Andy, This is great stuff. Keep asking these questions. On another U tube " Emily and Clarks Adventure" Clark has made a hybrid system with a big led acid bank and a smaller LIP companion all hooked together. I wish you would take a look and give your opinion. The idea is the LIP does most of the work each day and the LA stays at or close to float and only discharges in a big demand event like extended low charge days. This greatly reduces the cycles on the LA bank and extends their life. Many boats have a big LA bank so this is a way to get the benefit of LIP at a lower cost and not having to wait till the existing LA bank is dead.
One good reason to charge at a lower part of the top end knee is that your cells will reach full at different rates and you have some headroom for not cooking the weakest cells every time you fill them up if you stop early and let things equalise.
What does it mean - the weaker cell? Why it is weaker? Why it goes up during charge and goes down lower when discharge? Internal resistance- bigger or lower? Capacity, bigger or lower? Looks like weaker cells runs up, bms/balancers do them lower, and thay become not at the same SOC like others. Or if we use float, at 3,4v weak cells can absorb...
@@User1462uuw8w less capacity - eg take a 3 year old worn 8Ah cell and it will likely be 7.5Ah or so, put it in series with 3 brand new cells, discharge each to 2.5 individually then put 7.6Ah into the string, the old cell is full and at 3.65V and the other 3 are still needing more and at 3.2-3.3V on their way up. Continuing to charge without a balance solution overcharges and over voltages the weaker (lower capacity) cell.
I think you got it correct. I have a small pack 6Ah 24V and float the pack at 3.45V per cell. Keep in mind we are talking mV here when was the last time any of us have sent our meters/BMS systems out for calibration? Once my cells reach 3.45V the absorption current drops into the micro amp region. The Cells are chemical electroplating baths, ion movement is proportional to the current flow and not the voltage. No current or very minimal current results in no chemical reaction. However too high or too low a voltage may cause other side reactions that are not desired for cell life. I have no BMS or balancing ( just a fuse for over current) and so far the cells have maintained a good balance, but my loads are very low. My recommendation, charge to 3.40 to 3.55V then balance if you must with no load into or out of the pack. Most of your cell voltage divergence results from the heavier load current and the cell's differing internal resistance and not so much from the generally lower charging current.
Andy, since the smart shunt will re-set the SOC, what are the capacity numbers for your pack with your new charge curve? I bet you are still above 95% of the original 280AH. I really think you are on to some good ideas and thanks for your work.
The balancing sensing depends on particular BMS. Ideally, all cells are snapshotted for voltage at the same instant in time. Problem in real world situation of variable inverter load and variable solar charging is the battery current jumps around continuously. Only way to to get accurate voltage relationship between cells is to get all the cell voltages at same point in time with same battery string current. Cheap BMS's just average the voltage readings. It helps but not super accurate. Most cheap BMS's have fixed balancing trigger voltage of about 3.4 vdc. This makes it easier for cheap BMS because it pretty much means there is no inverter loading reducing how much the cell voltages jump around over time and the 3.4v trigger accentuates the voltage difference for a cell approaching full charge. The higher balance trigger voltage gives less time to balance before highest SOC cell overvoltage shuts down BMS. Balancing becomes a race between lowest cell reaching desired charge before the highest SOC cell triggers the overvoltage shutdown. The higher the balancing current the better the chance to win the race. Waiting until charge current drops off before balancing will not help. Are you going to ensure there is no inverter running? Not very practical. You will likely trip the overvoltage if cells need more balancing. 200 mA balancing is about the limit before you need to have separate cell voltage sensing wiring. On a common sense/balance dump current wiring the balancing current wiring voltage drop on sense wires impacts cell voltage readings. They shutdown balancing to make voltage reading but this reduces balancing time a bit. This is the main issue with cheap higher current active balancers. Charging is not just about voltage limits. The higher the charge current the sooner it reaches absorb voltage, but the longer the current taper off period lasts. The lower the charger current setting the longer it take to get to absorb limit but the shorter the absorb current taper off period. I would not advise using a timed absorb period for Li-Ion batteries unless you keep absorb voltage lower. If you time absorb for 2 hours at an absorb of 3.65v per cell it guarantees overcharging and battery degradation. If you have a BMS that allows you to lower balance trigger voltage it maybe good but depends on the BMS simultaneous cell voltaage sensing. Also lower balancing voltage trigger requires better accuracy of voltage sensing as the cell are closer in voltage without the benefit of the near full charge ramp up in cell voltage exposing the difference in state of charge of cells.
The purpose of balancing is NOT to reduce voltage. The purpose is to equalize the SOC of all cells. Top balancing tries to equalize all cells at 100% SoC. That is why balancing makes sense even when the charger is still pushing current. It's irrelevant to compare the charging current vs. balancing current. You need to compare the imbalance that you want to equalize to the duration of balancing multiplied by the balancing current. So if you have two hours to do the balancing and 200 mA balance current, you can at most equalize an imbalance of 0.4 Ah. The charger current is NOT relevant. Top balancing works on the assumption that voltage is an indication of SoC. This is a problem for LiFePo4. 1 mV imbalance around 3.35 V could mean a 1% or 10% difference in state-of-charge (slightly exagerated for illustration purposes). It's very hard to tell SoC from voltage with LiFePo4 below 3.45 V. It would require very high measurement precision, beyond what a typical BMS is capable of. 1 mV imbalance around 3.50 V means a very small difference in state-of-charge. That's why balancing at a higher voltage > 3.45V makes more sense. Now the question is: how much imbalance are you willing to tolerate in your pack? How much additional imbalance is created in one charge-discharge cycle? The usable pack capacity (Ah) is limited by the cell with the lowest SoC. You want the pack to be balanced to get to use the full capacity.
Maybe the balancing to within 10 percent state of charge is fine for him because he gets more longevity out of the cells. Think about it since he doesn't charge right up to 100 percent really fast his batteries slowly charge to 98-100 percent over a long day. So the time the battery is at 98-100 percent everyday is only a couple hours per day versus a fast charge to 100 percent would mean the battery is maintained at 100 percent for 5-6 hours every day. The time at high state of charge affects the life of the cells.
@@uhjyuff2095 I agree with you that charging to 100% SOC and keeping it there for a few hours every day does not really make sense. I believe it would be perfectly acceptable to charge to balancing voltage only once per month (Victron Energy actually recommends this for their LiFePo4 batteries). Or you could keep an eye on the amount of charge (Ah) that you put in and extract back in a cycle. If this drops noticeably, one possible explanation would be that the cells are drifting due to cycling and it's time to re-do the top balancing. But with a 200 mA balancing current, this is going to take a loooong time because the imbalance is high. Probably better to use a power supply to charge all cells individually to 3.50 V. 1. The available charge (Ah) of the pack is limited by the available Ah in the lowest charged cell. You balance to allow all cells to get to their maximum charge (100% SOC) so you can maximize the usable capacity of the battery. 2. If the pack is badly balanced, you risk that during a normal cycle, one cell gets to the BMS disconnect voltage (min or max) and the BMS cuts off the battery and you are left in the dark.
From what I've learned, the main purpose of cell balancing is to achieve a "collective higher battery voltage charge" thus increasing your overall battery capacity. Either you're going to bottom balance, or top balance; or, you're not going to balance at all. If you're not balancing for the sake of capacity, and you're only operating the cells in a voltage range that is safe, then why balance them. Simply set your battery low and high cut off voltages to the state of your unbalanced cells. Don't waste your money on a balancer if you're not going to use it for what it's designed for. Thank you for your educational videos. They are very helpful.
@Off-Grid Garage Andy, you rock! Your investigations are intriguing and very interesting even for people that are very savvy with solar. Keep up the excellent work. Great vids too.
Second post, yeah, you went down the rabbit hole I was, but you took the time to actually test and present. Thanks, this really helps my thought process. I agree. The pack that stays together, is a happy pack. I see the variances at the top/low ends. Why go there? Just build the pack bigger to compensate for the lost capacity, which is minimal.
So many good points. I've set mine (a Daly with only 30mA) to start balancing at 3.0v. The cumulative effect will add up to balance the cells nicely. 6 months in and it's been behaving really well. I should add, that the Daly only balances during the charging stage. And to detect that, it needs to be >1A in my setup.
Always a good amount of things to think about. I am running a 4S11P 100AH GBS kit. Grade B cells. Every few weeks I gather another pile of cells that I am able to P into more so I might be around 30P by now. I have chosen to limit my top voltage delivered by the SCC to 3.4 and do run a overkill 120BMS and watch the P voltage carefully. By limiting this top voltage I have never had the BMS try to balance cells…and I am happy with that because there is no way it would even make a dent in it with such a low balance rate. I only push 500W in and only run a simple freezer unit so the pack voltage is always less than 3.4 and when on cloudy days it may drop to 3.0 volts I have a transfer switch set to throw it back onto grid. Andy and Will P have filled in my every decision. Thank You.
I love these type of videos. We are learning a lot with the experimentation you are doing. Do you think you are getting close to saying something like; "ok, for a stationary setting if you want your batteries to last for a very long time here is the results of my experimentation. Set absorption voltage at XXXXX, set float voltage at XXXX and set float time at XXXX.? If you want to charge your batteries faster for like a mobile situation the settings I recommend are XXXXXXXXXXXXXXXXX with the knowledge that battery degradation will likely occur at a faster rate with these settings. Are we almost there?
Good stuff!!! Love getting into the nitty gritty, the core, the essence, substance, the brass tacks, nuts and bolts, the foundation, the meat & potatoes, da whole story. Sorry, I’ve been drinking. Anyway, keep up the good stuff.
I think you are right. The problem for longevity is lithium plating. This happens when the highest cell is pushed too high and current continues. Victron is risking that more than your settings especially if the BMS used has a low balance current. If I am correct balancing is not critical if you stay away from charging near full SOC . So victron is risking plating just to solve the problem of plating. You need to reset your shunt as infrequently as possible and balance at as low a voltage as possible even if it takes longer. Marine may be a more critical environment, they may need to compromise longevity for a little more capacity.
So after watching this video and understanding lifepo4 cells the way I do. Working and building battery systems experience has taught me a lot . You can't balance cells that are not fully charged meaning less than 100% soc . They never reach the high voltage knee .fact 1. Fact 2 you also cannot balance cells that are at a float of 3.325v . Why ? A cell will only show it's mismatch in capacity (i.e. requiring balance ) when it has a positive voltage applied - which is directly proportional to it's soc level . Applying voltage / current to a empty cell will never push it over in voltage because it hasn't reached its saturation point decreasing internal resistance . So here is a test charge all cells at 30amps or 60amps and watch their voltage sure you might see a few MV difference In a sort of balanced pack .take note of the voltages and keep track of the mismatch at a medium low soc . Allow the pack to charge up till close to full around 90% you might still only see a minimal difference . At around 95% soc you start to notice larger variances of voltage between cells that are closer to full and cells that are not so full .. And this is where balancing starts to work because you have the positive voltage differential allowing this process to show which cells actually require balancing . So ... To say you could balance cells at float .. after charge can be done yes but will it work ? No ... You need voltage to trigger the knees in all full cells allowing them to be balanced . Lifepo4 cells undercover agents they don't give information unless forced (charged ) to do so .
Cell deviation voltages are soc related and current related . Your target voltage is the first step sure you don't have to go to 3.55v aim lower but not below 3.4v .. on massive large cells like the ones you have you never going to heat the chemistry to degrade the electrolytes because these cells can take 280amps of charge .. recommended charge for cell life span is 0.5c which is 140amps of current . If I see correctly you still less than 0.25c which is less than 70amps . You never going to create heat internally in the cell at this charge rate even closer to the knee aiming for 3.5v per cell so don't sweat ..this factor is only important when charging small cells with a higher chargr current
@@adamsaiyad3959 But yet, he has only a 5 mv cell difference across the entire pack. WHY should he care, that the voltage difference might be higher with a higher voltage, given he will *NEVER* use a higher voltage?
