Lithium cell balancer reverse engineering with schematic

[@brettd5884]

Cell/Pack protectors usually monitor for several conditions - if any of them occur, the cell/pack is turned off. Those conditions are 1) over voltage (about 4.25V), 2) under voltage (about 2.50V), 3) high temperature, 4) low temperature, 5) high current (heavy load or heavy charge), and 6) short circuit. Pack controllers usually place a small load current across high cells (4.20V?), and may disable charging (using the protector circuit as a switch) while the single cell is being discharged. In a simple system like this, there is no communication between the balancing circuits and protection switch. Both cells could be shunted by their respective loads without the protector ever stopping charging. Fun factoid: While current is flowing through the MOSFET body diode (MOSFET turned off), the diode develops a voltage of about 0.7V. If the MOSFET is then turned on, the MOSFET will short out the diode - the voltage drop goes to 0V (=I * Rs). MOSFETs that are turned on conduct current equally well in both directions.

[@omsingharjit]

How body diode can be short out to when mosfet are turn on ? Body diode is connected reverse bised Respec to Mosfet Turn on Conduction !!I mean it will show but in opposite way .

[@davelowets]

@@omsingharjit The near 0 resistance of the Fet, when it's turned on, is across the diode and shorts it out. It would be no different than manually placing a wire across the leads of a diode in a circuit.

[@plainedgedsaw1694]

@@omsingharjit explained in another way, mosfet can switch AC (with isolated gate drive) but will always pass more than 0.6V in one direction

[@omsingharjit]

@@plainedgedsaw1694 I mean , AC like wave goes from zero to positive than zero two Negative but when you have circuit which operates at vcc and gnd than your signal can only go between zero to Supply Vcc , it can also be sine wave but perhaps it's called DC offset sine wave not AC sine wave , and in puss pull mosfet configuration it's possible to get ac through load by discharging and charging one of each cap at a time which is connected access single supply same topology used in Pc ATX smps using BJT .

[@plainedgedsaw1694]

@@omsingharjit yeah, i was just describing that mosfet is acting like a switch with diode across it.

[@brucemoyle7610]

An important thing to add is that there are two types of these BMS boards - 4.2V ones for LiPo, IMR etc and 3.65V ones for LiFePO4 ones. Sometimes eBay listings are slightly misleading if the seller is non-technical, so it's worth googling the part number written on the board to see which one you're getting.

[@robjordan63]

It would be interesting to see an experiment to check if your theory of a resistor between P- and B- would work as you expect.

[@tomvleeuwen]

I would expect the circut to work quite well even on high charge currents. I expect the cutoff to be slightly higher than the balancing circuit, so the balancing would kick in early. If this is the case, and the cells are quite well balanced to begin with, it would not take many cycles of switching on and off the charge circuit until the cells are balanced.

[@v12alpine]

The only downside I can think of is it would bypass the low voltage cutoff too, potentially leading to dead cells.

[@dantronics1682]

@@tomvleeuwen I will have a play with one of these to see if I'm correct but looking at the circuit we are assuming the charging transistor is completely switched off, It might be that its partially switch off so there is enough current to balance the cells since the balancing circuit bypass each battery when it reach the threshold (4.2v)

[@simonwatson2399]

The technique is called top balancing. When you charge a 2S pack, you just put 8.4v across the terminals (with a current limit). It charges at constant current then constant voltage. If the cells are mismatched or otherwise out of balance, one will reach 4.2v before the other. Hence detecting 4.2v for each cell and the shunt if 4.2v has been reached, as the other cell still needs to charge. The voltage across the shunt could in theory be up to the full 8.4v if the second cell has failed short circuit.

[@brettd5884]

In a multi-cell pack, more than 4.30V across a single cell would be an over-voltage fault, and sufficient reason for the pack controller to stop charging and disconnect the pack. Alternatively, if a cell decided to short, a low voltage condition (Vcell

[@ErrorMessageNotFound]

AFAIK, Typically all a simple cell balancer does is slowly drain the highest cell until it matches the voltage of the lower one.

[@lorang9979]

If you watch this video, you will learn that this balancer does not work in the way you have described.

[@dherrendoerfer]

Li-Ion chargers typically use CC-CV as the charge method, in CV mode the current near to the shutoff voltage is already so low that the bleeding circuit of the balancer has a significant effect. If the cell is too badly unbalanced it will strobe-charge the pack which is generally fine.

[@uhjyuff2095]

thats correct, but if the balancer and protection mosfets are correctly programed at the right trigger values the balancing function can still work without a CV slow trickle type charger. Imagine the trigger for cell balancing is 4.18v and the over voltage protection trigger is 4.25v and THEN the over voltage protection release is 4.2 volts the balancer will run continuously and the charger will activate every hour or so for a few seconds. If you keep the charger plugged in it is wasting energy at 60ma forever but at least the cells will get balanced over that time period no matter what type of charger you are using! Then after awhile once the cells are perfectly balanced maybe in a few days or so just unplug the charger and the balancer will drain the pack to 4.18v and then stop wasting the energy and you can have a perfectly balanced pack!

