entire pack will be limited by the cell with the lowest capacity. It'll also damage the other two cells (or at least the highest) by making it hit its protection cut off voltage all the time, which you don't want to do to keep a good lifespan of the battery. Just keep it to a balanced setup @ 4.2V per cell and it will work fine. P.S. If you ever decide you don't want those cells then let me know; I always have a use for lithium cells in projects :D
Can I please make a teardown request Clive? There are replacement rechargeable 9v batteries available that use 2 x LiPo cells and are USB chargeable. They sound promising but only if the cells are balanced and the output is modified to give near 9v over most of the discharge range... Thanks for the videos.
I have some but haven't really put them through their paces. The constant 9V output desirability depends on application. A lot of stuff regulates that down to 5V anyway.
I just ordered a couple from AliExpress because I hate single-use cells and batteries - especially the 9v which are quite expensive. That said the rechargeables aren't cheap either so I hope they work out OK.
Me too. I haven't really checked the capacity but the main use is occasional use so recharging frequently is not an issue and if I accidentally leave the devices on, I haven't wasted an expensive battery.
It has one of the micro small 1 amp max charge controller. I have some of those, but I purchased some that can handle 3 amps rather than 1. I blew one of the 1 amp boards.
Hi Clive, I want to create a 12 to 18v battery pack using lithium cells, say 3x4.2v (12.6v Max). X4 packs in parallel (3000mah per pack) making approx. 12000mah pack. Can you tell me what the best way of charging this packwould be and best way of protecting it with say a bms board? Much appreciated. Look forward to your reply. Love your videos. Excellent work
Ok, here's the thing about balancing. People get a little too hung up on voltage. Not all cells have the same capacity, so on one cell, you may put 1900mAh into it before it hits 4.2, and on another, you may put 1980mAh. This discrepancy is normal, even perfectly matched cells will age differently over time. So you have to balance them to get the most capacity out of the pack. Now, you can top balance - that is, make sure they all hit 4.2V simultaneously. But, because the capacities differ, when you pull 1900mAh back out, you will deplete the "smaller" cells before the "larger" ones. This means the smaller cell will hit its 2.5V (or whatever threshold you set) while the others are still above that. So when do you cut off, exactly? When the cell hits 2.5V under load? Or when it hits 2.5V with no load? Batteries sag a little under load, so 2.5V while supplying 2A is not necessarily empty. But if you keep discharging, things start turning sour. The voltage curve at the end of charge is very steep, so it doesn't take much to drop the cell voltage down to 2V, or 1V, or 0V. When that happens, you set yourself up for reverse polarized cells, and then comes our good friend, fire. If you bottom balance, then all cells hit the steep part of the discharge curve at about the same time, so the ability for the battery to deliver current falls swiftly and it becomes quite clear that it is truly empty. Even still, draining to 0V when all cells are balanced is not necessarily terminal, since the possibility of reverse polarity is minimal. (There's no current left to reverse charge the first cell to hit 0.00V.) But.... you DO have to watch the top end - when charging. One cell will hit 4.2V first, and at that point, you need to stop. It doesn't matter if the other cells are only 3.7V - you've filled the capacity of the lowest cell, and that's all you're going to get. When you discharge, you'll take out what you put in, and all cells will again hit bottom at the same time. The top voltage is all but irrelevant. And since you typically charge at a constant current, it's a lot easier to control the HV threshold trip point than trying to guess whether you've really hit the LV trip point under load. Monitoring circuitry is still nice to have, but it's not really critical to see every cell's voltage. You're always charging from 0 to some capacity, and then taking that same capacity out. Since LiIon cells don't self-discharge unless they're damaged, they won't drift apart appreciably over time. A per-cell HV cutoff is really all you need. In fact, monitoring individual cells puts the tiniest load on them, and creates an opportunity for imbalance since the cell monitoring current (albeit small) won't ever be precisely the same. Then you DO need re-balancing circuitry, and you've just made things more complicated. So this simple design is not terribly safe in the event of corner cases -- such as internal shorts or some other abnormal imbalance - but, in the normal course of operation, it does exactly what it needs to do to keep the cells happy and no more. (Although 4.25V is a little too high for my taste.)
If you've ever seen professional balancing circuits, they have transistors which shunt resistors across the cells to make it balance more effectively and not overcharge any cells.
Yeah, that's "top balancing" -- the art of chasing unicorns... until you eventually become aware of the fact that they are actually dragons. The goal of top-balancing is to have all the cells charged up to their maximum voltage, plus or minus margin-of-error. Shunts across a cell are a controlled discharge, with the aim of bringing that cell down below its maximum so that the entire series string can be charged up some more. The highest cell is continually knocked down until all of them arrive at the same high watermark voltage, all at the same time. Sounds like a laudable endeavor, but there are some fundamental problems with that approach. 1) The voltage will drop after charging stops. The chemistry of the cell settles for a little while after end-of-charge. If you measure right away, then measure in 30 minutes, then measure tomorrow ... the voltage will have dropped. This will cease after some time, but it can be days before the difference gets lost in the noise of the measuring device. Balancing circuits will dutifully pull those voltages back up again, by discharging and recharging, in an obsessive attempt to keep them at 4.200V. The more they mess with it, the longer it takes, and it does nothing at all beneficial. The additional energy put into the cell wouldn't light the charging indicator LED for 30 seconds. 2) Li-Ion cells do not particularly appreciate being at their max voltage. This is stressful for the cell. You're far better off skipping the constant-voltage portion of a charge cycle, and just cutting off right when it first hits 4.2V. Or even a little less. How about 4.1V? If you want >1000 charge cycles, maybe go with 4.0V. 3) You are not increasing the capacity of a series pack by "topping off" the cells. The battery will hold a set amount of energy, no more no less. It's not wasted capacity, it's just a surface voltage that will dissipate almost immediately under load. 4) Continuing on the theme of #3, when you balance at the top, you necessarily create an IMbalance at the bottom, which is a LOT more difficult to monitor effectively during a variable-load discharge cycle. You then MUST have a monitoring circuit that prevents over-discharge of a single cell - a problem that is not trivial to solve, due to the reasons explained in my first post. 5) All the added complexity to pull off this feat of engineering becomes a liability. The batteries themselves are predictable and reliable. Think about it... How many cells spontaneously burst into flame, just sitting there attached to nothing -- compared to self-made disasters where the electronics went off into the weeds? As always, KISS.
At least it has protection. Though, if one adds a diode and resistor over each protected battery, then when the protection kicks in, current can still flow through the diode and resistor. So that one at least could charge it more fully and even it out at the top at least. (The diode is so that we don't discharge the batteries when not charging, and the resist is there to make sure we aren't dissipating the full 4.2 volts in a diode. And yes, this "balancing circuit" doesn't really do anything during discharge at all.)
Have you ever torn into a DJI drone smart battery? They cost $50 and do their own balancing and communicating with the drone and transmitter/controller.
Had a few of these i use for bike lights (security camera battery packs). I was pondering adding some leads for balanced charging using a lipo blance charger i have. Dont want to burn the house down though. Whats the explosion risk? 😞 disapointing to know i wont get full use of the capacity if one of the cells is dipped. Any workarounds for that? (Im awful with circuits but here to learn 😅)
3 cells is a battery, it doesn't have any other battery to balance with...each cell has a maximum voltage after that cells protection circuit opens and stops charging only that cell the others continue charging ultimately all equal in the end...