@@MrSummitville valid comment . 5mv at what soc of his entire pack ? Is this 5mv at 60% or 99% soc . You bring up another interesting topic . How much energy capacity is created by unbalanced cells . This term is loose as unbalanced can be a full cell and a cell that is at 50% soc . Then I digress abit further and into more detail. If charging up 290ah cells . And of a group of 16 you have say over 80% of the pack at 3.48v and the remaining 20% at 3.9v at the end of a charge with no balancing how much capacity would the say pack have lost . Probably around 10ah at most maybe 20ah Over all on a pack that is this huge it's a minimal capacity loss for this unbalance . Then brings the next topic if the cells are of the same batch . And are closely matched from the supplier on IR and capacity. And are top balanced technically the pack should go out of balance . A top balanced once and never again balanced pack should remain the same through out the life span of the cells . Same for a bottom balanced pack .. So is balancing necessary debatable . Good quality cells require very little balancing on each cycle and a small balancing Bms can handle this easily . But on the other hand with large cells require a larger balancing current rule of thumb is again a misnomer . Because balancing can still take place as rightly mentioned in the video at a low enough chargr current towards the end of the charge profile in the CV stage.
@@MrSummitville ps as I mentioned in my first right up wanting to balance out a pack at a low voltage will never yeild correct balancing results . In essence might not balance at all . You can only truly balance when you are in the full cell knee and IR lowers because of soc where the propensity of the cell to raise in voltage requires very little current to raise above ocv. Open cell voltage when full . I didn't say he has to charge to 3.55v . I did say that he should use a value above 3.4v to achieve the positive voltage difference to allow current to flow into the cells in a ample amount of time. Charging at 3.30v will charge a cell . But it might take 20 years to full . Charging at a greater voltage differential allows for more current lowering time taken hence speeding up the charge process. We not applying 3.65v per cell as per manufacturer specifications . We applying by far lower voltages taking longer to charge but still getting the knee activated and allowing some form of correct balancing if required and needed by the user.
Hey Andy, I love that you are a testoholic like me! With your new settings, what is the highest mV deviation you see on the next charge up to 3.45V per cell? Myself, I've been charging to 3.45 with a long absorption time and balance it at 3.45. I can see with a system as large as yours is, keeping it in the flat of the curve would be easy to do.
As far as I know, LiFePO4 batteries do not degrade more if held at 3.65V or 3.55V continuously for several years versus 3.45V or 3.35V for the same number of years. High voltage induced degredation is caused by electrochemical side reactions resulting in permanent loss of electrolyte as it gets consumed in those side reactions and deposits insoluble breakdown products over the active electrode surfaces. However, these side reactions start around 4.0V and reach excessive rates by around 4.3V. Li-Ion and LiFePO4 cells use the same organic electrolyte, so theoretically you can safely and continuously use a float voltage of almost 4.0V on your LiFePO4 cells without any harm. The reason LiFePO4 cells aren't rated above 3.65V is because they store no extra energy above this, meanwhile their essentially infinitely steep Voltage/SOC curve above this allows individual cells to skyrocket over 4V extremely fast when charged in a series connected battery pack and still under the normal full pack charge voltage. I.e. high voltage is bad for Li-Ion lifespan, but merely difficult to control and of no benefit in LiFePO4 batteries.
LoganV, Affirmative. Every thing starts with theory and then into practical. Where as, in YT, mostly what you see is the other way around ie: even trial and error method., which’s dangerous and wrong. The Author; here, has done tremendous amount of painstaking research to get to the bottom of some intricacies and has been successful, in unfolding issues. True, for people like me, it takes a wee bit of time to comprehend the essence of what is being lectured. But eventually gets clear. Thanks to the Author.
Hi Andy. Nice video. I want to add another "option" for you. I think you cant balance your cells at 3.35v. 1mV could be a lot of SOC %. I think it is better to make balance at higher voltage. Probably with 3.4v is enough to appreciate SOC % differences by measuring voltage. (Please, now wait to read all text. I dont want to ecualize cells, i want to use those settings like other way to do top balance) What about using the ecualization utility of your mppt ? I mean, you can go higher once a month, for example. Let's go up 3.5v once a month with a ecualization current limit (perfect to balance cells as it will be limiting the maximum current that you want) and a controlled ecualization time. It will be our configurable top balance "voltage, time and current" mode. Please check your mppt settings. It makes me sense for me, dont it ? Charge your cells daily up to 3.35 or 3.40, but only once a month (or whatever) go up to 3.5 in which you will have more possibility or it will be easier to balance cells. Sorry, this option may top balance our heads also 🤯
Thank you Fernando. Using the Equalisation settings for a routine maintenance... hmmm... that's actually not a bad idea. I'll look into it. Thanks for your suggestion, that is great.
I think that the objective of balancing is you have the same state os charge in every cell, and you only can get that in high voltage, 3.4 or above. If you have set it lower the diference will not be noticeable and when you pack goes full, you migh have cells that are much lower or much higher then the other, so not in the same state of charge. This will afect you on low voltage as well. You can have a diference that your pack may trigger the bms in some point. This as happened to me, i had one cell so high it would trigger the bms and the pack voltage was 55volts. My settings is bulk to 56 volts and i float at 55. I cant use absorb so i use the balancing starting at 54.9. My bms balances at 2amps so it is very efective. I had to do a new top balance and now they are similar and the bms is only balancing if it detect 0.003 diference and above 54.9. This is more efective i think for balancing.
@@MrSummitville yes. If he charges to 3.6 he will see that the diference will be much bigger than that. Andy, thats a test you could do. Try to charge to 3.6 ou 3.65 to see what is you real delta between cells. For what i notice in my system the bms will trip because the state of charge can be much diferente and more noticeble at higher voltage. My pack is not perfectly top balanced and i only notice the lack of balance at 3.55 or more voltage. At 3.6 is perfectly noticeble that some cell are more high than other. I get 0.01v at 3.62, and at 3.4 i get 0.003v delta max.
@@MiguelSilva-mb6mb But WHY do we care that their might be voltage difference, at a higher voltage? He is never using that higher voltage. So, how does that impact operation at lower voltages? After three months, it has had absolutely no impact on his operation ...
Me, I like your conservative settings, they mostly make good sense. I dislike drain-style balancing though, but for your purposes it's adequate. The one thing I've learned from you that I didn't learn on my own is that you can float them to full capacity through absorption at lower voltages, I thought the voltage to state of charge curve was fixed, but actually, it's only once the cells are close to full that the charger can overwhelm the ability to absorb energy and send the voltage up higher and that's actually a side effect and thus *any* higher voltage over static "full" is sufficient to produce near full capacity. That's really interesting and really useful. Thanks :-)
I made a video. I'm blushing. :D I experimented with balancing at 3.35v and it does *usually* work. However in my experiments I found that at 3.35v it would often pull current from "strong" cells rather than weak ones which can actually worsen the overall imbalance of the pack. Why does it pull current from "strong" cells instead of "weak" cells? In my experiments I noticed that weak cells would hit 3.5-3.6v first, but then when the charger would shut off or switch to float mode (Or if my shed load turned on) the weak cells voltage would drop faster than the stronger cells would. So during "absorption" your weak cell might be at 3.55 and your "strong" cell might be at 3.45v. Now you switch to float and your weak cell quickly drops to 3.33v while your strong cell sits at 3.37v. Now your BMS starts to suck power away from your stronger cell further worsening the imbalance between these two cells. I can already hear you asking: Why would the weak cell go from being higher voltage than the strong cell to being below the voltage of the strong cell so quickly? Two reasons: 1) it has a lower overall capacity, so it's voltage will drop a bit faster and 2) Weaker cells generally have a higher internal resistance so even if they have identical capacities their voltage will be lower because more of their capacity is lost to this internal resistance. Also: Andy already mentioned this but because of how flat the curve is at this voltage it's really hard for the BMS to tell which cells are weak and which are strong. You have to push the cell a bit into the knee to see which cells "stand out" from the pack. I found that I couldn't trust the BMS to balance at anything less than 3.4v. I tried 3.37 and 3.38 as well. Anything under 3.4 and I found the BMS balancing strong cells way too often for my liking. This not only wastes energy (passive balancing just burns the excess power as heat) but it can also worsen imbalance which defeats the purpose of balancing altogether. So I just try to minimize the time the battery spends above 3.4v for balancing. My shed load technique has worked well for this: Anytime a cell gets above 3.4v it turns on a load. If there's lots of sunlight and the solar panels produce more than that load can pull away. (IE: My 3.2 kW array will first turn on one 1,000 watt A/C unit and if there's more than 1,000 watts of sunlight and it continues pushing the battery voltage above 3.4 it'll eventually turn on a 2nd unit for a total of 2,000 watts of loads). This gives the battery just enough time during peak sunlight periods to let the cells get above 3.4, balance a bit, and then drop back down to levels where degradation is not an issue. So basically my cells only get balanced on especially sunny days (And even then only for a few minutes at a time) which helps me minimize the time they spend above 3.4v. Over time this has been enough to maintain decent balance of my pack. I also have active balancers which help to minimize this time as well because I have a couple cells that I have replaced over the years and those new cells are obviously much stronger than the rest of the pack which make maintaining balance a bit more challenging.
Sorry Andy, I don't quite agree here. There are many ways to implement the balance process. Gently balancing while charging is like gently steering while you are driving on the freeway (and it works regardless of the speed you are driving), plus it is better for the cells since it slightly reduces their peak voltage by subtracting SOC earlier. Small nudges make a difference in the state of charge, as small steering adjustments change the vehicle position within the driving lane. A good battery pack only needs small nudges to keep balanced. If a pack needs more balancing it won't be balanced at the end, but that's a problem with the pack. Whether the balancer burns 200mA for say 1 hour while the battery is charging or afterward either way it has drained the same 0.2Ah and the same SOC percentage from the cell. The earlier it drains this the lower the peak voltage the cell experiences and the better it is for the life of that cell. Either technique works to balance the pack. There are many ways to balance. Gently balancing is valid as soon as the imbalance can be correctly detected, and the optimal time to start balancing is as soon as a valid detection of imbalance can be made.
Thanks for your comment Alan. This balancing method you explaining here (charge balance) does not work with LiFePO4. You cannot balance these cells at a lower voltage. It makes no sense and causes imbalance. Once the voltage rises finally, the BMS has no time to counter this action any more and the battery will never get balanced ever. These balance methods may work with li-ion as these cells drift quite early but LFP, no chance. Balancing only works after the voltage knee which is around 3.4/3.45V.
@@OffGridGarageAustralia Hi Andy, I'm sorry that you misinterpreted my posting. You were saying lower voltage, I neither said that nor intended that meaning. Balancing can start at any voltage where imbalance can be CORRECTLY DETECTED. This voltage must be above the flat part of the charge curve as no proper determination of balance can be made until the voltage begins to rise at end of charge (note that this voltage is a function of temperature, so picking a fixed value that is too low may cause problems when the seasonal temperature shifts move the system voltages upward, the default values are chosen for a wide range of temperatures). The most important point here is that there is no need to wait until the charge current is zero. That is not important. As soon as imbalance is correctly detected the balancing operation may start. It does not matter that only a small current is subtracted from a larger charging current. The SOC is being steered, and that is the way balancing works. The SOC must be steered so the cells match in voltage. The balancing is adjusting that SOC so the voltages match. Gently steering the SOC is totally valid, as long as the information is there to do it correctly. Doing it at too low a voltage fails to work because the correct information as to what cell's SOC to adjust is not available. The cell voltages don't reveal the SOC during the flat part of the charge/discharge cycles. Thanks.
andy, this video i did like :) also i think,i like your ideas with the voltage.. but i think it would be a good idea to test this a bit over a longer timeframe. i will however in my mind never harm the pack or cells
Very interesting video! I am a newcomer to lifepo4. But I finde, that for my 24v system in UPS usage scenario without PV is good to use charge bulk 27,8v(3,475v) 45A without float and without absorbtion- dont reach high SOC for healthy. I have diff voltage usual at 0.005v at 3,32v per cell in rest. But, in my oppinion, due to the smollest difference in tempreture and internal resistence (but I mesured all the same 0.18mOm on each 8 cells 202Ah Lishen) under load and during chage, we get increasing unbalance day by day, that is not so visible at flat line 3v- 3,3v. Lower 3v and higer 3,55 diff goes up to 0.200v under load or on charge to 3,55-3,6v! But for more acurate balancing i my Must pv3024plus 24v I have option "equalization" and nobody doesnt explane me, that it is not only for AGM or Lead battarys and also good for LiFePO4 battary. In "equalization" mode it is the same charge curve that Andy shows in this video! In mode "eqialization", once for a month or once for a weak, whatever... I can set voltage higher, than my usual charge bulk 27,8v(3,475v) to 29v ( 3,625)v, for example. In this case voltage goes up in 2 steps: to 28v CC/CV 1st and than very slow to 29v with 10A to 1A increasing current and goes in flat line at 50mA carrent for equialization - some absorbtion ,for exemple for 60 minutes and Neey balancer works hard for eqialize, balance cells and charge current goes down to almost to 0A. After that time charge stops and voltage goes to prefered float voltage or fully disconect charge.