[@dherrendoerfer]

@@uhjyuff2095 Some chargers do exactly that - it's called the bleeding phase, or just balancing phase. The charger rises the charge voltage slightly and lets all cells reach the top voltage. Then it switches off completely. Sometimes the CV voltage is set so this happens automatically from time to time. This is rather stressful on a cell, but it allows for much better pack performance.

[@dodutils]

I really liked the explanation about MOSFETs and Diodes purpose

[@andrewwhite1793]

The same as DW01, overcurrent protection is going to have a wide tolerance as the current sense is the voltage drop across the MOSFETs. Mosfets are specified be maximum RDSon, not minimum. If you are unlucky to get a couple of really good MOSFETs the overload protection could be at a much higher current than expected.

[@jercos]

When charged appropriately up to a fixed voltage, the current flow through these balancers should be minimal, as each cell is held to the target voltage. An "unequal" charge will overcharge (and potentially damage) the "high" cell until the voltage across the lower cell rises to the point where it's charging at

[@great__success]

yes, it would be nice to reverse engineere those "active" BMS....Until recently I thought all BMS had this active balancing function....but when I opened the battery from even a very expensive Kalkhoff e-bike, it only had a passive BMS....which led to all sorts of strange behaviors - sometimes it wouldn't charge, sometimes it would show that the battery was empty, but it would go for another 100 km.....It would be nice if someone did review of aliexpress active BMS, since there are like hundreds types...

[@beniaeschbach2626]

The second scenario you describe is called "active balancing". (Rather than just burning excess off, as we see here) These boards are available too, on AliExpress and the likes. Thanks Clive!

[@beniaeschbach2626]

Haha, some one was quicker to reply. Sorry, should have read to the bottom first.

[@twoina]

Matter of consideration. The additional electronics could be more wasteful that the burnt energy? Indeed, if you go through the numbers, a ebike should definitly have an active balancer, a small 2 cell battery pack should not.

[@gabrielenitti3243]

@@twoina i mean, passive balancing has an efficiency of 0%, as the balancing consists in turning extra energy into heat. Even if the charge pump is only 20% efficient, that's still more than 0%

[@kapegede]

You explained it so well! Thank you very much. So this is a passive balancer (the 68 ohm resistor reduces the capacity a bit until the weaker cell catches up). Furthermore there are active balancers out there, too. And I'm keen to know how they work exactly, too.

[@davelowets]

One of the types of "Active balancing" uses a system that can "turn off" the charger power supply to the full cells, and back on to the cells that still need current, using PWM as the pack approaches it's full capacity.

[@yngndrw.]

@@davelowets Another active type uses a small flyback DC-DC converter to drain the highest cell and spread the energy across the entire pack (Or a larger portion of the pack) - Have a look at Linear's LT8584 IC.

[@justpaulo]

3:04 This is a detail, but the body diodes direction of both NMOS is switched... The reason is that they live in the same chip (as you pointed out it's a dual MOSFET chip). So, most likely, both NMOS share the same substrate and so the anode of the body diode is necessarily common to both MOSFET. Therefore the node between the 2 NMOS should be the anode and not the cathode.

[@rayoflight62]

In some of my very high reliability designs, I couldn't rely on the CMRR of the voltage sensing OP-AMPs to monitor the voltage of each cell, especially for very high voltage batteries with tens of cell in series. I used a temperature-stabilised opti-coupler for each cell as voltage sensors. This design is in totally opposite direction of the chinese designs, which prioritise low components count and short PCB tracks (to save copper), where reliability, immunity to noise and all static working points of active components are sacrificed to the cost of the board. Thank you Big Clive for this video, that exemplify the design criterias to the bones...

[@PaulSteMarie]

With regard to mosfets and diodes, the there's actually two diodes in Harrington the design semicolon but one of the two is hardwired short in most packages. The full outputs from a mosfet would be gate, drain, source, and substrate. There are inherent diodes between the substrate and the gate / drain. Usually source is bonded to substrate in the package since that junction would be forward biased, anyway.

[@maxhammick948]

60 mA through the 68 ohm resistorb also works out to a nice 1/4 W, which seems like a reasonable upper limit for a small surface mount component like that with good heatsinking to the PCB.

[@satiric_]

Never thought about using the mosfets for current sensing before. That's really cool. Not the most accurate but certainly good enough for this application

[@robegatt]

Not the most accurate resistance-on but cool anyways.

[@davelowets]

More accurate than one would think....

[@siberx4]

I believe you're correct that the balancing circuits only have a limited current capability to effect balancing; if you could somehow charge at 1A forever it would "overwhelm" the ability of the balancing circuit to keep the cells at the same voltage, resulting in (temporary) overcharge for the higher-voltage cell. The reality is that this condition really can't occur except in a scenario where the cells are extremely badly mismatched in terms of capacity, and in those scenarios your pack is pooched already. Since the balancing circuit is always active, there's no scenario where you can otherwise get the individual cells into a wildly different state of charge, and any time a single cell is at 4.2V, the whole pack will already be at or close to the CV part of its curve, limiting the current the pack will accept. Charging continues until some fractional current cut-off of the standard charge current (between 1/5 and 1/20 of the normal charge current) as it would in any single-cell charger, and that cut-off current is generally around the same order of magnitude as the balance current (60mA in this example) so that the CV portion lasts long enough to ensure all cells reach a mostly-balanced state before cut-off.