Ive seen them use the High Drain feature to get to their rated output, Just like those jump starters do, listing a 800 Amp output but 3 cells each about 4000mAh at 3.7V, searching specs will show a high drain feature of 25A and Im guessing they twist the numbers around and use the High drain and the peak surge to come up with a WHOPPING 800 Amp when actual is at the 4A at 11.0-12.6 V . I have SO MANY modules and batteries here Id be more than happy to share with you some like my huge flat Li-Ion cells that have real specs of 3.7VDC@ 8Ah and a high drain of 200 Amps (Get this and its true, I did a high drain of 100 A and stopped there LOL)
Clive, you're regularly showing us that cheap and nasty stuff off e-bay is just that, as well as frequently dangerous too. Well the list of things to avoid (and even throw-out in cases where one was the once proud owner) is getting longer with each video. The list of "great" stuff is very shor, although you've occasionally come accross some "OK" designs, could you please keep an eye open for a (globally distributed and internationally available) item, such as a high capacity (10,000mAh+) USB power bank which is EXCELLENTLY designed and implemented, and share which one it is? Something which has over-discharge protection, over-charge protection, maybe something which even looks after its cells keeping each one of them in that long-total-lifetime sweetspot of 30%-80% capacity...
I've got a "S8 30000mAh made in china" pink bank with a solar panel. Exposing it to sun made one cell fucking explode. The circuit inside is so inefficient, that it discharges cells with the solar panel, rather than charging them, and the circuit is so bad, that even though there's 3x2700mAh3.7v cells, they charge just 5% of a 2700mAh phone in 1 hour, then go dead flat 3.0v. EDIT: One of these i.ebayimg.com/images/g/UhcAAOxyi-ZTYbJ0/s-l225.jpg
this is interesting, I know its not normal for a balancer to fail in a way that it can inversely charged any cells promoting damage to the cells. but I wonder if the cell chemistry is getting to the point of the lower amperage cells are safe enough to do this? or will the difference in voltage force a current out put between the different charge batteries would be just as destructive as it as been supposable has been in the past. I would imagine that so long as the cells are as close as they are of don't stray from there safe charge/ or discharge levels, It wouldn't be to bad. I have been using 12 LGdbhg218650 and B18650ca wired up for 12.8vdc max charge using only inline single cell protection. but I also have not ran either cell type up to its max discharge rate. they have been discharged using a 5v 50w regulator running a Raspberry Pi 3 B, a 7 inch touch screen and a keyboard as well as a mouse, a few times with a usb to Sata connector for the Raspberry Pi. In this configuration, no cell dropped below 3.7 independently and only a 0.01v variance between cells.
Yes, over/undervoltage protection, but since they're in series,as soon as one cuts out, the whole string is interrupted... there is no separate power feed to each cell.
The cells will 'drift' over time due to different internal resistances of the cells (ESPECIALLY since they are not a brand name cells manufactured with higher consistency,) and resistance within the soldering connections and slight variations within the protection circuits themselves. When we use 18650 cells in series for high powered flashlights for eg, IDEALLY, we use cells with the same internal resistance (not much of a problem with Sanyo, Panasonic, etc) and charge them to the same voltage before use, AND we do not discharge the cells till dead to reduce the chance of damage to that one cell that may be weaker than the others. The cautious user will use for example three protected cells in series knowing that the protection circuit will kick in and prevent overdischarging of a single cell, but, the observant/ aware user will simply use three unprotected cells and monitor them carefully during use.
Hi Clive, probably not the best place to ask this but what would be your recommendations for powering a homemade bluetooth speaker. Input to the bluetooth amplifier card is between 7 and 20v and ideally want to charge it as I use it connected to a USB charger port. Any help gratefully received.
Looks like an amateur-made battery pack. Someone has acquired a bunch of li-ion cells, stacked them up, up to 12V, and jammed them into a generic electronics enclosure together with some simple circuitry.
in some mobilephones, the Li-Po battery is rated to be charged up to 4,35 or 4,4V and all apps mesure this voltages in the battery so, maybe some batteries can be charged with up to 4.4? instead of 4.2??
Not only is the whole pack capacity limited to the lowest capacity cell, it is also limited by the difference in state of charge of the cells. In the worst case, if one cell is fully charged, and another cell is fully discharged, the total pack amp hour capacity will be zero. When you try to charge the pack, the fully charged cell voltage will rise above the cuttoff 4.2x voltage and break the circuit, and when you try to discharge the battery the fully discharged cell voltage will drop and break the circuit. You can "fix" this problem by manually charging each cell to full capacity with a bench power supply, and hope that the state of charge doesn't "drift" out of balance too quickly. Sometimes you will get a bad lithium cell which self discharges sitting on the shelf- something to do with spikey crystal structures that grow inside the cell, which will cause the pack to go out of balance more quickly. That sound right to you guys?
almost all power tools ice taken apart dont have balancing in their 18650 packs, even the namebrand ones, not even balancing cable for the charger. i heard that manganese cells dont require one, not sure about that
I suspect that the 10S battery on my e-bike has a similar dodgy protection topology, but the battery case is so tightly packed that I can't safely remove the individual cells and restuff it. Any ideas on reconditioning the cells in situ?
I tried to make a 12 volt battery for my hand held CB radio which came to me used with no battery pack. I soldered three of those same type batteries together and charged them up. Well somehow the CB was left on for a few weeks, and when I picked it up, the battery was dead. I opened the battery box and the center battery had swelled to the max, the other two batteries seemed OK. I tossed the swelled battery and still have the other 2 however I decided to use an external battery pack when needed for that old CB. (large GE hand held looks like a WWII Walki-Talki but in near new condition) 3 channel 3 watt radio with huge pull out antenna!)
hi bigclive. i am wanting to build a battery bank. but im not 100% sure as to the charging of it. i intend to use a bms board to control the charging and discharging, but need to know if the charger for this system needs to be a dedicated lithium battery charger (due to the needs of the cc/cv stages). or can you simpmy use a generic battery charger and let the bms take control of the charging stages? can you shed any light on this for me? much appreciated. keep up the good work! love your videos
This kind of nonsense makes me want to build power banks with all of the cells in parallel. Need a higher voltage? That's what boost converters are for...
if not balanced and not precisely matched, I'd expect the difference in charge between the top charged and lowest charge ones to gradually increase over time until the pack stops being usable for no reason other than you can't charge the top cell or discharge the bottom cell anymore and you'd think it dead.
At 8:22 a little crater you say, so that is where the magic smoke did a great escape then? 9:42 is that the good smell or the bad smell? 10:50 J3 Smudge - get the flux out of there
Doing the volts comparison at 5:00. Isn't it ironic that the smallest looking cell (top one) has the lowest voltage, highest voltage is the middle cell which looks to be the largest looking cell & the midway voltage is the bottom cell which looks to be in-between in size comparison?