Once top balanced, you should never have to balance your cells for months (depending on your use). They will NEVER be balanced perfectly and that is okay.
Fundamentally with any real world and thus non identical cells your latter statement is a bit off. It's not that they will never be balanced perfectly, it's that they cannot be balanced except by perfect manufacturing practice. All we can do, as end users, is top balance, bottom balance, or dynamically try for both IF the cells are close enough together AND for some reason you have to push both ends of the envelope AND your active balancer can pull enough current to avoid an over voltage scenario for individual cells/banks.
Even a year, or more. LiFePO4 cells tend to stay in balance for a very long time, as long as you do not fully discharge the pack, which is why the relatively low balancing current the BMS is able to provide is plenty. -Matt
Andy I wish to respectfully disagree a little on balance curent. Yes the balancing cell is not discharging but it also means all the other cells are charging more. Therefore they are getting more coloumbs. The question becomes is the difference enough over the time it takes to go from balancing to overvoltage of the highest charged cell. Really we need a integrated BMS/charger that can taper current based on individual cell voltage.
If the cell with higher voltage is charged less and the other ones are beeing charged more, that's actually good and part of the balancing. I just don't believe the BMS is capable of handling these high capacity cells. The 0.2A balance current is just not enough to make a difference in the time available to it. A charger which can handle all cells individually would be amazing. I know Peter here is working on somethign like this.
Well since you are using a discharge BMS balancer then what you say works. My BMS has active balance so my balancing takes place while floating but also when charging as long as the charge is below 10 amps.
Why has everyone on youtube started teaching to use 100% of the batteries capacity? What happened to leaving the top and bottom 15%? We know you get a higher cycle life. Good work.
Everyone? Nobody here, says that. Many do say, "... Use 100% of the capacity ..." So, that they can justify the high cost. But then in the real world they end up using only about 70% of the capacity ...
062521/0540h PST 🇺🇸 @ Paul>>> It’s a universal truth that the said country has always been a forerunner to almost many inventions that we see today. Many may not acknowledge that truth, though. Thank you for your sentiments, as I too acknowledge the truth. Stay safe and 73s…
Nice Andy you’re proving the theory with data real data not only with manufactured specifications but with real data that’s wonderful I which I have all the test equipment you have to do the same test years before it is making totally sense all your data prove the real stuff I’m 100% with you thanks for all the help and information I really enjoyed your channel 💪🏼👍🏻💪🏼👍🏻
The basic answer is no. For LiFePO4 you can only properly balance cells when the voltage is modestly higher than the cell's resting voltage. A Float of 3.375V is not high enough. Sure, you can try to balance at a float voltage of 3.375V, but equalizing the cells at that voltage won't actually balance them. At all. Why? Because the balance current would be too low and the voltage has too much freedom to move around with very little current when its that low. I would go as far as to say that balancing is basically impossible at 3.35V or 3.375V. So basically, balancing is only effective at the Absorption voltage (that is, a voltage modestly higher than the cell's resting voltage). In terms of balancing while in bulk... actually this can be done to some degree. If the BMS is smart enough and stores the cell history, it can figure out what kind of balancing is typically required during Absorption and it can try to apply it during bulk. It would basically be an educated guess, but once it finally gets to absorption it would then have less work to do to clean-up the cells with a final balancing. I don't know any BMSs which do this though. -- Battery degradation is more a matter of the SOC, *NOT* the bulk target voltage. Remember that the actual break-down voltage for LiFePO4 is 4.2V. Other lithium chemistries have less margin, but LiFePO4 has a huge margin. Bulking to 3.55V isn't going to hurt the battery. -- Remember that the proper target voltage for the charge controller depends heavily on the charge rate. If charging a battery slowly, lower target voltage is necessary to avoid over-charging. If charging a battery quickly, a higher target voltage is necessary to avoid under-charging. There is no 'right' answer... the answer is 'it depends on the average charge rate'. With the amount of solar you have now, you are probably charging at a high-enough rate to warrant a 3.50V or 3.55V bulk target. And I'll give you another reminder that with the number of strings you now have, blocking diodes are now mandatory. Look into getting a proper solar combiner box. -Matt
You said, " ... Because the balance current would be too low ... ". Oh, really? At 3.351 Volts the Balance current would be ~90% of the balance current at 3.6 Volts. So, how is that, "too low" ? In the real world, the Balance Current is not too low ...
Interesting concept have my 4s EVE pack charged to 3.55 and balancing starting at 3.4 but i use a 2A active balancer with the REC ABMS. i do look at your comment of does this really matter? these cells last so long that even if you cane them and reduce there lifespan by a few years they still last 15 odd years and by then there will be something with a much larger energy density
I have the same setup with REC and EVE cells but have balancing set to start at 3.35. Have noticed the BMS actively controls the charge parameters and does most of its balancing around the 13.8v point on the way down from 14.2v after the some of the cells hit top voltage at 3.55v, do you find similar?
062521/0549h PST 🇺🇸 @ BoatingTube … what is your BMS and the Charge Controller make? These two devices play a major role in setting up charge/volt parameters. FYI, even if your BMS (?) has default value and is not configurable, you will be able to increase and or decrease charging current/volts per your choice through your Charge Controller. [ Assuming your CC is VICTRON ] There’s no need for AB if your CC and BMS are from reputable manufacturer. 73s…
Before even watching I maintain you can only properly balance at the rapidly changing bits of the graphs you have been generating. ie at the very high or low voltages. And only properly check the balance at these extremes too.... now I unpause and watch :)
and maintain it still. However I only think you need to worry about balancing once a year. I DARE you to turn OFF the balancing for 3 months - then do a higher voltage/cell check and see how little it has gone. I don't think your balancing is working, but because your pack is already balanced, it doesn't have to do anything - so compare to 3 months with no balancing.
@@KevIsOffGrid You are right the balancer may not do anything because the battery is perfectly balanced, but it might change in the future so having an automatic balancer is the way to go. He doesn't want to make balancing batteries his full time job. Just turn it on and let it do the work for you!
@@uhjyuff2095 I have a similar pack, 24 cells. I check the balance maybe twice a year (no balancer) its in balance always. We use new cells - no balance issue. We cycle at low C rates - this isn't an EV with massive current draw, Andy's system 1C is 280Amps, all summer he had 1 charge controller at 35Amps - so couldn't even get to 0.2C charging, he could just about discharge at 0.2C if the 3000w inverter was flat out on the tesla charging... Balancing isn't a full time job, it isn't even a part time job, its just a non-issue at all. Just use the battery and that's it, Solar is so gentle on large battery banks :)
@@stupidmonkeykev Well check this out for an example. He starts charging his tesla and the low voltage cutoff on his inverter is set too low and then the BMS has to shutdown the discharge. The low cell is what caused the bms to cut off. Then he does this once a week. So one cell is getting slightly degraded more than the others due to lower voltage discharge every time and then the cells slowly start drifting out of balance more and more year after year. Having the balancer on for long periods because the packs are so large and the balancer is so slow is just a great feature to have and something that just works without having to think about it. Obviously, this is a worst case scenario but most people have no idea about these lithium batteries and how they work.
@@uhjyuff2095 - In what video did you see ... one cell is getting more degraded than than the other cells and then the cells drifted out of balance? We see no evidence of that happening ...
I don't kehr - Schornsteinfeger verweigert Arbeit... I think the same. As long as you monitor and are able to recognize problems, you can do whatever you like. And now let the sun shine ;-)
Wouldnt it be Smart if the BMS tells the Charger or VRM oh today my cell deviation is high today i wanna charge to a high Voltage? And on other days you dont charge to a high Voltage to not stress your cells?
You can do whatever you want to with your own batteries through your batteries I say that's fine I do something similar as long as you don't randomly want to go to a higher voltage and expect one cell not to jump away and go higher than the rest there's no problem at all doing that there's even balance boards that just continuously balance all the time and then there's some that just turn on a resistor when the voltage is above say 3.6 or 3.65 volt and the resistor just stays on forever as long as the voltage is high enough
Great topic and video. This has been an area on my mind for awhile. The only catch I see, is many BMS's are hard set to start balancing at 3.55v. The best I can figure, it so set your bulk charge voltage to 56v (3.5v), thus, when that first or second cell hits 3.55 it will start balancing off the the high voltage to the pack. With the voltage being near full, the risk of over voltage should be minimized. Thoughts?
Im probably thinking of this all wrong, but if you want 'balancing' during the float period, would it not function better to set the balancing at 3.349v instead of the 3.351v that was selected? Im presumably misunderstanding the situation, but surely if the float voltage is being maintained at 3.35v/cell(equivalent), then the balancing function will not be active if set at 3.351v (for most of the time, since loads are being fed from SCC)??? Furthermore, I would have thought, that at 3.349v/cell the highest cell would constantly be getting 'bled off' , all whilst the SCC attempts to maintain the float voltage setting point of 3.35v. In my crazy mind, this seems to me that the balancing would get done very gently using very small currents(not bouncing all over the place) , over a longer period of time, producing more stability???
I remember you use to say keep your state of charge at about 80-90% to keep your battery from degrading. Is this still true? Or can you keep your battery at 100% state of charge just at a lower voltage? Thanks for another great video Andy.
@@MrSummitville Most people won't need to go above 100% SOC. I want to know if it degrades your LifePO4 battery if you have it at 100% SOC at lower voltage.
@@MrSummitville Your question to me is "Why do you need ot go above 90% SOC ?" My answer is, I don't need to. I want to know what Andy's thoughts are currently about being at 100% SOC at a lower voltage.
11:42 what you say about where your balancing occurs is not correct, it actually starts on the up slope to the left of that - it's not important that it doesn't drop the voltage of the cell with higher voltage, it *IS* pushing more energy into the others relative to that one, and is therefore charging them more quickly and working towards an equilibrium. The trouble is that at those higher voltages your HV cells (low capacity cells) can be getting damaged while your higher capacity cells are just charging normally. That's why I much prefer full time active capacitor balance boards. They can shift more current and they can run full time with low draw, so as soon as the full cell voltage starts to rise they kick in immediately and start to shunt current into the bigger cells. This matters on the top and the bottom equally. But if you do have a genuinely different sized cell and you do reach the bottom AND the top then the chances are this will not be enough to avoid an over voltage scenario. Hence it's better to stay in the constant-voltate part of the curve on the lower side and only push the top end on the high side as gently as possible to generate the cell difference required to activate current flow between cells. Once top balanced you should see very little cell drift assuming no self discharge differences riding on top of the capacity differences.
There is ample evidence and a lot of papers showing that STATE-OF-CHARGE is the stressing factor (along with temperature and charge / discharge rate). Charging to 100% IS stressing the cells. Pushing and pushing and pushing the lithium ions into the graphite (aka INTERCALATION) is what stresses the cells during charging. There is almost no evidence that voltage is the stressing factor below 4.2 V (yes, even for LiFePo4). That's the voltage where some chemical reactions start to happen that degrade the electrolyte (if I remember correctly). Your cells SEEM to stay in balance at mid-voltages...but 4 mV imbalance could be 4% state-of-charge imbalance. You just don't see it. If you have one cell at 100% and another at 96%, you will only be able to discharge 96% of the battery capacity because the lower-charged cell will hit the bottom sooner.