[@lorang9979]

Overcharge will not happen not even "temporary". The protection circuit will detect the high voltage and open the mosfet and terminate the charging process. This happens, when the balance is poor and this balancer will never be able to correct it (as you explained). This is why it is a bad idea to put a resistor across the mostfet as it bypasses the protection and may destroy the cell.

[@dandearman2871]

That shines a light on something I was also wondering about. Thanks.

[@rose-ey6ct]

What a brillinat description of about how cell balancing and mofsets work! I suspected that was how cell balancing worked, but it seemed such an odd method, I was unsure. Thank you!

[@8001010]

I add a Led and resistor in parallel to 68 resistor to see balancing

[@piconano]

Even my Daly 100A 48V LiFEPO4 BMS for my scooter, only balances at 50mA. Charges at 500W until the first cell hits the threshold, then it stops and goes into trickle charge mode @~14W. Daly sells an active balancer with Bluetooth, but the price was the same as the BMS, so I just went with the BMS and it's 50mA limitation. I thought this value was too small, but after graduating from Google university, I found pretty much all brands are like that. Thankfully, I had to balance them once by leaving the charger on for 24Hrs and let it trickle charge all cells @50mA, to the 3.65V maximum per cell. Once balanced, they all seem to charge and discharge the same amount and don't require balancing. I contribute that to matched internal resistance and capacity of each cell. I don't know if that will change with time or not. I figured the worst thing would be that one cell dies sooner and shuts off the BMS and when that happens, I'll let it on the charger for 24hrs until all cells are full and balanced once again.

[@davelowets]

As the pack starts to age, and gets closer to it's end-of-life, it will get out of balance more and more, and take longer and longer to fully charge.

[@piconano]

@@davelowets All other chemistries do the same thing.

[@petski999]

You missed how these actually work. Each circuit operates independently but not necessarily at the same time. You are correct that the large mosfets will disconnect when one cell overcharges, but the balance circuit stays operating and will slowly discharge the high cell. Once the voltage of this high cell drops to a preset voltage, the main mosfets reconnect and charging resumes for both cells. This cycle repeats until the lower cell catches up to the higher voltage cell. With each cycle the lower cell gets an additional amount of charge equal to what was discharged from the higher cell while disconnected. This does not require the charger to limit current or know anything about the current state of charge of the cells.

[@MahatmaMichael]

Thank you Clive for this video.

[@tze-ven]

Almost all MOSFETs sold will have body diodes; it will not go away (anytime in the future). A MOSFETs "without" body diode will not have a fixed threshold voltage (Vth) because it varies with both Source and Drain voltages in reference to the 'body'. Also a fourth terminal (connected to the body) will be given as well and you have to connect it to a voltage that will always be reversed bias (or 0v) to the Source and Drain terminals. This makes the MOSFET difficult or inconvenient to use. That is probably the reason why the 'body' is always connected to the Source terminal so that Vth is fix and you don't have to deal with the fourth terminal, but you get the (unwanted or free, depending on how you appreciate it) body diode as the result.

[@user-rs8zg8ey2b]

Your master control chip has 6 pins, you forgot Vcc. This is the method 90% of rc chargers use to charge balance packs, but the mpu senses the voltages and turns on the bal fets. This also has the benefit of lowering the output current to the ballance capability. Thanks for the video.

[@UpLateGeek]

In theory the current should be well and truly winding down by the time your cells are approaching 4.2V. I believe typically Li-Ion charger are supposed to cut off the charge when the current drops to 10% of the constant current level, so this would work fine if your max constant current is 500mA. As long as the cells aren't horrendously out of balance, it should be approaching 50mA by the time the first cell reaches 4.2V, then the bypass circuit should switch on and dissipate that current until the second cell reaches 4.2V, then the current should start winding down further until it reaches cut-off.

[@celeron55]

It's CC-CV. The constant current phase continues specifically _until_ the voltage is 4.2V/cell, and then the current starts dropping to keep the voltage from not going above 4.2V/cell.

[@detritus10001]

Big Clive, the circuit breaker himself! Always good to see you're talking right down to earth in a language that everybody here can easily understand. (I even played the riff after typing it) Keep up the great explanations, my friend!

[@gimmieliberty6514]

i charge my 10s cycle packs using external balance leads via a hobby charger [b6ac clone] when a cell group reaches 4.2 v charge current drops to a fluctuating 0.1/0.2 amps, which can take a very long time to bring the other groups up just 0.05 volts, so in effect bulk charging with a restistor soaking away current on full groups doesnt work, looking at building a charger with 10 five volt suppllies [for electrical seperation] supplying 10 tp 5100mu charging modules monitered by a couple of volt displays each displaying 5 groups

[@wouterrobers224]

About the charger: Li-ion chargers are CCCV, meaning they charge with a fixed current until they reach a voltage limit (4.2v per cell). Then they keep that voltage. So basically like your lab power supply: a voltage limited current (or a current limited voltage). Once the cells are balanced, there is very little that will get them out of balance, because all the current goes through both cells. Typically there is no trickle charge, because overcharging Li-ion batteries is not like Lead-acid or NiMH. You can overcharge and destroy a Li-ion battery with the tiniest current. So the voltage limit is a hard limit.