Yup, using the battery safety circuits for cut offs are the new cheaper batteries. I had a Bluetooth speaker that did the same. It’s not good, the protection is also within %5 to 10% accuracy as well. Protection circuits were for preventing one cell running away into another cell and stopping thermal runaway. It was never intended for charging, one less thing to make it cheaper and charge you the same cost to buy. I don’t like this method at all.
so its really only a 7Ah powerbank and the 12v figure is way optomistic, though it at least claims that the 12v is not regulated. too bad the 5v chip is toast so we cannot see the efficiency.
Having worked in the hobby industry since lipo cells had no balancing input! I can concur what actually happens is you get 3x cells that should be 3.7v nominal so 11.1v 3x 4.2v no load per cell means 12.6v but as the cells drop off you still get 12.6v no load but spread between different cells.... I.e 12.6v fully charged but with silly cell killing variances like 4.2v cell one 4.0v cell 2 and 4.6v cell 3 this caused the rugby ball "puffing" of Lipo cells and even explosions and fire of generation 1 cells, gen 2+ saw protection circuits and balancing introduced as well as fibreglass tape to stop the over expansion of cells and inevitable fire etc afterwards.
Well, in the bad old days of NiCd and NiMH battery packs for cordless drills, they didn't use balancing either. The weakest cell would dictate the lifespan of the pack. It would ruin it once it had degraded far enough to allow reverse polarizing, and I believe that especially NiCd wouldn't recover from that (not sure about NiMH). IMHO the most important reason to use balance charching on (plain) Li Ion cells is the safety. Mild overcharging NiCd or NiMH wasn't particularly dangerous, but something Li Ion can have a much more severe reaction to, as we know. I guess this particular power bank was slapped together using protected cells and a "meh, that'll do" attitude. A self respecting manufacturer would have used balance charging to also optimize the overall lifespan of the pack.
Jonah Beale You want to see a tear down of an airbag... I read there were some that had to be replaced as part of a recall... should be able to get one cheap.
Balancing is often seen as a good thing but in fact that is a bit useless. If at manufacturing they are discharged and then charged until one is full then even if one is only 80% full then it's just that that one has 20% more capacity and all the cells actually hold the same absolute charge. When you discharge them again then they all go back flat at the same time. The balancing circuit is really helpful if you want the maximum possible energy out of the system because it makes sure that the charge is delivered back with the highest possible voltage but the Ah capacity remains the same, limited by the weakest cell. Another slight advantage of balancing is that aging is probably more even because it's charging to full capacity that wears the cell. But this can be seen as a drawback because it doesn't reduce the wear of the weakest cell, it just increases the wear of others...
You're assuming that if not "top balanced" that the cells will somehow stay "bottom balanced" and always reach their minimum voltages together. This is definitely not the case. For any hope of long term consistency you need some kind of balancing. This becomes especially true as the cells age, or the self discharge rate goes up a bit in one of the cells.
Cells have a coulombic efficiency very close to 1 ie charges in and out are equals. But the one with the lowest efficiency (0.999) are the one with biggest charge % so yes, if there's a loss, it's on the weakest one. So it's the one that reaches 100% charge that will reach 0% charge first. So actually we still have one of the cells that does the full 100% - 0% trip. It's just that with more wear of the weakest one, the others will not do a 0% - x% trip but a y% - x+y% trip that delivers a little bit more power because of higher voltage. And of course I assume that the cells still have individual protection of over charge and discharge.
Hi Clive, You should be able to create the imbalance fairly easily by discharging one of the cells partly, no? From then on the imbalance should persist for a fair number of charge cycles.
FWIW (I imagine big Clive knows this) J3Y seems to be a pretty common marking on cheapo surface mount NPN transistors, so that would be a very plausible auto-complete for the flux blob.
They were probably saying it's going to be a 4800mAH equivalent if used as a 5v power supply. Buck converters step down voltage and output lower voltage at higher current with high efficiency, so you still put most of the energy from the battery to the load regardless. It's just that the battery is discharged at much lower current than if it was a regular 3.7v pack of 3 cells in parallel with a boost circuit.
These cells are probably so crappy that some of them will never reach the cutoff voltage. So if you use a balance circuit it would never stop charging. You can't sell that, so off with the balance circuit and fingers crossed you hope you get at least 1 good cell out of 3.
![NOSFERATU - Official Trailer [HD] - Only In Theaters December 25](http://i.ytimg.com/vi/nulvWqYUM8k/mqdefault.jpg)
Nice vid Mr.Clive. Fun fact: You can tell if a number is divisible by 3 if the sum of the individual digits is. eg 543. 5+4+3=12. It is divisible.
Witch!
5+4+3=12; 1+2=3
Yay!! ... A relaxing Sunday big clive video! Happy days!
shlibber buy a Frey Bentos, open it with a tin opener, throw away the the contents, wash the dish, hey presto, explosion resistance!
give the contents to the dog at least?
Trey Quattro
That would be considered as animal cruelty
Haha. The thought did cross my mind...
I will not lie. I was slightly (more than slightly) hoping that gang of angry-as-hell envelope cells would have gone thermonuclear. I sort of chanted in my head ala Andy (photonic) style "I'm not avin' it!! Flames!! I want FLAMES!!! Crank it RIGHT UP!!"
Side note, I miss Andy and his old self.
Where is Andy? Was the best no thinking, fun channel on RUclips
we popped it!
I would like a slice of explosion containment pie please...
Dave Rogers, why have it contained in the pie dish? We want to see more burns on the bench lol.
He is back!!!! New video uploaded :)
Ah yes, the classic J3-Smudge transistor
BangDroid I wonder if that is a variant of the total smudge transistor?
how about a follow up video, where you show us how to make a charge curcuit with proper balancing to fix these cheap power banks.
No point making such a circuit. The parts you'd need to get (from China) to built one cost more than a ready made module you already can get (from China).
Yep. They can generally be had for $1-2. Like pretty much everything else, it seems.
The design for a balance circuit for 2 to 4 cells is usually in the datasheet for the part you are thinking of using. They are used in just about everything these days.
Actually the one that reaches the high cutoff voltage first is likely the lowest capacity one, and they will all reach the low cutoff voltage at roughly the same time because even though one might only be charged to a lower voltage it would still contain the same capacity.
You said that balance charging would bypass the cell when it got to it's full voltage. That's interesting - I've never seen a circuit like that. All balancing circuits I've seen actively discharge the highest cell to bring it's charge down in line with the rest.
Wouldn't it be a very complicated circuit if it were to bypass a battery and still maintain the correct voltage across the remaining cells?
I feel like it may be prudent to get some type of lid for your explosion containing pie dish :)
That would greatly amplify any real explosion. Leaving it without a lid directs any explosion upwards in a relatively "safe" direction. A fast flare-up and a fire is more likely than a BOOM type explosion in the first place, though. Have a look at detonation velocity on wikipedia, it gives some explanation of this.
... so you *want* a bigger explosion then? :-P
This is a fire containment dish. It holds the cell so he can carry it outside before it burns the house down. And lithium cells don't explode. They burn vigorously. Hand grenades explode. See the difference?