I agree, not the voltage that degrades the cells it's the state of charge that does it. Lots of studies showing it is SOC not just the voltage. His balancer is just not smart enough to balance really accurate, maybe in the future balancers will be more accurate, but maybe not because it comes down to the science of these batteries. NMC seems to be ez pz to balance but LifePo4 is more complicated.
@Daniel Ardelian - I agree that time spent 100% SOC does reduce lifetime capacity. But I ask you ... Why do most manufactures specifically state .. Do not Float Charge at 4.2 volts per cell? Because ... 4.2 Volts is significantly above 100% SOC ...
At 53.2 you're only at about 80% soc so of course the cells are close together in voltage. It's when you get to the extremes (fully charged or near empty) that you will see an out of balance condition.
If there is imbalance then the cells are always out of balance, you just don't see it due to the flat voltage-to-SOC characteristic of LiFePo4 at mid-voltages. You don't see it but it's there...one cell has more energy than another...Then when you discharge, the cell with lower SoC will hit the lower voltage limit first and this will limit the usable capacity of the whole bank.
Why can't you use zener diodes to balance the cells? They come in 3.3V . Though it really seems an ideal solution to have the right voltage of solar cells to charge each cell individually. 6x 0.55V cells in series, however many lots of six in parallel as you like.
You cannot balance any cells at 3.3V, that is far too low. You want to be in the part of the charge curve, where the voltage is rising, so 3.45V for LiFePO4 and around 4V for Li-ion. And this may be a different point for different batteries and also may have to be adjusted once the cells get older.
@@OffGridGarageAustralia yeah and I guess you'd need another diode, perhaps an LED in series with the Zener but opposite polarity. Which could increase the charging bypass voltage to 3.5, and stop the cell discharging through the Zener. I'm no electronics whiz, but that seems to be the simple standard voltage regulator setup. If an LED was used then you would get a nice indicator light telling you the cell is at voltage, when the Zener breaks down at the cutoff limit. Just a thought. Living on a boat and wanting stuff that can handle marine environments and be easily diagnosed and fixed from spares you have on hand. I mean, its life or death if you can't keep vital systems running at sea. I've got bought digital voltage readouts for each of my six cells. I think I can do alright piloting the vessel on electric propulsion by spade switching neighbor cells in a series of six from series to parallel for throttle and top and bottom balancing purposes while the solar and motor are connected to the bank. I would be nice to have a backup for the voltage readouts incase they should malfunction. Though switching the cells to parallel would diagnose that quickly. Complicated electronics at sea is nightmare city.
Some BMS is out there will not allow the balancing function if the battery is not charging I had one like that I would rather prefer it to only balance above a certain pack voltage and above a set difference between cells but allow that anytime With my current setup I use my victron shunt relay to turn on a separate balancing board when it reaches 100% state of charge and then I have set 1 hour absorption at 3.5 volts on mine
I have a 8s 24v EVE system. After doing my first balance at 3.55v (No need to go higher IMHO) my cells stay within 10/15mv from 3v to 3.4. I exclude both ends of the charge/discharge curve as usable because it is really negligible. If I were to need that extra juice, it means I did not size my system properly.
Hey Andy just catching up with you.. not sure if I can go that low effectively. Probably need more panels at this point. Certainly is good for longevity though at 3.35v...
Try not to balance you cells at all. You will be surprised how little will change to what you are doing now. You can take 3 years vacation, and when you come back, your cells will be still where they are today. I didn't absorb, float or whatever for almost a year with my pack now. My BMS would need a trigger voltage of 3,55V to burn 30mA away. None of my cells ever saw that voltage. The only time I had to remove a bit of energy from one cell group was, because it was a bit ahead from day one already. LiFePo4 and balancing, is the most overexaggerated topic ever :)
Andy, It looks like you got your system fine tuned or at least making friends with it. Unfortunately, many of us don’t have the flexibility that your system has or it’s for a totally different use, but we all are fighting the same gremlins. I was doing some tests on a quick turn around charge for 12v 280ah battery for the trolling motor in an aluminum fishing boat. Use all day, charge nightly at home and be ready early the next day. Not being able to control the absorption time (dumb charger) and not having the luxury of time, it gets charged at hard at 50 amps. The BMS simply doesn’t have the time to work the imbalance with the feeble 50 milliamp balance current. I started seeing too much imbalance and the Overkill BMS would end the charge because of a high cell and with it the balancing. Things would only get worse each cycle. I found the Heltec capacitive active equalizer/balancer. I use the 3S 4S 5A model for my 12 volt project and the 13S - 17S 5A model for my 48v systems. They are efficient and haven’t noticed any heat because it doesn’t waste energy through resistors but channels voltage from the higher cells to the lower ones. Standby current is negligible but it can be put into hibernation by adding a switch. They work together (parallel) to the BMS’s leads at the battery. My delta? Usually within .004 Volts 30 minutes after my charger finishes at 14.5 V and never varies more than .012 V during the charge. Right now it’s fluctuating between .001 and .002 delta at 70% charge and the equalizer has been off since it’s last been used fishing a few weeks ago. If it’s not going to be used right away we don’t charge until the day before since they don’t like being stored at full charges. The thing is a beast close to 4 times the useable power and 10lbs lighter than the 100ah AGM it replaced. More running voltage to the trolling motor = faster, far longer and may outlast the boat lol. Cheers
Sounds like you have a handle on it, so to speak lol. What if you had two batteries that way you could give your batteries the time if needed and then switching your batteries every other day instead. More cost of course, anyway just a thought.
@@MrSummitville No, try charging any group of top balanced, high ah (270-320) grade A, impedance matched cells with a 1/4 C rate and virtually no absorption/ balance time with the most common BMS’s (only .050 milliamp balance current). Remember most BMS’s only balance during the last part of the charge and when the charge is disconnected the balancing is stopped , but necessarily complete. Eventually you’ll end up with an imbalance as you get to full charge. This is a a common issue for the avid fishermen, golf cart, fork lift, applications that get high use but need more than mediocre balancing current. Another thing people don’t realize is temperature difference of cells. The cells in the core vs the ends have different rates of heat gain or loss, either self generated (charge/discharge) or environment. Temperature differences can effect voltage/capacity. Then there’s time. Theses things are new. How matched will these things be 5 years from now? But there’s an easy inexpensive cure or a bit of insurance.
@@SkypowerwithKarl Charging at C/4 is typical for a 280 Ah 48 Volt Off-Grid battery bank with a 3,500 Watt PV array. It is done every day. And if you need a 5 Amp Active Balancer then one of your cells is BAD or failing - ie it does not match the other cells. Also, do not use a balancer that stops balancing when charging stops. In your situation, you created the need, for a High Amp Active Balancer.
@@MrSummitville Hello, balancing after it’s done charging is an active balancer. Most popular BMS’s for internal battery/case use don’t have a large balancer because they can’t get rid of heat (resistance). Capacitive equalizers (balancer) are the future, very low heat. They may even open the doors for B grade non top balanced systems. Keep in mind that a solar system isn’t constantly disconnected and the battery won’t see a charge till it’s depleted. Solar is a constant give and take where it highly likely to have long intervals of absorption available.
I run 2 banks, and allow 2 days for internal balancing (1 day minimum) before use, charged to 3.4v per cell, average. More ammo for your 2 bank solution!
I´m using 18S Strings 3 in Parallel 280Ah so far i have no negative experienced, i can balance at 3.53-3.54 per cell (64V pack voltage), i have used cells so i have to rebalance every time i get to 10% or 100%. My BMS (batrium) cant keep up so i´ll be instaling active balancers soon aswell.
Hey OGG i seem to have a goofy Shunt also. I think the voltages and current readings i get are close but not exact. I have Valance Batteries (2) and i have no clue what each cell reads, I haven't been able to get my laptop and batteries to communicate. Do you think the BMS in those batteries are smart enough to regulate the voltage cells safely and properly on its own? BTW today is the first day my Victron Charge controller stayed on Float all day. Ive never seen that before. Ran my AC and for about an hour and the Float light never left Float mode.
@11:47 - You drew a GREEN line and said, "This is where my balancing *STARTS* " Actually, the balancing started when the Cell Voltage first went *ABOVE* 3.351 Volts, during BULK Charging. How can your balancing "start" so late - when the voltage is actually FALLING below 3.351 volts per your graph? Doesn't your balancing actually terminate ( it does not begin ) when the cell voltage falls below 3.351 volts?
Question that you may have already answered... When you "float" at 53.6 I've noticed on my shunt/monitor that if there is a load, it momentarily uses battery and then the solar charge controller adds power until it is pretty much 0... perhaps 0-4W in either direction... If the load is larger and you remove it, momentarily it seems to charge the battery before backing off/etc. Is that a bad thing? Seems like at one point, the thought was to bulk charge and then essentially set the float low enough that it doesn't re-trigger any charging unless there is a decent drain on the battery. Perhaps on a related note, the unit i'm using is a Growatt unit. I notice that the CC charging up to whatever voltage completes after which it "absorbs" for a few minutes at CV. As soon as that finishes, it goes into FC and floats at my desired float charge. The odd thing (i'm thinking this is strange) is that the battery meter shows that even with no load and inverter off, perhaps up to 200W starts being drained from the battery and tapers off until it reaches the float voltage. So it appears on those, that the unit actively discharges from higher voltage down to the float voltage. (FYI, my 24v system is using 26.8 as float and 29.2 as constant current charge per battery manufacturers recommendation - when i had CC charge to 27.8 they advised me it was too low as their BMS would never balance the battery).
Probably because it would run for too long at the higher voltage and endanger weaker cells during that period. His goal with the 10 minutes was only to trigger his SOC reset to occur, not for charging purposes.
@@fredio54 #1 He is charging at a modest voltage of 55.2 so endangering cells is not an issue. #2 On a sunny day and the charge amperage is high, he will hit the absorption voltage much quicker so the charging will stop prematurely compared to an overcast day which will have a much longer charge time at a lower amperage. If you stop charging a battery at 20 or 30 Amps compared to stopping at 2 to 5 Amps (at the same voltage), you will be at a lot lower state of charge when you're done. Very inconsistent.
@@SpeakerKevin #1 I beg to differ, danger is defined by the following simple math: 55.2 - (3.25 * 15) = 6.45V the voltage one cell will get to if the others have not become saturated and started to exceed nominal voltage. That's fried/toast territory. #2 Not true if he maintains *any* voltage higher than nominal for an extended period as proven on this channel rather conclusively.
@@SpeakerKevin To be perfectly clear, the voltage for a 16s pack you have to charge at to be 100% sure to not overcharge any cells is defined by (3.25 * 16) + 0.4 = 52.4 - any higher is at risk of damage without appropriate balancing techniques and equipment.
@@fredio54 His delta voltage between cells is a mere .005V. Plus he has a fully programmable BMS that will shut down the charging if any cell gets outside of a safe voltage.
Can you help me ?? Hey Andy, there are 2 cells in 48v pack that are too high and shuts off the BMS. One video of yours showed you lowering a cell with light bulbs or something. How did you do that?
I just use some car 12V light bulbs and connect them to the cells which are getting too high to discharge them a bit until all other cells have caught up.