[@wouterrobers224]

Source: worked on charging systems for Li-ion batteries since 2006.

[@davelowets]

IF the cells are very closely matched when the packs are manufactured. People building these crazy "powerwalls" today, with many different cells all lumped together, are completely asinine.

[@wouterrobers224]

@@davelowets Matching is important, however after balancing, cells do not tend to drift away from balance at 100% SoC. Still the pack will perform as well as the weakest cell, so matching cells is really important, but not that much for balancing.

[@henrybecker2842]

Thanks for the analysis Clive. Learned a lot today.

[@twocvbloke]

Kind of akin to how solar panels have a bypass diode, so if one panel is in the shade, it'll just pass the current from the other panels it's connected to until it has the power to join in with the fun... :)

[@jspencerg]

We just rewatched one of my all time favorite movies. "Waking Ned Devine" has your oft mentioned storms and some beautiful sunny days to show off your home. How about some vintage explorations? It would be great to see your thorough explanations applied to some non-microprocessor based units.

[@tigercat3864]

Typically a Li-Ion charge controller like TP5100 is used with a 2S balancing BMS, which limits charge current properly and works well. But buyer beware - many "chargers" (e.g. eBay "8.4V 1A US plug 1000mA charger adapter for Lithium Ion Battery Li-ion LiPo 2S") sold are just 8.4V power supplies with no Li-Ion charging logic.

[@davelowets]

Such garbage. They take forever to charge a cell, you never know for certain if the charger has completed a charge or not, and it can harm cells that are out of balance by over volting one cell while the other one is still low. 4.2+4.2=8.4 volts, but SO does 4.6+3.8=8.4 volts.... 👎

[@maxtorque2277]

"balancing" without any kind of organised higher level BMS is not really balancing, but actually "charge shunting" Ie say Vmax is 4.2v, the first cell to reach say 4v has it's shunt resistor turned on, and as a result is now being charged slightly slower than the cell with the voltage below that same 4v point. The aim is that by the time the 4.2v and the hard shut off is reached, that charging imbalance will have had enough time to have caught up the voltage on the cell not being balanced. Proper "smart" balancing that is necessarily controlled by some form of higher level control logic can both do this earlier, ie at a lower cell voltage and hence have more time for balancing, but also can signal a reduction in charging current to also help the balancing occur as necessary. For small simply cell systems, then the basic shunt type arrangement works, but it mostly doesn't do that much, and it take a significant time to do it. It relies on the fact that hopefuilly the cells are both well matched initally but also age at a similar rate!

[@jimsmindonline]

A follow up on the capacitor based balancing would be awesome. (Julian Ilett covered it a while back)

[@LibbyTarde]

Big Clive is so rich and wealthy! He has abundance in the ultimate commodity of intelligence and passion!

[@brynyard]

A fun problem with many high-power mosfets (and also diodes) is that they have a failure mode where they short all pins. Haven't dissected one to see why, but this often creates havoc in the rest of a circuit (for some bizarre reason).

[@jercos]

High voltage breakdown, a little static shock will do it. This happens to diodes as well, and is put to use in one-time programmable memory. (antifuses)

[@brynyard]

@@jercos I've primarily seen this first hand in high current devices, like a truck-battery charger supplying 48v/100+A (with quite spectacular results), though this could have been cause by static discharge.

[@luxmonday]

A lot of battery safety testing is performed with the FETs deliberately shorted... you must have a UL rated fuse for short circuit. Bonus points if the fuse is a "three terminal" version with secondary over-charge detection... an entirely separate set of circuitry blows the fuse on over-charge. That way you can have 1 fault (a shorted FET, either charge or discharge) and still pass over-charge and short circuit testing. Adding a large Zener diode to the output helps prevent damage to the FETs from ESD. Also adding 2x series caps (again 1 can fail) across the charge and discharge FETs routes the ESD spike to the cells and away from the silicon. By the time you've finished designing the BMS most of the circuitry is there for edge conditions (balancing, shorting, over charging, ESD, over temperature etc)

[@dimitar4y]

As far as I know (before video) they just slap on a resistor across the batteries that are too high to bypass them during charging and to help discharge them a little. An isolated comparator per cell.

[@cmuller1441]

Don't forget that the cells are supposed to be initially balanced. Before connecting then in series you should charge them to 4.2v in // for many hours. Also li cells almost don't loose electrons. IE in a cycle you get the same number in and out. The only losses are from unwanted side reactions that ultimately lower the cell capacity but that's a very slow process. The balancing circuit purpose is just to maintain the balance not create it from a set of completely unbalanced cells. When reaching the 4.2v limit, the differences is in the order of mV.