So batteries that have their protective board if you were to take 3 lipos like that and not run them in parallel and you ran them in series would it be okay to charge them without a balance lead or a balance board set up? because they all have their own protection board or would you still run into the same problem with them just charging at different rates and getting different voltage? I was just curious because I was looking to put a bigger battery in something, its just a 3.7 lipo but I want more mAh and I didn't know if it was okay to do that or if I needed to figure out how to make some kind of balance charger board or buy something that does that. thanks for the great videos and the awesome content! I love your stuff so much, you're so knowledgeable and I've learned so much from you! keep up the great work!
You'd get the same behaviour.
Thx for the reply! I was worried that they were going to get out of sink or something and controller would go up in smoke 😂
Pretty lame to not have any balancing especially the boards are so cheap, but at least each cell had protection.
I'd put a 3S 4 pin JST connector on them and use a lipo charger occasionally just to knock them back into sync.
That's exactly what I was about to suggest.
The next time you have got one of those construction lights, take a look into its battery-pack. The one in mine is constructed similar to this one: 2 cells in series, each with its own protection circuit on a common board but no balancing.
I just recently bought a very similar one off of eBay from a company that makes them called talent cell
3:07 (Chuckles) I’m in danger!
damn, i actually have this power pack, good to know.
I have actually noticed a slight trend with DW-01 and dual mosfet protection modules
upon cutoff in series with other cells on discharge they tend to act like a reverse bias diode at a few percent of their max current allowance
so if a pack like this was, say, used on a short LED strip(the likely intended use case) it would dim down to pack-1 cell voltage(6-8.4v) and keep chugging along as long as the current remained under about 10% of the module limit
still not ideal(I would slap a cheap Schottky 10-15a reverse bias in parallel with each cell ) then use this on a voltage variant tolerant system like the USB buck port
actually with the diodes this pack would be fairly ideal for an "always on" usb port possibly even just using a 13v input with a diode and letting it stay plugged in forever and using the usb to power something like a small network switch that takes 5v
power goes out and you keep your basic access point/switch going without a fully AC UPS device
heck even the ~12v port would do that well as most "12v" network gear takes in 12v but then instantly bucks down to 5v or less
and fuck my worn laptop keyboard the K key needs to be jabbed so hard to register
Even if the batteries are made to be fully matched at the factory, they will inevitably vary with use due to external factors. That's why I think that balancing is major important.
Laptop batteries do use this technique but they connect each two cells in parallel without any means to disconnect any of them in case of failure. Sad but it happens.
The best I’ve found for a 6# pack I made was a BMS that would monitor the cells and disconnect them whenever any was outside it’s operating range, and a balancing module which would stick a 40 (I think?) ohms resistor across the cells that went over 4.2
Is there any module to be bought that does what clive showed, with diverting all the current and balancing the cells in a non-wasteful way ?
Since we grow ever closer to the permanent move from our sticks and bricks home in South Dakota, where the coyote's howel, the snow piles up so high that cars are found by the radio aerials on their roofs, and the winters so harsh to freeze animals where they stand on the prairies, into our 36 foot Class A Motor Home, which we have used for the past 3 winters to escape to Sunny Arizona, where it is always warm and sunny, well except when it isn't. I have been planning on more 12 volt supplies for the toys that I build using Arduino's and ESP8266's I have all the parts necessary, as well as many recovered Lipo cells that I have torn from old laptop batteries, old tablets, and so forth. I have some circuit boards that supposedly regulate the cells so as to charge them properly. I currently have two built, one with a step up power supply, and another with a step down. Both work wonderfully, and I have equipped them with those cheap power displays so I know pretty much what they are putting out at the terminals, but they are rather large and bulky, I have them housed in white plastic boxes with clear lids so I can see the display through the top, and drilled holes to access the rheostats to regulate the power. But I really want one like you have covered here. I believe I have plenty of parts to build one,however I never did build my 3d printer, so I will have to mount it all in wood, or worse yet steel cut from old power supplies out of ancient computers. So at long last, I must end my run on sentence with an expression of thanks for covering this device, you see in order to be able to make our transition from house to RV, I must MUST absolutely MUST stop buying crap from China, don't believe me, just ask my wife of 48 years and she will let you know in no uncertain terms that I am NOT allowed to even THINK of EBAY this summer, lest I find myself grounded in our old house here in South Dakota and suffer through another year of cold, wind and excessive snowfall, and man I just cannot take that shit an more. After 67 years of snow in the winters, only interrupted by a few years living in Texas, a year in Nam I find the very thought of shoveling one more scoop of snow revolts me, and makes my heart beat like it may be the last of such abuse. So thanks a million for showing me the make up of this one, I fully intend on building one, what I have in mind is to use wood to build a box that will appear to be a miniature 12 volt battery with the caps and all, perhaps even paint it black, where the posts should be will be the banana plugs, and perhaps I shall have the caps light up when power is turned on. I will use a power brick that used to power my old router before it gave up the ghosts this spring for charging it, and will have a flip door where the display hides so I can check the condition of my home built 12 volt cell. I am even considering making it with more flat cells so I can have enough amps there to power my Ham Radio.
Frankly, a pack with no balancing and with the battery manufacturer's protection boards is likely much safer than a pack with a chinese balancing circuit.
With protection boards, balancing is just a matter of long-term performance, not safety.
Huh? It won't be "the battery manufacturers protection board", but just a single cell protection board that was added later down the production line, and then the battery + protection PCB (also made in china) repurposed for this product. Whilst I agree that balancing is more about performance than safety, since all this stuff is made in china anyway, there is no basis for a claim that it's safer than a proper PCB that would have cost 50c more which had individual cell balancing and whole pack protection.
13:00 I often cheat on the figures to make it look bigger.
With the psu since the light lit when it was plugged into the power bank and not the wall indicates to me that you would get a nasty shock if you bridged the two bottom prongs
I nearly wrote a comment complaining about the way this pack is wired, then I suddenly remembered that about 95% of all of the sub-$60 Li-ion gadgets I've dissected to date have been set up exactly the same way -- cells in series, no balancing, and basically zero f*cks given by the manufacturer. Re-purposing 18650s, I've come across my fair share of dead or reversed-polarity cells.
In some cases it is possible to have Li-Ion packs without balancing, if you use cells with LiMn2O4 cathode material. They are often used for older ebikes or powertools, for 18650s they are mostly 1,3 or 1,6Ah cells. Best examples are early Makita packs, Dyson 21,6V or Bionx ebike batteries. Those cells have a self balancing behavior if they are closely matched (which you need for any pack with cells in series) little bit like NiMH did. If you have a Makita battery with no balancing and dead cells, it‘s because one cell died. Balancing
wouldn‘t have done a thing in this case.
you should show your batter capacity and discharge method
I will not lie. I was slightly (more than slightly) hoping that gang of angry-as-hell envelope cells would have gone thermonuclear. I sort of chanted in my head ala Andy (photonic) style "I'm not avin' it!! Flames!! I want FLAMES!!! Crank it RIGHT UP!!"
Yeah that's not ideal. Unlike NiMH, Li will not find an equilibrium between cells so they will "run away" from each other so the voltage differential will get larger and larger. You have protection so there is little danger but the effective capacity will plummet. Shove a 4V2 zener between cells, will be reet haha
I'd add regular diodes so they only balance when you fully discharge them. Otherwise I reckon it could get quite warm.