Passive Balancing is slow & for Low Capacity cells. Typically below 2A to 10A for large capacity cells, Transfers from Hi cells to Lo Cells. 3.65V per cell is the "Max Allowable" for LFP. They always settle to 3.55 within an hour and that IS the 100% point in reality. Remember Nominal os 3.200V per cell and full working power-curve is 3.000-3.400. EVE SPECS INFO: 4.2) Standard Charge The standard charge means charging the cell with charge current 0.5CA and constant voltage 3.65V at (25±2)℃, 0.05C cut off. 4.3) Standard Discharge The standard discharge means discharging the cell with a discharge current 0.5CA and cutoff voltage 2.5V at (25±2) ℃. If required, the battery can be discharged at 1.0CA constant current to a cutoff voltage of 2.5V. EVE 280AH Cell: (280 X 0.05C) = 14A HERE are Settings used with a Midnite Classic Solar Charge Controller for LFP Charge Profile. A QNBBM Active Balancer installed on all LFP Packs with the BMS. All equipment MUST BE Voltage Corrected & Calibrated (VERY IMPORTANT) - Divide Values X2 for 12V. Multiply X2 for 48V. Absorb: 28.2 for 15 minutes (3.525vpc) (some call this boost) Equalize: OFF Float 27.9V (3.4875vpc) MIn Volts: 22.0 (2.750vpc) Max Volts: 28.7 (3.5875vpc) Rebulk Voltage: 27.7 (3.4625vpc) End Amps: 14A (*1) This get's the bank charged to full with high amps (Constant Current) and then float (Constant Voltage) tops off so the cells are on average between 3.475-3.500. I am running 7/24/365 so float is used up by the Inverter + provides whatever the packs will take to top off. (*1): End Amps is calculated from the {Highest AH Battery Pack} in a Bank. IE: 200AH X 0.05 = 10A 280AH X 0.05 = 14A. ** Coulumbic Efficiency for LFP is 99%
I charge to 100% , turn off charging until 90% at the Same day and only then recharge. It is not good or healthy to keep the cells above 3.45V for a long time. But i think victron does not allow such a setting Based on SOC
Does your BMS balance on float ? certainly my Daly BMS will only balance while the battery is charging. Having said that I agree with your voltage settings. I believe your BMS software has the option to balance when not charging. It would be interesting to see that in action.
Yes, it will start on the set balancing turn on voltage when the voltage is going UP and will continu as long as the voltage is above that and current is being delivered. I think the approach Andy takes is not working, but he will have to find out himself.
@@HansKeesom The key phrase there is "current being delivered". On float there is very little being delivered, only enough to maintain battery losses, so balance would not work.
Hi there i follow you with great interest in the U.K.At the moment i run a 15kw powerwall made of laptop batteries and very nice to .But i want to go for brand new batteries so i watch your youtube every episode.Can i ask one thing.In the summer in uk(HA HA) my batteries are fully charged by 10 oclock in morning(30amp).MYy question do you bring your batteries up to float conditions so the battery can absorb all day at low amps.I am still trying to get this float thing lithium ion rise to voltage and the amps drop off a cliff the last 10% takes about 1 hour then .1amp forever.Is this the same with lifep04??.I love the flat curve(charge/discharge) my inverter can do 60v so i am planning to run 17s @ 3.45V So the float will be about 9hrs???
Bulk and Absorb charge to 90% SOC and then you can try Float Charging the LiFePO4 battery bank at approx 3.35 Volts per cell or whatever voltage pushes ~0.1 amp into a 100AH cell ...
Everyone should use voltage that work best for them. It is Capacity (amp-hours) vs # of Cycles vs Calendar Aging trade-off fro everybody. Nobody should blindly change their voltage just because this *might* work for Andy. This is *NOT* an off-grid home ...
I think you are onto something here, but I would (for science purposes) run it like you said for 2-3 months, and then charge it to 3,6V per Cell and see how far they drift apart. Just to see what happens.
@@MrSummitville To make sure that the cells dont drift apahrt SOC wise. I mean you would not notice it by the cell voltage in the middle of the charge curve.
Why don't you take Jeremy's advice and only charge to 54.4V (3.4V/cell). You could use a longer Absorption time for balancing, or better yet, set the tail current at 1-2 Amps. This should give you a resting voltage of around 53.9V/3.37V.
I am now confused about the balancing process. I have a Daly BMS and it only shows that it is in balancing mode when the BMS is in charge mode. Once the BMS is switched off because it reached the designated charge voltage, there is No balancing. So I wonder if balancing is not only to reduce individual cell voltages or does it also increase those cell voltages which are too low? Danke
The way I understand BMS is it always lowers the highest cell down to the lower ones voltage via mosfet. It will continue to do that until all of the cells are charged to your set point.
How can the BMS balance your cells with say 30mA when you still charge your cells with 5A at the same time. That does not make any sense... Only an active balance will decrease voltage of cells which are too high and transfer this energy to cells with lower voltage. Normal BMS don't do that, they just burn off the energy through a resistor.
I would imagine that different manufactures configure their balancing act many different ways. I’m also willing to bet that some do a piss poor job at system logic and can only come up with a sub-par product. I would consider that Daly BMS a shitty design if it only balances (or tries to balance) while charging only. It will never achieve its purpose in my view if that is the case. I agree with Andy, this makes no sense. Must be a design flaw.
@@Do_the_Dishes what if it is actively balancing and instead of depleting high voltage cells, it is supplying the lower cells with additional charge current?
@@OffGridGarageAustralia An Active Balancer might be able to reduce the voltage of one cell, if and only if, the Active Balancer can transfer *more* the 5 Amps of charging current, many (most?) do not. And if it is a Switched Capacitor type of Active Balancer and it is rated at "5 Amps", then it will probably only average 2.5 amps transfer between the highest & lowest cell. BUT ... Cell Balancing, whether Active or Passive, should *only* occur during the end of the Absorb Charge or during Float Charge when the charge amps are LOW ...
i dont think scc are that smart from what i have scene the batterys only offers current if the solar is not producing enogth power to run the load so then takes power from the battery untill solar comes into sun again and can then power the load and the charge the battery this will happen with out a charge controller
i myself like active balancers as these will start working any time the battery is 30mv diference and takes power from the higher battery and dump it into the lower battery i feel this is less wastefull on power then dump load balancers
A 30mv difference between adjacent cells or 30 mv across the whole pack? What Mfr & model of A.B. does 30mv across the whole pack? Andy's battery bank only has a 5mv difference across the whole pack, and it only operates near the end of the charge cycle = very limited time.
@@MrSummitville active balancer work all the time no matter what voltage andy's cells are still knew i suspect as they get older they will start to drift more the active work between cells if one is 30mv lower then over this could even be 10mv im just going off the one i have it will start taking power from the higher cells and putting it into the lower charged cell and thay will trigger no matter the point of charge it just sees the set voltage diference and starts to balance
Hi Andi Wie immer super Video Schau dir mal das Baterium BMS an bis 7 A mit Kühlung und 2 A ohne Kühlung und swipe mal durch das Video th-cam.com/video/m2MuEpuMSiw/w-d-xo.html Grüsse von der Schweiz
Danke Dir Daniel! Weiss ich, habe ich mir schon alles angeschaut. Die sind ja gleich hier um die Ecke. Super teuer und keine Sicherheitsfunktionen. Hmmm, komme da immer noch nicht dran...
balancing is/should not discharge a cell that is high, it is/should bypass it partially with 0.2 A. it start that at 2,7 to have enough time to do this. The balancing turn on voltage is when the voltage is going UP, not when going down, so I fail to see how your idea at 12:02 is gonna work. At the end you have failed logic. You moved the location of what you consider 100% charged, but that does not mean you have the same capacity/energy of wh's stored. It's like 95% is the new 100%, hurrah. If you have less drift, that does not mean the balancing is working just as good, it just need to do less work. So that is in the case of this batteries a good reason to go for a lower top voltage. So instead of reducing the absorption time, it would have made more sense to lower the absorption voltage.
So, what is "wrong" with enabling / starting the Cell Balancing process during the Bulk Charge mode, when any cell goes ABOVE the 3.351 Volt set point ? Nothing is wrong, you just "claim" it wrong. In fact, it is good thing to enable / start the Cell Balancing as soon as any cell goes above 3.351 volts in Bulk Mode, to obtain *maximum* Balance Time. So, I do not see WHY you think is better to *DELAY* the start of Cell Balancing, long after the cell voltage has raised *ABOVE* 3.351 volts. This video, your graph & your "logic" make no sense what-so-ever ...
Hard to say what is right and what is wrong. If someone can explain, what does it mean, if several cells run up during charge and go lower at rest or discharge... I think, becose of the slightly difference of internal resistance, and/or capacity of cells during charge or discharge using I=U/R we can explain the deviation during charge higer then 3,45. In my case, I always have 3 cells, which always runs up on 0.020v at 45A and 27,8v ( after 3,4v until 3.55v) and daly bms slightly slow down there run-up. And when charge is on finishing line, voltege increased above 28,2 and current starts to go down my Neey balancer starts the job, that 3 quiqly goes down under 3,45, and other 5 cells that was lower on 0.020v becomes higer on 0.050v and goes up to 3,630v! at high battery voltage and decreasing current. So that 3 cells, which looks higer becomse lower and look like not fully charge... So, if would set start voltage at 3,35v for Neey 4A balancer, it will discharge that 3cells, and in the end the they become not fully charged, and it will look like 5 cells at 3,612 and 3cells are still at 3,437v. I described my real situation. And for balancing I leave it in float at 3,588v for several hours to get deviation at 0.002v and then tern of. Iven ufter that, after rest, that 3 cells looks lower on 0,002v... Maybe nex time I'll do experiment and try to charge to 3,600v all battery and then separately that 3 cells one by one with automatic 3,65v Lifepo4 charger to 3,65 and some absorbsion. Maybe that 3 cells become fully charged. But maybe in this case they could run up much more quiker and bms and balancer would discarge tham, and again.... 😞
Andy, This is great stuff. Keep asking these questions. On another U tube " Emily and Clarks Adventure" Clark has made a hybrid system with a big led acid bank and a smaller LIP companion all hooked together. I wish you would take a look and give your opinion. The idea is the LIP does most of the work each day and the LA stays at or close to float and only discharges in a big demand event like extended low charge days. This greatly reduces the cycles on the LA bank and extends their life.
Many boats have a big LA bank so this is a way to get the benefit of LIP at a lower cost and not having to wait till the existing LA bank is dead.
I am a fan of Will Prowse and I must say your channel is getting as good as his with content like this,
Great job Andy! Regards from an Uk froggie 😜
Awesome! Thank you!
Juan, there’s a lot of difference between an Elephant and a Goat. You get my drift? Education makes a lot of difference than narrating from a script.
One good reason to charge at a lower part of the top end knee is that your cells will reach full at different rates and you have some headroom for not cooking the weakest cells every time you fill them up if you stop early and let things equalise.
What does it mean - the weaker cell? Why it is weaker? Why it goes up during charge and goes down lower when discharge? Internal resistance- bigger or lower? Capacity, bigger or lower? Looks like weaker cells runs up, bms/balancers do them lower, and thay become not at the same SOC like others. Or if we use float, at 3,4v weak cells can absorb...
@@User1462uuw8w less capacity - eg take a 3 year old worn 8Ah cell and it will likely be 7.5Ah or so, put it in series with 3 brand new cells, discharge each to 2.5 individually then put 7.6Ah into the string, the old cell is full and at 3.65V and the other 3 are still needing more and at 3.2-3.3V on their way up. Continuing to charge without a balance solution overcharges and over voltages the weaker (lower capacity) cell.
I think you got it correct. I have a small pack 6Ah 24V and float the pack at 3.45V per cell. Keep in mind we are talking mV here when was the last time any of us have sent our meters/BMS systems out for calibration?
Once my cells reach 3.45V the absorption current drops into the micro amp region. The Cells are chemical electroplating baths, ion movement is proportional to the current flow and not the voltage. No current or very minimal current results in no chemical reaction. However too high or too low a voltage may cause other side reactions that are not desired for cell life. I have no BMS or balancing ( just a fuse for over current) and so far the cells have maintained a good balance, but my loads are very low. My recommendation, charge to 3.40 to 3.55V then balance if you must with no load into or out of the pack. Most of your cell voltage divergence results from the heavier load current and the cell's differing internal resistance and not so much from the generally lower charging current.
Andy, since the smart shunt will re-set the SOC, what are the capacity numbers for your pack with your new charge curve? I bet you are still above 95% of the original 280AH. I really think you are on to some good ideas and thanks for your work.
The balancing sensing depends on particular BMS. Ideally, all cells are snapshotted for voltage at the same instant in time.
Problem in real world situation of variable inverter load and variable solar charging is the battery current jumps around continuously. Only way to to get accurate voltage relationship between cells is to get all the cell voltages at same point in time with same battery string current.