[@Fluxkompressor]

Exactly. If your cells are good from the beginning (ie no self discharge) then you should not need a balancer at all (although it is good to have one for that little bit of drift) If one cell degrades sooner than the others, this one cell is under much more stress since it reaches the charge and discharge cutof voltage before the other cells, so it degrades even faster. At this point your balancer cant help you much anymore. But since your cells were supposedly equal from the beginning, all the cells should be near their end of life now anyway

[@myfavoriteviewer306]

I think that screen printing is the most safety conscious and responsible thing I've seen on any of the circuits on this channel. Pretty neat, although I might just be easily amused 😁

[@markeverett7642]

Our chargers (and I can't say who our is) stop the high current charge before the protection kicks in. So the balancer comes into play during the maintenance trickle.

[@qbitsday3438]

Subscribed Instantly ! Your Videos are Great !

[@wherami]

yes both chargers I have used do the balancing at around .6ma.

[@dennisyoung4631]

Odd thought: perhaps replace the 68 Ohm resistor with an LED and a 15 Ohm resistor? This might give a (simple and crude) *indication* when balancing is actually happening? Another thought is to put a pair of tip-jacks in one of the (metal cased and dismantlable) Ebike chargers I have, so as to more-readily test the actual output voltage of the thing?

[@TehFreek]

ISTM that you could do the charging via B+, BM, and B- with a smart three-terminal charger, and use P+ and P- for output only (or brute force quick-charge, of course). That would involve more wires, but would maximize balancing effectiveness.

[@poptartmcjelly7054]

Some smarter balancer IC's also balance based on cell difference. For example some of the BQ series chips from texas instruments define balancing voltage as a difference between cells. So it balances at all times during the entire charge cycle and not strictly just during the CV phase of the charge cycle. This is better, because lithium cells don't like trickle charging and for these dumb balancers to work the pack needs to be trickle charging. The dumb balancers also pose a risk of micro-cycling, where and overcharged cell might be discharged to balance and then allowed to charge again to reach balancing voltage only to be discharged again, therefore causing many micro-cycles.

[@carlubambi5541]

Great break down as always .handy for making independent power sources for usb charging phone or powering small devices

[@agentmango]

My guess is that if one cell is full and other isn't, the full cell's protection resistor will be dissipating in excess of 62 mA (i.e. in excess of 4.2 V). The resistor, plus the non-full cell, in series will be taking however much current gives 8.4 V. Thus the non-full cell will (only) asymptomatically approach becoming full. The high level takeaway from this video is that some battery packs will only get balanced when charged to full. It would be nice to see BMS balancing systems that work at lower SoC become more ubiquitous.

[@asificam1]

Active balancers can balance a LOT more current and most of them operate regardless of voltage... they just care about relative voltage from cell to cell. Sadly, they are often standalone modules.

[@Ed.R]

I use these on the 2S 18650 battery pack on my bicycles automatic shifting. I got them for the protection and didn't believe the balancing part of it. However the packs have stayed perfectly in balance and maybe this is why. I have never balance charged the pack and it receives a lot of charge from a front hub dynamo.

[@davelowets]

That would be why... 👌

[@richardturton6900]

Didn't Julian Ilett do a load of videos on cell balancing a couple of years ago? I think it was mainly "super caps" he was working with instead of batteries.

[@keithking1985]

Thanks Clive 👍🇮🇪💚

[@apanoiu]

High quality content. Thank you!

[@FireballXL55]

I have always thought that the charge circuit needs to back off and only charge at the BMS bypass current. But do not see this implemented very often.

[@phonotical]

Imagine trusting protection from a major failure to a few microns of paint...

[@yngndrw.]

Linear have some interesting active cell balancers, take a look at the LT8584.

[@acmefixer1]

What I would like to know is where does each HY2213 get its positive and negative power from? Another thing that's important to anyone who stores lithium cells is how much current does this circuit drain from the battery while it's sitting on the shelf?? Thanks, Clive.

[@SeanBZA]

Power comes from the sense inputs, and they normally use a tiny amount of current, 100ua or so, or even less for the ones with low power as intended use, with integrated mosfet switches, so they basically will operate over a year on your standard 18650 cell, without using more than the self discharge already does.

[@janemba42]

This video started quicker than ElectroBOOM can say full bridge rectifier.

[@gabrielenitti3243]

As you can't just charge a single cell as the input voltage is meant for two cells, why not balance between themselves? Like for example, when one gets to 4.2V, you turn off the charging mosfet and then you connect the cells in parallel with a resistor in between. When the voltages get equal, you start charging again until one cell gets back to 4.2V and then you repeat until both are at 4.2V and you stop charging. It may not be the fastest, and not the most efficient (half of the balancing energy is wasted) but quite simple and effective

[@pleasecho2]

Nice detective work on ID'ing the chip. There are SO many 2S charge modules out there. How is this any different or better?

[@omsingharjit]

I always Wondered how to increase charging current of the board ? I guess I could change the value of I sense resistor but I have to try .

[@redstoner2006]

so... i have used something similar, and for the sake of seeing how well it was charged, i shoved some LED's across those big resistors. the protection IC most likely wont trip unless it really needs to and it has some room to let the ballancer ICs do their job. also, the protetion IC will constantly trip the overcharge if once cell is grosly overcharged and it lets the balance IC bring that cell down when you're using higher currents at the earlier charge cycles.