Actually, assuming the charger is fairly dumb, putting a zener as high as 5v would fix the problem. The higher voltage would drop the charge current down low, allowing a small amount of charge to bypass the fully charged cell to top up the other ones. When all are charged, almost no current should flow across the 15v of all three.
True but NiMH is not impervious. The balancing needs to be done at a fairly low current, otherwise the charged cell with get very hot (and potentially vent/explode). But, balancing them this way also uses up the chemicals inside the cell faster, which causes the cell to die sooner.
Interesting. My brother bought one of those, hoping they did something better than this, as he paid a fair price. But no. Also, one of the cut-off circuit chips was broken and (also) gave a terrible ripple, which worked horribly bad on what he wanted to use it for. Nice fix for it: on ebay or aliexpress you van buy charge balance circuits for around $2.
In the early days of RC lipos a very good make didn't have balance leads. So after a few years of use i fitted balance leads and all 3 cells were the same voltage
Can you give us a tour of your shop? Always like those types of videos. This one was cool too though
I found this exact same arrangement inside the first generation Mini-rig portable speaker. Interestingly the mini-rig charges from 5v and uses a boost converter to charge the 3 cell LiPo pack.
Most of the "balancing" lithium-ion battery pack chargers don't do that "bypassing" trick you mentioned. I wish they did... They tend to charge the serially connected cells directly from both ends and check the individual cell voltages while doing so. When one of the cells reach above the set voltage, it is then slowly discharged by applying a resistance on to it thru a mosfet, keeping the cell voltage at the set level while waiting for the other cells to keep up. I never understood this approach, it is a waste of energy and it reduces the battery life significantly since it constantly cycles the cells. Not to mention it takes more time to charge the pack this way since some of the energy that put in is converted to heat by the discharging mosfets.
In my mind a resistance parallelled to a fully charged cell in a set of hard wired cells doesn't necessarily discharge it as long as there's a charging current present.
I'm more inclined to see it as a parallell path that shunts part of the charging current across the fully charged cell in order to continue charging the others at full speed, while keeping the former topped up.
The parallel resistance would only discharge the cell if the voltage across it from the charging current would drop below the cell voltage it's parallelled to.
"...J - 3 - Smudge. I think that's a standard transistor..." LOL
Obviously, that must be a standard transistor :-)
What if you stick a beefy 4.7V Zener across each protected cell (on the protection PCB)? Wouldn't that correct any imbalance exceeding 0.5V total by bypassing the cut-off fully charged cell until the other two cells exceed 3.95V average? Extreme imbalance might charge two cells to 4.25V and the last to 3.2V. A lower Zener voltage could improve balance at the cost of higher leakage.
You kids get the flux off my lawn!!!
we love you clive
jehugarcia said in a video, that after running he's electric converted Samba, with 18650s from old laptop batteries, and only bulk charging them (cut of at the max voltage for the series, ie: 12.6v), as long as all the cells have been tested and shown similar capacity between them and no self discharge after a charge while sitting for a few days (no internal shorts), they've shown very little imbalance between the cells.
Ultimately balancing is about performance, not safety (assuming your individual cell protection exists and works).
I think realistically that especially if you only have a few cells, balancing is probably well into diminishing returns territory, assuming of course that the cells were well balanced and matched for capacity to start with.
If one cell gets out of balance so far that it's going to be a significant benefit to balance then that cell is probably knocking on death's door anyway, all the balancing in the world isn't going to help a pack if one of it's cells is dying.
Balancing adds a fair amount of complexity, even most of the cheap balancing boards, where they just connect a small load across high cells to bring them back down, that's a fair old accumulation of components.
The only thing I would really like to see is for the DW01 to have a just slightly lower high voltage cut off 4.25 or even 4.35 (4.3+-/0.05 tolerance for the DW01+G and you never know what you'll get from China) is just a bit high for my liking, actually I'd like a DW01 (or workalike) with a 4.1v cut off.
Probably be easier to just lug a 7Ah 12v SLA battery about, at least then you know it's 7,000mAh... :P
i actually have a old style SLA battery pack has 2 12 volt ports that see in cars. a small flashlight thing on the back and has a 9/6/3 volt ports like the one in the vid. orignal cell was 4Ah but its replacement is 4.5 not that huge but its handy for certain things id be scared with at a lithium cell level. even though bit dated i love this thing for what it does aha.
Admittedly even lead acids suffer from imvalance I really wish someone made them wish connector for each cell.
It'll only be 7Ah for about 100 or so cycles, in case you forgot Lead Acid batteries shouldn't be taken below half their capacity if you want to get a reasonable amount of cycles out of them, so yeah I think I'd still rather have a fake 20,000mah power bank over an SLA.
Yeah, or even make affordable modular batteries.
many years ago i would agree but with how lithium cells today have better chemistry than SLA to go beyond 40% discharge without damage unlike SLA...
if you got to power 12v most people will aim for 3 cells, yet at the lowest end of the cell 2.8v you get only 8.4v not 12 and generally too low for the equiptment, sooo then people go for 4 cells but their upper voltage is to high at 16.8 (4x4.25 max) so this is where you need to think, can you use a low current buck-boost converter (.>3amp) or do you need higher current and suffer with too low voltage cut off..
have a look at the size of 4s3p 2600mah 18650 cells, yes i agree its about the same size as the 7ah SLA but the 18650's can produce higher current right down to near empty! the SLA might only give you a full empty of 7ah if current is only 250ma...... then the SLA might be damaged from full discharge
I recently came across a technology I haven't seen before. It's a "powerline" internet adapter. It's basically a box that you plug into a power outlet and the box has an Ethernet cable going in. It actually creates an internet connection over the power network in your house so by just by plugging another box into an outlet you get a hard line internet connection. Would be interesting to see inside one.
Those things are fantastic and neat, I got to use them in the early 2000s and some of my family still uses them at home. It's probably quick enough for realistic 1-computer needs. You ever get a chance to use power line networking over the last few years?
Nah I have ethernet cables going everywhere. Never had a need for the powerline adapters
Would it really have been that hard to put in a simple 3S balance charging BMS? I got 2 of those for 2 dollars and use them for low voltage cutoffs (they do need to be reset by connecting a charge pulse to the output every time they activate, but that can be also reset by connecting the output negative to input negative, after recharging and reconnecting the battery, of course. That resets the charge chip). They could even have kept the individual one cell BMS' as they wouldn't have interfered with the 3S balance charging BMS.
I made a similar thing with a slim aluminium extruded case and 3 18650s inside to allow me to use a little 12V power tool I have without having to use the mains adapter, but for the sake of an extra quid fifty I stuck a 3 cell BMS and balance board from ebay in there too.
and re the voltages of the DW01, it depends which version. theres a DW01a, DW01b, c etc etc, all the way up to g I think, each with slightly different detection and release voltages for over and undercharge
That will suffer a quick loss in capacity every charge cycle.
The least charged cell will die quicker and quicker each cycle since the main current support exists in the 4-4.2v range and anything below 3.7 suffers much greater voltage drop at the same amperage.
Because the highest charged cycle will perform at 100% it will require less charge which will create a domino effect on the least charged cell getting ever less recharged.