Cheap BMS's just average the voltage readings. It helps but not super accurate. Most cheap BMS's have fixed balancing trigger voltage of about 3.4 vdc. This makes it easier for cheap BMS because it pretty much means there is no inverter loading reducing how much the cell voltages jump around over time and the 3.4v trigger accentuates the voltage difference for a cell approaching full charge. The higher balance trigger voltage gives less time to balance before highest SOC cell overvoltage shuts down BMS.
Balancing becomes a race between lowest cell reaching desired charge before the highest SOC cell triggers the overvoltage shutdown. The higher the balancing current the better the chance to win the race. Waiting until charge current drops off before balancing will not help. Are you going to ensure there is no inverter running? Not very practical. You will likely trip the overvoltage if cells need more balancing. 200 mA balancing is about the limit before you need to have separate cell voltage sensing wiring. On a common sense/balance dump current wiring the balancing current wiring voltage drop on sense wires impacts cell voltage readings. They shutdown balancing to make voltage reading but this reduces balancing time a bit. This is the main issue with cheap higher current active balancers.
Charging is not just about voltage limits. The higher the charge current the sooner it reaches absorb voltage, but the longer the current taper off period lasts. The lower the charger current setting the longer it take to get to absorb limit but the shorter the absorb current taper off period.
I would not advise using a timed absorb period for Li-Ion batteries unless you keep absorb voltage lower. If you time absorb for 2 hours at an absorb of 3.65v per cell it guarantees overcharging and battery degradation.
If you have a BMS that allows you to lower balance trigger voltage it maybe good but depends on the BMS simultaneous cell voltaage sensing. Also lower balancing voltage trigger requires better accuracy of voltage sensing as the cell are closer in voltage without the benefit of the near full charge ramp up in cell voltage exposing the difference in state of charge of cells.
Excellent comment 👍🏼
The purpose of balancing is NOT to reduce voltage.
The purpose is to equalize the SOC of all cells. Top balancing tries to equalize all cells at 100% SoC.
That is why balancing makes sense even when the charger is still pushing current.
It's irrelevant to compare the charging current vs. balancing current.
You need to compare the imbalance that you want to equalize to the duration of balancing multiplied by the balancing current. So if you have two hours to do the balancing and 200 mA balance current, you can at most equalize an imbalance of 0.4 Ah. The charger current is NOT relevant.
Top balancing works on the assumption that voltage is an indication of SoC. This is a problem for LiFePo4.
1 mV imbalance around 3.35 V could mean a 1% or 10% difference in state-of-charge (slightly exagerated for illustration purposes). It's very hard to tell SoC from voltage with LiFePo4 below 3.45 V. It would require very high measurement precision, beyond what a typical BMS is capable of.
1 mV imbalance around 3.50 V means a very small difference in state-of-charge.
That's why balancing at a higher voltage > 3.45V makes more sense.
Now the question is: how much imbalance are you willing to tolerate in your pack? How much additional imbalance is created in one charge-discharge cycle?
The usable pack capacity (Ah) is limited by the cell with the lowest SoC. You want the pack to be balanced to get to use the full capacity.
I fully agree with you. 👍
Thanks! Yes! The difference between the (for example) 10A and 9.8A could still be relevant for balancing.
@Daniel Ardelian Great explanation.
Maybe the balancing to within 10 percent state of charge is fine for him because he gets more longevity out of the cells. Think about it since he doesn't charge right up to 100 percent really fast his batteries slowly charge to 98-100 percent over a long day. So the time the battery is at 98-100 percent everyday is only a couple hours per day versus a fast charge to 100 percent would mean the battery is maintained at 100 percent for 5-6 hours every day. The time at high state of charge affects the life of the cells.
@@uhjyuff2095 I agree with you that charging to 100% SOC and keeping it there for a few hours every day does not really make sense.
I believe it would be perfectly acceptable to charge to balancing voltage only once per month (Victron Energy actually recommends this for their LiFePo4 batteries).
Or you could keep an eye on the amount of charge (Ah) that you put in and extract back in a cycle. If this drops noticeably, one possible explanation would be that the cells are drifting due to cycling and it's time to re-do the top balancing. But with a 200 mA balancing current, this is going to take a loooong time because the imbalance is high. Probably better to use a power supply to charge all cells individually to 3.50 V.
1. The available charge (Ah) of the pack is limited by the available Ah in the lowest charged cell. You balance to allow all cells to get to their maximum charge (100% SOC) so you can maximize the usable capacity of the battery.
2. If the pack is badly balanced, you risk that during a normal cycle, one cell gets to the BMS disconnect voltage (min or max) and the BMS cuts off the battery and you are left in the dark.
From what I've learned, the main purpose of cell balancing is to achieve a "collective higher battery voltage charge" thus increasing your overall battery capacity. Either you're going to bottom balance, or top balance; or, you're not going to balance at all. If you're not balancing for the sake of capacity, and you're only operating the cells in a voltage range that is safe, then why balance them. Simply set your battery low and high cut off voltages to the state of your unbalanced cells. Don't waste your money on a balancer if you're not going to use it for what it's designed for.
Thank you for your educational videos. They are very helpful.
@Off-Grid Garage Andy, you rock! Your investigations are intriguing and very interesting even for people that are very savvy with solar. Keep up the excellent work. Great vids too.
Second post, yeah, you went down the rabbit hole I was, but you took the time to actually test and present. Thanks, this really helps my thought process. I agree. The pack that stays together, is a happy pack. I see the variances at the top/low ends. Why go there? Just build the pack bigger to compensate for the lost capacity, which is minimal.
So many good points. I've set mine (a Daly with only 30mA) to start balancing at 3.0v. The cumulative effect will add up to balance the cells nicely. 6 months in and it's been behaving really well.
I should add, that the Daly only balances during the charging stage. And to detect that, it needs to be >1A in my setup.
Great advice, thanks for sharing!
@@OffGridGarageAustralia But balancing is better ( = a smaller final mv difference ) during lower charge amps, not with higher charge amps ...
Are you still using this method ? i have Daly built inn also in my 100A lifepo4 battery
Always a good amount of things to think about. I am running a 4S11P 100AH GBS kit. Grade B cells. Every few weeks I gather another pile of cells that I am able to P into more so I might be around 30P by now. I have chosen to limit my top voltage delivered by the SCC to 3.4 and do run a overkill 120BMS and watch the P voltage carefully. By limiting this top voltage I have never had the BMS try to balance cells…and I am happy with that because there is no way it would even make a dent in it with such a low balance rate. I only push 500W in and only run a simple freezer unit so the pack voltage is always less than 3.4 and when on cloudy days it may drop to 3.0 volts I have a transfer switch set to throw it back onto grid.
Andy and Will P have filled in my every decision. Thank You.
I love these type of videos. We are learning a lot with the experimentation you are doing. Do you think you are getting close to saying something like; "ok, for a stationary setting if you want your batteries to last for a very long time here is the results of my experimentation. Set absorption voltage at XXXXX, set float voltage at XXXX and set float time at XXXX.? If you want to charge your batteries faster for like a mobile situation the settings I recommend are XXXXXXXXXXXXXXXXX with the knowledge that battery degradation will likely occur at a faster rate with these settings. Are we almost there?
What did you learn from this video? The graph is wrong ...
Good stuff!!! Love getting into the nitty gritty, the core, the essence, substance, the brass tacks, nuts and bolts, the foundation, the meat & potatoes, da whole story. Sorry, I’ve been drinking. Anyway, keep up the good stuff.
Thanks Andy, you are demonstrating what I plan to do in this video!
I think you are right. The problem for longevity is lithium plating. This happens when the highest cell is pushed too high and current continues. Victron is risking that more than your settings especially if the BMS used has a low balance current.
If I am correct balancing is not critical if you stay away from charging near full SOC . So victron is risking plating just to solve the problem of plating.
You need to reset your shunt as infrequently as possible and balance at as low a voltage as possible even if it takes longer.
Marine may be a more critical environment, they may need to compromise longevity for a little more capacity.
So after watching this video and understanding lifepo4 cells the way I do. Working and building battery systems experience has taught me a lot .
You can't balance cells that are not fully charged meaning less than 100% soc . They never reach the high voltage knee .fact 1.
Fact 2 you also cannot balance cells that are at a float of 3.325v . Why ?
A cell will only show it's mismatch in capacity (i.e. requiring balance ) when it has a positive voltage applied - which is directly proportional to it's soc level .
Applying voltage / current to a empty cell will never push it over in voltage because it hasn't reached its saturation point decreasing internal resistance .
So here is a test charge all cells at 30amps or 60amps and watch their voltage sure you might see a few MV difference In a sort of balanced pack .take note of the voltages and keep track of the mismatch at a medium low soc .
Allow the pack to charge up till close to full around 90% you might still only see a minimal difference .
At around 95% soc you start to notice larger variances of voltage between cells that are closer to full and cells that are not so full ..
And this is where balancing starts to work because you have the positive voltage differential allowing this process to show which cells actually require balancing .
So ... To say you could balance cells at float .. after charge can be done yes but will it work ?
No ...
You need voltage to trigger the knees in all full cells allowing them to be balanced . Lifepo4 cells undercover agents they don't give information unless forced (charged ) to do so .
Cell deviation voltages are soc related and current related .
Your target voltage is the first step sure you don't have to go to 3.55v aim lower but not below 3.4v .. on massive large cells like the ones you have you never going to heat the chemistry to degrade the electrolytes because these cells can take 280amps of charge .. recommended charge for cell life span is 0.5c which is 140amps of current .
If I see correctly you still less than 0.25c which is less than 70amps .
You never going to create heat internally in the cell at this charge rate even closer to the knee aiming for 3.5v per cell so don't sweat ..this factor is only important when charging small cells with a higher chargr current
@@adamsaiyad3959 But yet, he has only a 5 mv cell difference across the entire pack. WHY should he care, that the voltage difference might be higher with a higher voltage, given he will *NEVER* use a higher voltage?
@@MrSummitville valid comment . 5mv at what soc of his entire pack ?
Is this 5mv at 60% or 99% soc .
You bring up another interesting topic . How much energy capacity is created by unbalanced cells .
This term is loose as unbalanced can be a full cell and a cell that is at 50% soc .
Then I digress abit further and into more detail.
If charging up 290ah cells .
And of a group of 16 you have say over 80% of the pack at 3.48v and the remaining 20% at 3.9v at the end of a charge with no balancing how much capacity would the say pack have lost .
Probably around 10ah at most maybe 20ah
Over all on a pack that is this huge it's a minimal capacity loss for this unbalance .
Then brings the next topic if the cells are of the same batch . And are closely matched from the supplier on IR and capacity. And are top balanced technically the pack should go out of balance .
A top balanced once and never again balanced pack should remain the same through out the life span of the cells .
Same for a bottom balanced pack ..
So is balancing necessary debatable . Good quality cells require very little balancing on each cycle and a small balancing Bms can handle this easily .
But on the other hand with large cells require a larger balancing current rule of thumb is again a misnomer . Because balancing can still take place as rightly mentioned in the video at a low enough chargr current towards the end of the charge profile in the CV stage.
@@MrSummitville ps as I mentioned in my first right up wanting to balance out a pack at a low voltage will never yeild correct balancing results . In essence might not balance at all .
You can only truly balance when you are in the full cell knee and IR lowers because of soc where the propensity of the cell to raise in voltage requires very little current to raise above ocv. Open cell voltage when full .
I didn't say he has to charge to 3.55v .
I did say that he should use a value above 3.4v to achieve the positive voltage difference to allow current to flow into the cells in a ample amount of time.
Charging at 3.30v will charge a cell . But it might take 20 years to full .
Charging at a greater voltage differential allows for more current lowering time taken hence speeding up the charge process.
We not applying 3.65v per cell as per manufacturer specifications .
We applying by far lower voltages taking longer to charge but still getting the knee activated and allowing some form of correct balancing if required and needed by the user.
Hey Andy, I love that you are a testoholic like me! With your new settings, what is the highest mV deviation you see on the next charge up to 3.45V per cell? Myself, I've been charging to 3.45 with a long absorption time and balance it at 3.45. I can see with a system as large as yours is, keeping it in the flat of the curve would be easy to do.