[@davelowets]

Have you ever heard of a multimeter? Or a $2.50 Lipo checker? LEDs across the resistors aren't going to show you much of "how well the battery is charged". 😕

[@superdau]

@@davelowets In this case the LEDs will show how well the battery is charged, because they light up when a cell is at 4.2V.

[@davelowets]

@@superdau an LED will light up over a whole host of different voltages. They certainly aren't going to indicate 4.2 volts. An led can light up anywhere from 2.5 volts on up to 4.2 volts.... WAY over the range of any lithium cell's diacharge curve of 3 volts to 4.2 volts.

[@davelowets]

@@superdau One could design a circuit to light an LED at precisely 4.2 volts, but it's going to alot more components than just a simple LED across a resistor in order to be accurate, and not just light over a wide range of voltage.

[@superdau]

@@davelowets You don't understand the circuit. There's either 0V or 4.2V across the resistor, and nothing in between, depending on the charge state of the cell.

[@grantrennie]

Thanks for the good video 👍

[@Simcore999]

It s a very slow balancing, indicating one of the two batteries is most likely defective anyway

[@spacecase0]

These are my least favorite cell balancing systems. The charger side certainly works with that circuit board in the battery. You can tell when someone has implemented this sort of thing because it takes many many hours to charge a battery instead of only a couple with other designs

[@davelowets]

"Many, many hours" is pushing it a bit.... If it's a HUGE pack with many cells, then yes it could. For a small 2S pack, then no.

[@weerobot]

Another Great Presentation...

[@l00t3R]

Love your vids where you're reverse engineering. What is this module used for? Or what is its common use?

[@richardhalliday6469]

Radio Control model batteries - drones etc. Which commonly use 2, 3 and up to 6 cell batteries.

[@SinKillerJ]

OffGridGarage may be worth a curiosity watch. Andy over there does alot of experiments and testing with cell balancing for solar purposes.

[@ELECTRICMOTOCROSSMACHINE]

Very cool.

[@Qscreator]

this balancer is cool for small bat. but i think it not proactive and one things we need noties is the current balancer smaller the charge current, so if battery on charging that ballancer is not possible

[@willforcey2921]

Love the video and diagrams you drew! Have you looked into / made a video about active balancers, ETA3000?

[@bigclivedotcom]

I've not looked at those yet.

[@uhjyuff2095]

if the balancer and protection mosfets are correctly programed at the right trigger values the balancing function can still work without a CV slow trickle type charger. Imagine the trigger for cell balancing is 4.18v and the over voltage protection trigger is 4.25v and THEN the over voltage protection release is 4.2 volts the balancer will run continuously and the charger will activate every hour or so for a few seconds. If you keep the charger plugged in it is wasting energy at 60ma forever but at least the cells will get balanced over that time period no matter what type of charger you are using! Then after awhile once the cells are perfectly balanced maybe in a few days or so just unplug the charger and the balancer will drain the pack to 4.18v and then stop wasting the energy and you can have a perfectly balanced pack!

[@jean-clauded5823]

A passive balancer. Wish you would do one on a larger, active balancer for Li Ion batteries (like an 18650)

[@dcallan812]

clever little board, 2x👍

[@plainedgedsaw1694]

It's called end of charge balancing, people built these with transistors and voltage references.

[@lezbriddon]

I like the old 'flying capacitor' cell balancing for its improved efficiency but i would hate to try to build it.... (a capacitor is connected by a network of transistors to the highest voltage cell, when charged its connected by the same network of transistors to discharge into the lowest cell) (sold on sites as active balance bms)

[@wherami]

neat. i have been working on these recently as well. just completed my first rebalancing with a home built battery.

[@restojon1]

It's MosFet niftyness, I tell you!

[@cortanajpn]

Not entirely related to the video, but it does involve power supplies, so it’s tangentially related - and I’ve no idea how to make a comment on a channel instead of a video or even if that’s possible. Anyway, has anyone seen a computer power supply fail where (I think) a MOSFET explodes? (Obviously, after watching this channel, I had to salvage something from what is likely a very expensive problem, by dismantling the PSU to identify the fault).

[@paulmccoy2908]

That’s amazingly simple. I wonder if you could do that with some very specific zeners, especially with something like Lifepos.

[@shanejohns7901]

The circuit should work across all types of rechargeable battery chemistry. The issue is the cut-off voltage. The different types of batteries tend to have different max voltages. The controller chip usually works across all types of chemistries as well. But the trick is telling the controller which cut-off voltage to use. And that's often done through a resistor or a pair that tell it which cut-off voltage to use.

[@davelowets]

@@shanejohns7901 No. Not suitable for all types of chemistry. Ni-Mh and Ni-Cd cells use a totally different charge profile and the way the end of charge is determined is completely different than a lithium cell. Lead acid is once again way different.

[@shanejohns7901]

@@davelowets It would work so long as the target voltage is under the max voltage. That may not ensure the fullest charge on the battery, granted.