I once had an exact same looking 12 V power brick explode on me on first use . .
Should have the pictures somewhere still.
In work we have some batterys that are 24v 10Ah. They contain 28 18650 cells ( 7s4p ) and they have NO Balancing or protection. The carger for the packs is a 10A charger.. Things can go really bad :)
No protection circuit under the shrink wrap?
Nice video, didn't even know these existed! They're going to be a bit dangerous though, with the lack of balancing, as soon as the cells drift (which they will do, all lithium cells in series eventually become unbalanced/one cell loses capacity first) then the entire pack will be limited by the cell with the lowest capacity. It'll also damage the other two cells (or at least the highest) by making it hit its protection cut off voltage all the time, which you don't want to do to keep a good lifespan of the battery. Just keep it to a balanced setup @ 4.2V per cell and it will work fine.
P.S. If you ever decide you don't want those cells then let me know; I always have a use for lithium cells in projects :D
Can I please make a teardown request Clive? There are replacement rechargeable 9v batteries available that use 2 x LiPo cells and are USB chargeable. They sound promising but only if the cells are balanced and the output is modified to give near 9v over most of the discharge range... Thanks for the videos.
I have some but haven't really put them through their paces. The constant 9V output desirability depends on application. A lot of stuff regulates that down to 5V anyway.
I just ordered a couple from AliExpress because I hate single-use cells and batteries - especially the 9v which are quite expensive. That said the rechargeables aren't cheap either so I hope they work out OK.
Me too. I haven't really checked the capacity but the main use is occasional use so recharging frequently is not an issue and if I accidentally leave the devices on, I haven't wasted an expensive battery.
It has one of the micro small 1 amp max charge controller. I have some of those, but I purchased some that can handle 3 amps rather than 1. I blew one of the 1 amp boards.
Who else would like to see Clive do a vid of just cells going nuclear? He can bring the containment pie dish outside.
And I thought patrons were better than that “first” bollocks.
They get first dibs on firsting.
Apparently not! 🙄
bigclivedotcom now that's funny right there.
Maybe that's the reason we pay!
It's the main benefit of being a Patreon, lol. :)
Hi Clive, I want to create a 12 to 18v battery pack using lithium cells, say 3x4.2v (12.6v Max). X4 packs in parallel (3000mah per pack) making approx. 12000mah pack. Can you tell me what the best way of charging this packwould be and best way of protecting it with say a bms board? Much appreciated. Look forward to your reply. Love your videos. Excellent work
Ok, here's the thing about balancing. People get a little too hung up on voltage.
Not all cells have the same capacity, so on one cell, you may put 1900mAh into it before it hits 4.2, and on another, you may put 1980mAh. This discrepancy is normal, even perfectly matched cells will age differently over time. So you have to balance them to get the most capacity out of the pack.
Now, you can top balance - that is, make sure they all hit 4.2V simultaneously. But, because the capacities differ, when you pull 1900mAh back out, you will deplete the "smaller" cells before the "larger" ones. This means the smaller cell will hit its 2.5V (or whatever threshold you set) while the others are still above that. So when do you cut off, exactly? When the cell hits 2.5V under load? Or when it hits 2.5V with no load? Batteries sag a little under load, so 2.5V while supplying 2A is not necessarily empty. But if you keep discharging, things start turning sour. The voltage curve at the end of charge is very steep, so it doesn't take much to drop the cell voltage down to 2V, or 1V, or 0V. When that happens, you set yourself up for reverse polarized cells, and then comes our good friend, fire.
If you bottom balance, then all cells hit the steep part of the discharge curve at about the same time, so the ability for the battery to deliver current falls swiftly and it becomes quite clear that it is truly empty. Even still, draining to 0V when all cells are balanced is not necessarily terminal, since the possibility of reverse polarity is minimal. (There's no current left to reverse charge the first cell to hit 0.00V.) But.... you DO have to watch the top end - when charging. One cell will hit 4.2V first, and at that point, you need to stop. It doesn't matter if the other cells are only 3.7V - you've filled the capacity of the lowest cell, and that's all you're going to get. When you discharge, you'll take out what you put in, and all cells will again hit bottom at the same time. The top voltage is all but irrelevant. And since you typically charge at a constant current, it's a lot easier to control the HV threshold trip point than trying to guess whether you've really hit the LV trip point under load.
Monitoring circuitry is still nice to have, but it's not really critical to see every cell's voltage. You're always charging from 0 to some capacity, and then taking that same capacity out. Since LiIon cells don't self-discharge unless they're damaged, they won't drift apart appreciably over time. A per-cell HV cutoff is really all you need. In fact, monitoring individual cells puts the tiniest load on them, and creates an opportunity for imbalance since the cell monitoring current (albeit small) won't ever be precisely the same. Then you DO need re-balancing circuitry, and you've just made things more complicated.
So this simple design is not terribly safe in the event of corner cases -- such as internal shorts or some other abnormal imbalance - but, in the normal course of operation, it does exactly what it needs to do to keep the cells happy and no more. (Although 4.25V is a little too high for my taste.)
Nick Wallette Thanks, nice analysis. Also liking, "and then comes our good friend, fire" :-)
Wow, I've been working with lithium batteries for years and I had no idea how much went into balancing!
If you've ever seen professional balancing circuits, they have transistors which shunt resistors across the cells to make it balance more effectively and not overcharge any cells.
Yeah, that's "top balancing" -- the art of chasing unicorns... until you eventually become aware of the fact that they are actually dragons.
The goal of top-balancing is to have all the cells charged up to their maximum voltage, plus or minus margin-of-error. Shunts across a cell are a controlled discharge, with the aim of bringing that cell down below its maximum so that the entire series string can be charged up some more. The highest cell is continually knocked down until all of them arrive at the same high watermark voltage, all at the same time.
Sounds like a laudable endeavor, but there are some fundamental problems with that approach.
1) The voltage will drop after charging stops. The chemistry of the cell settles for a little while after end-of-charge. If you measure right away, then measure in 30 minutes, then measure tomorrow ... the voltage will have dropped. This will cease after some time, but it can be days before the difference gets lost in the noise of the measuring device.
Balancing circuits will dutifully pull those voltages back up again, by discharging and recharging, in an obsessive attempt to keep them at 4.200V. The more they mess with it, the longer it takes, and it does nothing at all beneficial. The additional energy put into the cell wouldn't light the charging indicator LED for 30 seconds.
2) Li-Ion cells do not particularly appreciate being at their max voltage. This is stressful for the cell. You're far better off skipping the constant-voltage portion of a charge cycle, and just cutting off right when it first hits 4.2V. Or even a little less. How about 4.1V? If you want >1000 charge cycles, maybe go with 4.0V.
3) You are not increasing the capacity of a series pack by "topping off" the cells. The battery will hold a set amount of energy, no more no less. It's not wasted capacity, it's just a surface voltage that will dissipate almost immediately under load.
4) Continuing on the theme of #3, when you balance at the top, you necessarily create an IMbalance at the bottom, which is a LOT more difficult to monitor effectively during a variable-load discharge cycle. You then MUST have a monitoring circuit that prevents over-discharge of a single cell - a problem that is not trivial to solve, due to the reasons explained in my first post.