As far as I know, LiFePO4 batteries do not degrade more if held at 3.65V or 3.55V continuously for several years versus 3.45V or 3.35V for the same number of years. High voltage induced degredation is caused by electrochemical side reactions resulting in permanent loss of electrolyte as it gets consumed in those side reactions and deposits insoluble breakdown products over the active electrode surfaces. However, these side reactions start around 4.0V and reach excessive rates by around 4.3V. Li-Ion and LiFePO4 cells use the same organic electrolyte, so theoretically you can safely and continuously use a float voltage of almost 4.0V on your LiFePO4 cells without any harm. The reason LiFePO4 cells aren't rated above 3.65V is because they store no extra energy above this, meanwhile their essentially infinitely steep Voltage/SOC curve above this allows individual cells to skyrocket over 4V extremely fast when charged in a series connected battery pack and still under the normal full pack charge voltage. I.e. high voltage is bad for Li-Ion lifespan, but merely difficult to control and of no benefit in LiFePO4 batteries.
hi. have a link that explains this info you just said?
@@cgmarch2359 Do *NOT* Float Charge any LifePO4 cell at 4.0 volts. It will cause very early death. All Cycle Tests show this to be true.
"It is a bit theoretical" Yep - but you explain it very well.
Keep it up
LoganV, Affirmative. Every thing starts with theory and then into practical. Where as, in YT, mostly what you see is the other way around ie: even trial and error method., which’s dangerous and wrong. The Author; here, has done tremendous amount of painstaking research to get to the bottom of some intricacies and has been successful, in unfolding issues.
True, for people like me, it takes a wee bit of time to comprehend the essence of what is being lectured. But eventually gets clear. Thanks to the Author.
Very helpful, thanks. I've been wondering about these details as I get ready to set up my system, now I know!
Glad it was helpful! Thank you.
Hi Andy. Nice video. I want to add another "option" for you. I think you cant balance your cells at 3.35v. 1mV could be a lot of SOC %.
I think it is better to make balance at higher voltage. Probably with 3.4v is enough to appreciate SOC % differences by measuring voltage.
(Please, now wait to read all text. I dont want to ecualize cells, i want to use those settings like other way to do top balance)
What about using the ecualization utility of your mppt ? I mean, you can go higher once a month, for example. Let's go up 3.5v once a month with a ecualization current limit (perfect to balance cells as it will be limiting the maximum current that you want) and a controlled ecualization time. It will be our configurable top balance "voltage, time and current" mode.
Please check your mppt settings. It makes me sense for me, dont it ?
Charge your cells daily up to 3.35 or 3.40, but only once a month (or whatever) go up to 3.5 in which you will have more possibility or it will be easier to balance cells.
Sorry, this option may top balance our heads also 🤯
Thank you Fernando. Using the Equalisation settings for a routine maintenance... hmmm... that's actually not a bad idea. I'll look into it. Thanks for your suggestion, that is great.
Always a pleasure watching your videos and learning things.
I think that the objective of balancing is you have the same state os charge in every cell, and you only can get that in high voltage, 3.4 or above. If you have set it lower the diference will not be noticeable and when you pack goes full, you migh have cells that are much lower or much higher then the other, so not in the same state of charge. This will afect you on low voltage as well. You can have a diference that your pack may trigger the bms in some point. This as happened to me, i had one cell so high it would trigger the bms and the pack voltage was 55volts. My settings is bulk to 56 volts and i float at 55. I cant use absorb so i use the balancing starting at 54.9. My bms balances at 2amps so it is very efective. I had to do a new top balance and now they are similar and the bms is only balancing if it detect 0.003 diference and above 54.9. This is more efective i think for balancing.
what active balancer do you use?
@@AJTarnas i use the JK BMS 150 amps balance current 2 Amps
@@MiguelSilva-mb6mb But yet, after 3 months, Andy only has a 4mv - 5 mv difference ...
@@MrSummitville yes. If he charges to 3.6 he will see that the diference will be much bigger than that. Andy, thats a test you could do. Try to charge to 3.6 ou 3.65 to see what is you real delta between cells. For what i notice in my system the bms will trip because the state of charge can be much diferente and more noticeble at higher voltage. My pack is not perfectly top balanced and i only notice the lack of balance at 3.55 or more voltage. At 3.6 is perfectly noticeble that some cell are more high than other. I get 0.01v at 3.62, and at 3.4 i get 0.003v delta max.
@@MiguelSilva-mb6mb But WHY do we care that their might be voltage difference, at a higher voltage? He is never using that higher voltage. So, how does that impact operation at lower voltages? After three months, it has had absolutely no impact on his operation ...
Me, I like your conservative settings, they mostly make good sense. I dislike drain-style balancing though, but for your purposes it's adequate. The one thing I've learned from you that I didn't learn on my own is that you can float them to full capacity through absorption at lower voltages, I thought the voltage to state of charge curve was fixed, but actually, it's only once the cells are close to full that the charger can overwhelm the ability to absorb energy and send the voltage up higher and that's actually a side effect and thus *any* higher voltage over static "full" is sufficient to produce near full capacity. That's really interesting and really useful. Thanks :-)
I made a video. I'm blushing. :D
I experimented with balancing at 3.35v and it does *usually* work. However in my experiments I found that at 3.35v it would often pull current from "strong" cells rather than weak ones which can actually worsen the overall imbalance of the pack.
Why does it pull current from "strong" cells instead of "weak" cells? In my experiments I noticed that weak cells would hit 3.5-3.6v first, but then when the charger would shut off or switch to float mode (Or if my shed load turned on) the weak cells voltage would drop faster than the stronger cells would. So during "absorption" your weak cell might be at 3.55 and your "strong" cell might be at 3.45v. Now you switch to float and your weak cell quickly drops to 3.33v while your strong cell sits at 3.37v. Now your BMS starts to suck power away from your stronger cell further worsening the imbalance between these two cells.
I can already hear you asking: Why would the weak cell go from being higher voltage than the strong cell to being below the voltage of the strong cell so quickly? Two reasons: 1) it has a lower overall capacity, so it's voltage will drop a bit faster and 2) Weaker cells generally have a higher internal resistance so even if they have identical capacities their voltage will be lower because more of their capacity is lost to this internal resistance. Also: Andy already mentioned this but because of how flat the curve is at this voltage it's really hard for the BMS to tell which cells are weak and which are strong. You have to push the cell a bit into the knee to see which cells "stand out" from the pack.
I found that I couldn't trust the BMS to balance at anything less than 3.4v. I tried 3.37 and 3.38 as well. Anything under 3.4 and I found the BMS balancing strong cells way too often for my liking. This not only wastes energy (passive balancing just burns the excess power as heat) but it can also worsen imbalance which defeats the purpose of balancing altogether. So I just try to minimize the time the battery spends above 3.4v for balancing. My shed load technique has worked well for this: Anytime a cell gets above 3.4v it turns on a load. If there's lots of sunlight and the solar panels produce more than that load can pull away. (IE: My 3.2 kW array will first turn on one 1,000 watt A/C unit and if there's more than 1,000 watts of sunlight and it continues pushing the battery voltage above 3.4 it'll eventually turn on a 2nd unit for a total of 2,000 watts of loads). This gives the battery just enough time during peak sunlight periods to let the cells get above 3.4, balance a bit, and then drop back down to levels where degradation is not an issue.
So basically my cells only get balanced on especially sunny days (And even then only for a few minutes at a time) which helps me minimize the time they spend above 3.4v. Over time this has been enough to maintain decent balance of my pack. I also have active balancers which help to minimize this time as well because I have a couple cells that I have replaced over the years and those new cells are obviously much stronger than the rest of the pack which make maintaining balance a bit more challenging.
Hi Jeremy. I was wondering which active balancers you are using and do you have more than one paralleled?
@@itsmeasis I use these: www.electriccarpartscompany.com/3v-1s-lithium-lighted-battery-balancers
And no I don't have multiple in parallel.
Thanks Andy for the video.
I have been using your figures and my system is working GR8.
Sorry Andy, I don't quite agree here. There are many ways to implement the balance process. Gently balancing while charging is like gently steering while you are driving on the freeway (and it works regardless of the speed you are driving), plus it is better for the cells since it slightly reduces their peak voltage by subtracting SOC earlier. Small nudges make a difference in the state of charge, as small steering adjustments change the vehicle position within the driving lane. A good battery pack only needs small nudges to keep balanced. If a pack needs more balancing it won't be balanced at the end, but that's a problem with the pack. Whether the balancer burns 200mA for say 1 hour while the battery is charging or afterward either way it has drained the same 0.2Ah and the same SOC percentage from the cell. The earlier it drains this the lower the peak voltage the cell experiences and the better it is for the life of that cell. Either technique works to balance the pack. There are many ways to balance. Gently balancing is valid as soon as the imbalance can be correctly detected, and the optimal time to start balancing is as soon as a valid detection of imbalance can be made.
Thanks for your comment Alan.
This balancing method you explaining here (charge balance) does not work with LiFePO4. You cannot balance these cells at a lower voltage. It makes no sense and causes imbalance. Once the voltage rises finally, the BMS has no time to counter this action any more and the battery will never get balanced ever.
These balance methods may work with li-ion as these cells drift quite early but LFP, no chance. Balancing only works after the voltage knee which is around 3.4/3.45V.
@@OffGridGarageAustralia Hi Andy, I'm sorry that you misinterpreted my posting. You were saying lower voltage, I neither said that nor intended that meaning. Balancing can start at any voltage where imbalance can be CORRECTLY DETECTED. This voltage must be above the flat part of the charge curve as no proper determination of balance can be made until the voltage begins to rise at end of charge (note that this voltage is a function of temperature, so picking a fixed value that is too low may cause problems when the seasonal temperature shifts move the system voltages upward, the default values are chosen for a wide range of temperatures). The most important point here is that there is no need to wait until the charge current is zero. That is not important. As soon as imbalance is correctly detected the balancing operation may start. It does not matter that only a small current is subtracted from a larger charging current. The SOC is being steered, and that is the way balancing works. The SOC must be steered so the cells match in voltage. The balancing is adjusting that SOC so the voltages match. Gently steering the SOC is totally valid, as long as the information is there to do it correctly. Doing it at too low a voltage fails to work because the correct information as to what cell's SOC to adjust is not available. The cell voltages don't reveal the SOC during the flat part of the charge/discharge cycles. Thanks.
andy, this video i did like :)
also i think,i like your ideas with the voltage..
but i think it would be a good idea to test this a bit over a longer timeframe.
i will however in my mind never harm the pack or cells
Very interesting video! I am a newcomer to lifepo4. But I finde, that for my 24v system in UPS usage scenario without PV is good to use charge bulk 27,8v(3,475v) 45A without float and without absorbtion- dont reach high SOC for healthy. I have diff voltage usual at 0.005v at 3,32v per cell in rest.
But, in my oppinion, due to the smollest difference in tempreture and internal resistence (but I mesured all the same 0.18mOm on each 8 cells 202Ah Lishen) under load and during chage, we get increasing unbalance day by day, that is not so visible at flat line 3v- 3,3v. Lower 3v and higer 3,55 diff goes up to 0.200v under load or on charge to 3,55-3,6v!
But for more acurate balancing i my Must pv3024plus 24v I have option "equalization" and nobody doesnt explane me, that it is not only for AGM or Lead battarys and also good for LiFePO4 battary.
In "equalization" mode it is the same charge curve that Andy shows in this video!
In mode "eqialization", once for a month or once for a weak, whatever... I can set voltage higher, than my usual charge bulk 27,8v(3,475v) to 29v ( 3,625)v, for example.
In this case voltage goes up in 2 steps: to 28v CC/CV 1st and than very slow to 29v with 10A to 1A increasing current and goes in flat line at 50mA carrent for equialization - some absorbtion ,for exemple for 60 minutes and Neey balancer works hard for eqialize, balance cells and charge current goes down to almost to 0A. After that time charge stops and voltage goes to prefered float voltage or fully disconect charge.
This sound's good to me andy keep up the greate work.
Once top balanced, you should never have to balance your cells for months (depending on your use). They will NEVER be balanced perfectly and that is okay.