[@paulmccoy2908]

The cutoff voltage for Lifepo cells is typically 3.65v. 3.6v is a pretty common zener voltage.

[@davelowets]

@@shanejohns7901 We want chargers that CHARGE our batteries...

[@Ni5ei]

Is this from a powerbank or did you buy this is as is?

[@bigclivedotcom]

Bought as is from eBay.

[@rpavlik1]

I always thought the whole "balance by wasting energy" was weird but I guess it's pretty common. The weird thing about this implementation, is that unless you charge it *all the way* it won't balance, so if you try to stop charging at a lower point to save battery wear you might get an incredibly unbalanced pack.

[@ursodermatt8809]

ignoramus! 1) practically all battery packs are balanced at the top end. 2) you cannot balance a battery pack somewhere in the middle, it will almost certainly cause imbalance. 3) if you really must "save battery", it does not really matter how unbalanced the pack is. (unless you then suddenly totally discharge or fully charge the pack. (which in your case would be expected) 4) your wasting energy bleeding a high cell is an absolute tiny amount compared to what is going in and out of the pack.

[@stepheneyles2198]

@@ursodermatt8809 Thanks for your explanation but please don't call people 'ignoramus' - not everyone watching this channel knows as much as others...

[@ursodermatt8809]

@@stepheneyles2198 ah yes, i partly agree with you. however rpavlik1 states his opinions as facts. my attitude is if i do not know anything about a matter, i might ask questions or investigate it, but do not make ignorant statements.

[@dantronics1682]

@@ursodermatt8809 He said he thought, he never said it was a fact, unless he edit his original posting. or is it that in the west when someone says "they thought" or "I guess" it means its a fact?

[@ursodermatt8809]

@@dantronics1682 what kind of balancers do you have in the "east" ?

[@murraymadness4674]

I had exactly the same question about how the 'balance' chips can work if the over voltage protection shuts off the charger input? It looks like they can't unless the protection trips at higher voltage than the balancer, so if the charge current exceeds the bleed rate, the voltage will rise and the charge input shuts off. Maybe that is what you said?

[@agw5425]

On a related note, I would love to see you do a deep dive on the budget/cheap spot welders used to build lithium packs with, like the 10-20£ wish 12v ones. I often thought of buying one but are worried that they may do damage to my pack or my 12v battery/source. Could you see if a regular iup battery and/or a car battery charger (6 Amp) could work as a source of power to run it, would a super capacitor be better/required to make it work/not kill my 12v lead acid battery?

[@SAHILKHAN-lu8oq]

U need atleast 200+ Amps for spot welding. A charger certainly couldn't do that. Super capacitor are not suitable either. Best choice is obviously High CCA rated Lead Acids or high C rated Lipos. One thing u also have to make sure that during the spot welding pulse, under that 200+A load, your battery voltage shouldn't drop below 11V. If it drops below that, then the spot welder kit will destroy itself (MOSFETs blast). And you'll end up saying that the spot welder kit is bad. But it's the battery u r using is bad. One more thing u have to keep in mind that the battery discharges a little bit with every weld spot. To prevent that, and to make every spot weld identical u should keep a very low current charger (1-2A) attached with the battery while welding. I expect I cleared all your queries by now.

[@agw5425]

@@SAHILKHAN-lu8oq So do you think a motorcycle start battery is enough with the charger plugged in or must it be a full car start battery with high crank amps, I´m on a budget so I try find what is enough but not to much. Super caps can start a V8 car and give verry high amps, would that not be better than a lead acid battery?

[@PipeManPeep]

@@agw5425 a bike battery (4Ah) won't supply enough current. You'll want a car battery with a few hundred CCA. A lead-acid battery would be cheaper than super caps.

[@d.t.4523]

Good luck! 👍

[@t1d100]

Interesting. Thank you.

[@stoptheirlies]

Hi Clive, good video mate, what about the third pad, the temperature shut down circuit?

[@AnonymousAlcoholic772]

Hey i was wondering if anyone else experienced this problem. I have a set of Howard Leight headphones, the green ones. They are not bluetooth but they have an audio jack and a weird microphone that amplifies spoken words while dampening stuff like gunshots, as they are intended for shooting. Anyway the mic is what seems like a rheostat where it clicks and turn off on one side and other direction turns on and seems to increase gain on the mic. Anyway I wanted to use these headphones by just plugging in to my PC speaker with the audio jack. As soon as this thing went near my computer, as long as it was on, but before I even hooked up the audio jack, they started making weird ticking sounds. The sounds did not vary with the gain on the mic, and only stopped if I completely turned off, the headphones. Distance from other electronics does seem to play a role in how strong the ticking is. However, it’s not Bluetooth because even when I make my computer go to airplane mode, it has no effect on the ticking sound. As long as the headphones are near the computer, they make the sound. It’s not the router cause I moved it to the other room. Any ideas what’s causing it?

[@suryapatnaik5720]

And they charge 3 dollars for a single piece(board).