5) All the added complexity to pull off this feat of engineering becomes a liability. The batteries themselves are predictable and reliable. Think about it... How many cells spontaneously burst into flame, just sitting there attached to nothing -- compared to self-made disasters where the electronics went off into the weeds? As always, KISS.
Easily one of the best youtube comments I've ever read. Thanks.
At least it has protection.
Though, if one adds a diode and resistor over each protected battery, then when the protection kicks in, current can still flow through the diode and resistor. So that one at least could charge it more fully and even it out at the top at least. (The diode is so that we don't discharge the batteries when not charging, and the resist is there to make sure we aren't dissipating the full 4.2 volts in a diode. And yes, this "balancing circuit" doesn't really do anything during discharge at all.)
Have you ever torn into a DJI drone smart battery? They cost $50 and do their own balancing and communicating with the drone and transmitter/controller.
Had a few of these i use for bike lights (security camera battery packs). I was pondering adding some leads for balanced charging using a lipo blance charger i have. Dont want to burn the house down though. Whats the explosion risk? 😞 disapointing to know i wont get full use of the capacity if one of the cells is dipped. Any workarounds for that? (Im awful with circuits but here to learn 😅)
Could you discharge one of the three cells and dis-balance and see how they charge
3 cells is a battery, it doesn't have any other battery to balance with...each cell has a maximum voltage after that cells protection circuit opens and stops charging only that cell the others continue charging ultimately all equal in the end...
Ive seen them use the High Drain feature to get to their rated output, Just like those jump starters do, listing a 800 Amp output but 3 cells each about 4000mAh at 3.7V, searching specs will show a high drain feature of 25A and Im guessing they twist the numbers around and use the High drain and the peak surge to come up with a WHOPPING 800 Amp when actual is at the 4A at 11.0-12.6 V . I have SO MANY modules and batteries here Id be more than happy to share with you some like my huge flat Li-Ion cells that have real specs of 3.7VDC@ 8Ah and a high drain of 200 Amps (Get this and its true, I did a high drain of 100 A and stopped there LOL)
Clive, you're regularly showing us that cheap and nasty stuff off e-bay is just that, as well as frequently dangerous too.
Well the list of things to avoid (and even throw-out in cases where one was the once proud owner) is getting longer with each video. The list of "great" stuff is very shor, although you've occasionally come accross some "OK" designs, could you please keep an eye open for a (globally distributed and internationally available) item, such as a high capacity (10,000mAh+) USB power bank which is EXCELLENTLY designed and implemented, and share which one it is? Something which has over-discharge protection, over-charge protection, maybe something which even looks after its cells keeping each one of them in that long-total-lifetime sweetspot of 30%-80% capacity...
I've got a "S8 30000mAh made in china" pink bank with a solar panel. Exposing it to sun made one cell fucking explode. The circuit inside is so inefficient, that it discharges cells with the solar panel, rather than charging them, and the circuit is so bad, that even though there's 3x2700mAh3.7v cells, they charge just 5% of a 2700mAh phone in 1 hour, then go dead flat 3.0v. EDIT: One of these i.ebayimg.com/images/g/UhcAAOxyi-ZTYbJ0/s-l225.jpg
this is interesting, I know its not normal for a balancer to fail in a way that it can inversely charged any cells promoting damage to the cells. but I wonder if the cell chemistry is getting to the point of the lower amperage cells are safe enough to do this? or will the difference in voltage force a current out put between the different charge batteries would be just as destructive as it as been supposable has been in the past. I would imagine that so long as the cells are as close as they are of don't stray from there safe charge/ or discharge levels, It wouldn't be to bad. I have been using 12 LGdbhg218650 and B18650ca wired up for 12.8vdc max charge using only inline single cell protection. but I also have not ran either cell type up to its max discharge rate. they have been discharged using a 5v 50w regulator running a Raspberry Pi 3 B, a 7 inch touch screen and a keyboard as well as a mouse, a few times with a usb to Sata connector for the Raspberry Pi. In this configuration, no cell dropped below 3.7 independently and only a 0.01v variance between cells.
dont each battery have an independent charging board? as youd see in a single cell
Yes, over/undervoltage protection, but since they're in series,as soon as one cuts out, the whole string is interrupted... there is no separate power feed to each cell.
They were like, "Let's not completely lie (6600mAh) but let's not tell the truth either (2200mAh). Just slap on '4800', that sounds believable."
Pretty interesting device
The cells will 'drift' over time due to different internal resistances of the cells (ESPECIALLY since they are not a brand name cells manufactured with higher consistency,) and resistance within the soldering connections and slight variations within the protection circuits themselves. When we use 18650 cells in series for high powered flashlights for eg, IDEALLY, we use cells with the same internal resistance (not much of a problem with Sanyo, Panasonic, etc) and charge them to the same voltage before use, AND we do not discharge the cells till dead to reduce the chance of damage to that one cell that may be weaker than the others. The cautious user will use for example three protected cells in series knowing that the protection circuit will kick in and prevent overdischarging of a single cell, but, the observant/ aware user will simply use three unprotected cells and monitor them carefully during use.
Hi Clive, probably not the best place to ask this but what would be your recommendations for powering a homemade bluetooth speaker. Input to the bluetooth amplifier card is between 7 and 20v and ideally want to charge it as I use it connected to a USB charger port. Any help gratefully received.
Looks like an amateur-made battery pack. Someone has acquired a bunch of li-ion cells, stacked them up, up to 12V, and jammed them into a generic electronics enclosure together with some simple circuitry.
in some mobilephones, the Li-Po battery is rated to be charged up to 4,35 or 4,4V and all apps mesure this voltages in the battery so, maybe some batteries can be charged with up to 4.4? instead of 4.2??
Not only is the whole pack capacity limited to the lowest capacity cell, it is also limited by the difference in state of charge of the cells. In the worst case, if one cell is fully charged, and another cell is fully discharged, the total pack amp hour capacity will be zero. When you try to charge the pack, the fully charged cell voltage will rise above the cuttoff 4.2x voltage and break the circuit, and when you try to discharge the battery the fully discharged cell voltage will drop and break the circuit. You can "fix" this problem by manually charging each cell to full capacity with a bench power supply, and hope that the state of charge doesn't "drift" out of balance too quickly. Sometimes you will get a bad lithium cell which self discharges sitting on the shelf- something to do with spikey crystal structures that grow inside the cell, which will cause the pack to go out of balance more quickly. That sound right to you guys?
almost all power tools ice taken apart dont have balancing in their 18650 packs, even the namebrand ones, not even balancing cable for the charger. i heard that manganese cells dont require one, not sure about that
I suspect that the 10S battery on my e-bike has a similar dodgy protection topology, but the battery case is so tightly packed that I can't safely remove the individual cells and restuff it. Any ideas on reconditioning the cells in situ?