[@twoina]

To estimate the current we have to consider the voltage drop on one diode of the first mosfet plus the drop on the on-resistance on the second mosfet? I thought, the drop on the diode is much higher (~0,6V) than the resistance (maybe 0,01V per Amp?). The voltage drop on the diode is lower at higher temperatures, that's not what we want here. :-) What am I missing?

[@daler2577]

The units that I used didn't seem to balance the cells very well as the cells were at different voltages after charging.

[@davelowets]

How long after charging did you check the individual cell voltages? They could have been exactly the same after charge termination, but after the cells sat for a while, they bled off voltage at differing rates. Your cells may be shot, or mismatched, and this could be why you're seeing this. To DIY build a quality lithium pack, you MUST choose cells that are VERY closely matched, AND of the same age.

[@daler2577]

@@davelowets the cells were of the same age and equally matched to internal resistance and voltage. And the voltages were checked within an hour of cell termination

[@davelowets]

@@daler2577 Then I'm guessing you had some crappy cells. 🤷🏻

[@andyash5675]

Does the HY2213 protect against situations where the cell is not charging, but the temperature has changed? If the cell voltage were to increase due to an environmental shift, the 68ohm resistor would act as a "safety valve" for the excess charge.

[@lambda7652]

Wat always worry's me with protected Pacs is what if i put them in Series? One pack will inevitably open first, exposing the mosfets to X time the pack voltage (x being the number of packs in series)

[@station240]

No current flows if the circuit is open at any point, so it doesn't matter what the voltage is. Voltage is relative to the circuit's connections, so the middle pack BMS is only exposed to the the + and - of the cells it connect, eg 7.4V not 74V (10x 7.4V). The only exception would be if the metal case is connected to 0V, and the BMS is bolted to that, then you have to consider the insulating materials (eg the PCB fibreglass).

[@lambda7652]

@@station240 Maybe i did not explain it right. Lets make it simple and go with 1,5V battery's assume you have 10 in series so 15V, a switch and a 15Ohm load. The switch needs to switch 15V and 1A. if the switch is open there is 15V drop over the switch) Now if the switch is supposed to be a Mosfet it also needs to withstand 15V and 1A. So now if there is a switch/mosfet to swtich off in every cell. it needs to withstand 15V not 1,5V

[@piconano]

I was wondering if partial charge and discharge of a cell, counts as a charge/recharge cycle? After googling for hours, I can't find a definitive answer to that. Only opinions. No research papers! For this reason, I don't charge my moped after each little trip to the store. My range is 40Km, and I recharge every 30Km

[@casemodder89]

partial charge/discharge obviously gives you a partial cycle. logic right ? but 3x 33,3% cycle is a tiny bit less then 1 fullcycle as the DoD is lower. and the lower the DoD the longer a battery lasts. that is the reason to discharge a lead acid battery not below 50% SoC. you might get 300x 100% DoD, but 750x 50% DoD. similar with Li-batts. but much higher numbers. like 500 full cycles for a Li-Ion and 3000 full cycles for LiFePo4.

[@piconano]

@@casemodder89 Thanks for that Captain obvious. Plug you phone back on the charger every-time you use it a bit, and see how the battery won't hold charge after 3 years, just as if they were fully charged and discharged each time. Next time you see a question, make sure it was for you little turd. Now be gone little troll.

[@daveclucas4663]

Any chance of you doing a teardown of a solar charge controller. I know the basic principles but I have a few queries in my mind. Does it work just by controlling current by sensing the voltage or does it control the voltage as well. In particular I was wondering if the solar panel is outputting a voltage lower than battery voltage would it be worth putting a buck/ boost convertor after the solar panel to enable it to use more of the solar panel output?

[@andrewwhite1793]

The balancing circuit works when the cells have different capacities after time in use. The weaker cell gets charged first. The balancer turns on stopping it being overcharged whilst the other cell fully charges.

[@macmroz]

It would be nice to have letters description on the schematic diagram, just for less thinking to guess which is which. Anyhow you should have some knowledge to read the schematic diagram, but why not to have them?

[@darkknight145]

Now you need to do a breakdown of an active balancing board.

[@PrincetonTV]

Hello I really enjoy your videos. Can you help me? I want to know if a Gunn Diode is helpful on an AC Solenoid. The Solenoid is a GE 9500 C102r2z with a 115v 60Hz coil. Coil is 15D 9G102. Thank you . I have been researching this for a while .

[@brucereichert6509]

I wonder if this would work on the large solar lawn lights, that take 2 18650 batteries. Because those are built so cheaply that one battery always seams to suffer more than the other.

[@PipeManPeep]

I've only ever seen them with the cells in parallel.

[@brucereichert6509]

@@PipeManPeep I have 4 of them. 4 years ago I got them from Menards. They worked great for about 2 years, then I noticed they wasn't lasting very long. Everything tested good except the batteries. I replaced both batteries and a year later, I have the same problem again.

[@PipeManPeep]

@@brucereichert6509 are the cells are in series?

[@abx42]

Interesting little piece of board, what did you rip that out of??

[@bigclivedotcom]

It came as a bare module.

[@W4BIN]

The diodes in MOSFETs is called parasitic, I believe they were not added, but accidental and unavoidable. Ron W4BIN