I tried to make a 12 volt battery for my hand held CB radio which came to me used with no battery pack. I soldered three of those same type batteries together and charged them up. Well somehow the CB was left on for a few weeks, and when I picked it up, the battery was dead. I opened the battery box and the center battery had swelled to the max, the other two batteries seemed OK. I tossed the swelled battery and still have the other 2 however I decided to use an external battery pack when needed for that old CB. (large GE hand held looks like a WWII Walki-Talki but in near new condition) 3 channel 3 watt radio with huge pull out antenna!)
hi bigclive. i am wanting to build a battery bank. but im not 100% sure as to the charging of it. i intend to use a bms board to control the charging and discharging, but need to know if the charger for this system needs to be a dedicated lithium battery charger (due to the needs of the cc/cv stages). or can you simpmy use a generic battery charger and let the bms take control of the charging stages? can you shed any light on this for me? much appreciated. keep up the good work! love your videos
what if a fairy high value resistor was used to shunt the cells to each other? would that balance them and allow it to work normally too?
Electricians are supposed to make „eeeewww“ noises while handling dangerous tech to make sure nobody touches them if they go bzzzt
This kind of nonsense makes me want to build power banks with all of the cells in parallel. Need a higher voltage? That's what boost converters are for...
Oi Clive I want my screw drivers back 🤣🤣🤣🤣🤣🤣🤣🤣 hope you are OK and safe
I don't usually post but -thanks BC for your content, it's taught me a lot, thank you buddy.
Multiple protection curcuits are really key these days but can be lacking
if not balanced and not precisely matched, I'd expect the difference in charge between the top charged and lowest charge ones to gradually increase over time until the pack stops being usable for no reason other than you can't charge the top cell or discharge the bottom cell anymore and you'd think it dead.
At 8:22 a little crater you say, so that is where the magic smoke did a great escape then?
9:42 is that the good smell or the bad smell?
10:50 J3 Smudge - get the flux out of there
You and your 230V death sockets... LOL
Doing the volts comparison at 5:00. Isn't it ironic that the smallest looking cell (top one) has the lowest voltage, highest voltage is the middle cell which looks to be the largest looking cell & the midway voltage is the bottom cell which looks to be in-between in size comparison?
Yup, using the battery safety circuits for cut offs are the new cheaper batteries. I had a Bluetooth speaker that did the same. It’s not good, the protection is also within %5 to 10% accuracy as well. Protection circuits were for preventing one cell running away into another cell and stopping thermal runaway. It was never intended for charging, one less thing to make it cheaper and charge you the same cost to buy. I don’t like this method at all.
Please discharge one cell only, then charge all up and see if they balance again.
All bike light from ali are supplied with such 8.4v batteries without balancing. They must be periodically balanced manually.
Powerbank without balancing? I'd call that portable firestarter.
kjur18 portable hand warmer
so its really only a 7Ah powerbank and the 12v figure is way optomistic, though it at least claims that the 12v is not regulated. too bad the 5v chip is toast so we cannot see the efficiency.
Having worked in the hobby industry since lipo cells had no balancing input! I can concur what actually happens is you get 3x cells that should be 3.7v nominal so 11.1v 3x 4.2v no load per cell means 12.6v but as the cells drop off you still get 12.6v no load but spread between different cells.... I.e 12.6v fully charged but with silly cell killing variances like 4.2v cell one 4.0v cell 2 and 4.6v cell 3 this caused the rugby ball "puffing" of Lipo cells and even explosions and fire of generation 1 cells, gen 2+ saw protection circuits and balancing introduced as well as fibreglass tape to stop the over expansion of cells and inevitable fire etc afterwards.
Well, in the bad old days of NiCd and NiMH battery packs for cordless drills, they didn't use balancing either. The weakest cell would dictate the lifespan of the pack. It would ruin it once it had degraded far enough to allow reverse polarizing, and I believe that especially NiCd wouldn't recover from that (not sure about NiMH).
IMHO the most important reason to use balance charching on (plain) Li Ion cells is the safety. Mild overcharging NiCd or NiMH wasn't particularly dangerous, but something Li Ion can have a much more severe reaction to, as we know.
I guess this particular power bank was slapped together using protected cells and a "meh, that'll do" attitude. A self respecting manufacturer would have used balance charging to also optimize the overall lifespan of the pack.
Can I make a video request? Please do a tear down of car seat massager(cheap Chinese version).
Is it going to expose a meat grinder under the fabric?
It's going to be a disappointing small bauble the size of a Kinder Egg stuffed with a motor and an eccentric weight.
Not explosive enough. I wanna see one of the supplemental restraint units that blows up to yank you back into the seat.
Jonah Beale You want to see a tear down of an airbag... I read there were some that had to be replaced as part of a recall... should be able to get one cheap.
Cheating the figures? Are you implying that my 300,000mAh isn't really what it claims?
Yes. It may only be 299,999mAh.
Balancing is often seen as a good thing but in fact that is a bit useless. If at manufacturing they are discharged and then charged until one is full then even if one is only 80% full then it's just that that one has 20% more capacity and all the cells actually hold the same absolute charge. When you discharge them again then they all go back flat at the same time.
The balancing circuit is really helpful if you want the maximum possible energy out of the system because it makes sure that the charge is delivered back with the highest possible voltage but the Ah capacity remains the same, limited by the weakest cell.
Another slight advantage of balancing is that aging is probably more even because it's charging to full capacity that wears the cell. But this can be seen as a drawback because it doesn't reduce the wear of the weakest cell, it just increases the wear of others...
You're assuming that if not "top balanced" that the cells will somehow stay "bottom balanced" and always reach their minimum voltages together. This is definitely not the case. For any hope of long term consistency you need some kind of balancing. This becomes especially true as the cells age, or the self discharge rate goes up a bit in one of the cells.
Cells have a coulombic efficiency very close to 1 ie charges in and out are equals. But the one with the lowest efficiency (0.999) are the one with biggest charge % so yes, if there's a loss, it's on the weakest one. So it's the one that reaches 100% charge that will reach 0% charge first. So actually we still have one of the cells that does the full 100% - 0% trip. It's just that with more wear of the weakest one, the others will not do a 0% - x% trip but a y% - x+y% trip that delivers a little bit more power because of higher voltage.
And of course I assume that the cells still have individual protection of over charge and discharge.
I was about to buy one from amazon
They look like one poly lithium 1600 mah I've watched on eBay
hi i want to cut the power from hdd s with a switch 12 v , 5 v in one button can you help thx
9:38 - like "cool whip", right? Family Guy anyone? :P
Hi Clive,
You should be able to create the imbalance fairly easily by discharging one of the cells partly, no? From then on the imbalance should persist for a fair number of charge cycles.
The need for balancing depends, completely, on the manufacturing quality control program. i.e. no scrimping on anything.!!
FWIW (I imagine big Clive knows this) J3Y seems to be a pretty common marking on cheapo surface mount NPN transistors, so that would be a very plausible auto-complete for the flux blob.
They were probably saying it's going to be a 4800mAH equivalent if used as a 5v power supply. Buck converters step down voltage and output lower voltage at higher current with high efficiency, so you still put most of the energy from the battery to the load regardless. It's just that the battery is discharged at much lower current than if it was a regular 3.7v pack of 3 cells in parallel with a boost circuit.
Everyone should just go to mWH anyway
These cells are probably so crappy that some of them will never reach the cutoff voltage. So if you use a balance circuit it would never stop charging. You can't sell that, so off with the balance circuit and fingers crossed you hope you get at least 1 good cell out of 3.