Fundamentally with any real world and thus non identical cells your latter statement is a bit off. It's not that they will never be balanced perfectly, it's that they cannot be balanced except by perfect manufacturing practice. All we can do, as end users, is top balance, bottom balance, or dynamically try for both IF the cells are close enough together AND for some reason you have to push both ends of the envelope AND your active balancer can pull enough current to avoid an over voltage scenario for individual cells/banks.
Even a year, or more. LiFePO4 cells tend to stay in balance for a very long time, as long as you do not fully discharge the pack, which is why the relatively low balancing current the BMS is able to provide is plenty.
-Matt
@@junkerzn7312 Yep, that's what I said! :-D
Andy I wish to respectfully disagree a little on balance curent. Yes the balancing cell is not discharging but it also means all the other cells are charging more. Therefore they are getting more coloumbs. The question becomes is the difference enough over the time it takes to go from balancing to overvoltage of the highest charged cell. Really we need a integrated BMS/charger that can taper current based on individual cell voltage.
If the cell with higher voltage is charged less and the other ones are beeing charged more, that's actually good and part of the balancing. I just don't believe the BMS is capable of handling these high capacity cells. The 0.2A balance current is just not enough to make a difference in the time available to it.
A charger which can handle all cells individually would be amazing. I know Peter here is working on somethign like this.
Well since you are using a discharge BMS balancer then what you say works. My BMS has active balance so my balancing takes place while floating but also when charging as long as the charge is below 10 amps.
what active balancer do you use?
@@AJTarnas Smart Ant BMS with balance and 2amp ICgogogo active balancer. Andy has nice new batteries so he doesn't need one.
Why has everyone on youtube started teaching to use 100% of the batteries capacity? What happened to leaving the top and bottom 15%? We know you get a higher cycle life. Good work.
Everyone? Nobody here, says that. Many do say, "... Use 100% of the capacity ..." So, that they can justify the high cost. But then in the real world they end up using only about 70% of the capacity ...
@@MrSummitville I don't tend to read comments. By youtube i mean content creaters. I agree with your comments.
Need to publish a paper on this … do more experiments, I love how great engineers Germans are.
062521/0540h PST 🇺🇸 @ Paul>>> It’s a universal truth that the said country has always been a forerunner to almost many inventions that we see today. Many may not acknowledge that truth, though. Thank you for your sentiments, as I too acknowledge the truth. Stay safe and 73s…
Nice Andy you’re proving the theory with data real data not only with manufactured specifications but with real data that’s wonderful I which I have all the test equipment you have to do the same test years before it is making totally sense all your data prove the real stuff I’m 100% with you thanks for all the help and information I really enjoyed your channel 💪🏼👍🏻💪🏼👍🏻
Very informative explanation professor andy.,👏👏
Thanks Andy
The basic answer is no. For LiFePO4 you can only properly balance cells when the voltage is modestly higher than the cell's resting voltage. A Float of 3.375V is not high enough. Sure, you can try to balance at a float voltage of 3.375V, but equalizing the cells at that voltage won't actually balance them. At all. Why? Because the balance current would be too low and the voltage has too much freedom to move around with very little current when its that low. I would go as far as to say that balancing is basically impossible at 3.35V or 3.375V.
So basically, balancing is only effective at the Absorption voltage (that is, a voltage modestly higher than the cell's resting voltage).
In terms of balancing while in bulk... actually this can be done to some degree. If the BMS is smart enough and stores the cell history, it can figure out what kind of balancing is typically required during Absorption and it can try to apply it during bulk. It would basically be an educated guess, but once it finally gets to absorption it would then have less work to do to clean-up the cells with a final balancing. I don't know any BMSs which do this though.
--
Battery degradation is more a matter of the SOC, *NOT* the bulk target voltage. Remember that the actual break-down voltage for LiFePO4 is 4.2V. Other lithium chemistries have less margin, but LiFePO4 has a huge margin. Bulking to 3.55V isn't going to hurt the battery.
--
Remember that the proper target voltage for the charge controller depends heavily on the charge rate. If charging a battery slowly, lower target voltage is necessary to avoid over-charging. If charging a battery quickly, a higher target voltage is necessary to avoid under-charging. There is no 'right' answer... the answer is 'it depends on the average charge rate'. With the amount of solar you have now, you are probably charging at a high-enough rate to warrant a 3.50V or 3.55V bulk target.
And I'll give you another reminder that with the number of strings you now have, blocking diodes are now mandatory. Look into getting a proper solar combiner box.
-Matt
You said, " ... Because the balance current would be too low ... ". Oh, really? At 3.351 Volts the Balance current would be ~90% of the balance current at 3.6 Volts. So, how is that, "too low" ? In the real world, the Balance Current is not too low ...
Interesting concept have my 4s EVE pack charged to 3.55 and balancing starting at 3.4 but i use a 2A active balancer with the REC ABMS. i do look at your comment of does this really matter? these cells last so long that even if you cane them and reduce there lifespan by a few years they still last 15 odd years and by then there will be something with a much larger energy density
I have the same setup with REC and EVE cells but have balancing set to start at 3.35.
Have noticed the BMS actively controls the charge parameters and does most of its balancing around the 13.8v point on the way down from 14.2v after the some of the cells hit top voltage at 3.55v, do you find similar?
062521/0549h PST 🇺🇸 @ BoatingTube … what is your BMS and the Charge Controller make? These two devices play a major role in setting up charge/volt parameters. FYI, even if your BMS (?) has default value and is not configurable, you will be able to increase and or decrease charging current/volts per your choice through your Charge Controller. [ Assuming your CC is VICTRON ]
There’s no need for AB if your CC and BMS are from reputable manufacturer. 73s…
How do you get the 15yrs of life?. Is there any research which has proved such a very high cycle life?
Before even watching I maintain you can only properly balance at the rapidly changing bits of the graphs you have been generating. ie at the very high or low voltages. And only properly check the balance at these extremes too.... now I unpause and watch :)
and maintain it still. However I only think you need to worry about balancing once a year. I DARE you to turn OFF the balancing for 3 months - then do a higher voltage/cell check and see how little it has gone. I don't think your balancing is working, but because your pack is already balanced, it doesn't have to do anything - so compare to 3 months with no balancing.
@@KevIsOffGrid You are right the balancer may not do anything because the battery is perfectly balanced, but it might change in the future so having an automatic balancer is the way to go. He doesn't want to make balancing batteries his full time job. Just turn it on and let it do the work for you!
@@uhjyuff2095 I have a similar pack, 24 cells. I check the balance maybe twice a year (no balancer) its in balance always.
We use new cells - no balance issue.
We cycle at low C rates - this isn't an EV with massive current draw, Andy's system 1C is 280Amps, all summer he had 1 charge controller at 35Amps - so couldn't even get to 0.2C charging, he could just about discharge at 0.2C if the 3000w inverter was flat out on the tesla charging...
Balancing isn't a full time job, it isn't even a part time job, its just a non-issue at all. Just use the battery and that's it, Solar is so gentle on large battery banks :)
@@stupidmonkeykev Well check this out for an example. He starts charging his tesla and the low voltage cutoff on his inverter is set too low and then the BMS has to shutdown the discharge. The low cell is what caused the bms to cut off. Then he does this once a week. So one cell is getting slightly degraded more than the others due to lower voltage discharge every time and then the cells slowly start drifting out of balance more and more year after year. Having the balancer on for long periods because the packs are so large and the balancer is so slow is just a great feature to have and something that just works without having to think about it. Obviously, this is a worst case scenario but most people have no idea about these lithium batteries and how they work.
@@uhjyuff2095 - In what video did you see ... one cell is getting more degraded than than the other cells and then the cells drifted out of balance? We see no evidence of that happening ...
I don't kehr - Schornsteinfeger verweigert Arbeit...
I think the same. As long as you monitor and are able to recognize problems, you can do whatever you like.
And now let the sun shine ;-)
Wouldnt it be Smart if the BMS tells the Charger or VRM oh today my cell deviation is high today i wanna charge to a high Voltage? And on other days you dont charge to a high Voltage to not stress your cells?
I like the way you have an idea then prove it I also like using the BMS on my shoulders 🤔😁👍
His graph is wrong. He proved nothing ...
You can do whatever you want to with your own batteries through your batteries
I say that's fine I do something similar as long as you don't randomly want to go to a higher voltage and expect one cell not to jump away and go higher than the rest there's no problem at all doing that there's even balance boards that just continuously balance all the time and then there's some that just turn on a resistor when the voltage is above say 3.6 or 3.65 volt and the resistor just stays on forever as long as the voltage is high enough
Great topic and video. This has been an area on my mind for awhile. The only catch I see, is many BMS's are hard set to start balancing at 3.55v. The best I can figure, it so set your bulk charge voltage to 56v (3.5v), thus, when that first or second cell hits 3.55 it will start balancing off the the high voltage to the pack. With the voltage being near full, the risk of over voltage should be minimized. Thoughts?
Im probably thinking of this all wrong, but if you want 'balancing' during the float period, would it not function better to set the balancing at 3.349v instead of the 3.351v that was selected? Im presumably misunderstanding the situation, but surely if the float voltage is being maintained at 3.35v/cell(equivalent), then the balancing function will not be active if set at 3.351v (for most of the time, since loads are being fed from SCC)??? Furthermore, I would have thought, that at 3.349v/cell the highest cell would constantly be getting 'bled off' , all whilst the SCC attempts to maintain the float voltage setting point of 3.35v. In my crazy mind, this seems to me that the balancing would get done very gently using very small currents(not bouncing all over the place) , over a longer period of time, producing more stability???
I remember you use to say keep your state of charge at about 80-90% to keep your battery from degrading. Is this still true? Or can you keep your battery at 100% state of charge just at a lower voltage?
Thanks for another great video Andy.
Why do you need ot go above 90% SOC ?
@@MrSummitville Most people won't need to go above 100% SOC. I want to know if it degrades your LifePO4 battery if you have it at 100% SOC at lower voltage.
@@popcoingaming5086 No, my question to you was ... Why do you need to go above 90% SOC ? ( Your reply was about greater than 100% SOC
@@MrSummitville Your question to me is "Why do you need ot go above 90% SOC ?" My answer is, I don't need to. I want to know what Andy's thoughts are currently about being at 100% SOC at a lower voltage.
@@popcoingaming5086 He is experimenting and trying to get more subscribers to his channel ...
11:42 what you say about where your balancing occurs is not correct, it actually starts on the up slope to the left of that - it's not important that it doesn't drop the voltage of the cell with higher voltage, it *IS* pushing more energy into the others relative to that one, and is therefore charging them more quickly and working towards an equilibrium. The trouble is that at those higher voltages your HV cells (low capacity cells) can be getting damaged while your higher capacity cells are just charging normally. That's why I much prefer full time active capacitor balance boards. They can shift more current and they can run full time with low draw, so as soon as the full cell voltage starts to rise they kick in immediately and start to shunt current into the bigger cells. This matters on the top and the bottom equally. But if you do have a genuinely different sized cell and you do reach the bottom AND the top then the chances are this will not be enough to avoid an over voltage scenario. Hence it's better to stay in the constant-voltate part of the curve on the lower side and only push the top end on the high side as gently as possible to generate the cell difference required to activate current flow between cells. Once top balanced you should see very little cell drift assuming no self discharge differences riding on top of the capacity differences.
There is ample evidence and a lot of papers showing that STATE-OF-CHARGE is the stressing factor (along with temperature and charge / discharge rate). Charging to 100% IS stressing the cells. Pushing and pushing and pushing the lithium ions into the graphite (aka INTERCALATION) is what stresses the cells during charging.
There is almost no evidence that voltage is the stressing factor below 4.2 V (yes, even for LiFePo4). That's the voltage where some chemical reactions start to happen that degrade the electrolyte (if I remember correctly).
Your cells SEEM to stay in balance at mid-voltages...but 4 mV imbalance could be 4% state-of-charge imbalance. You just don't see it. If you have one cell at 100% and another at 96%, you will only be able to discharge 96% of the battery capacity because the lower-charged cell will hit the bottom sooner.
I agree, not the voltage that degrades the cells it's the state of charge that does it. Lots of studies showing it is SOC not just the voltage. His balancer is just not smart enough to balance really accurate, maybe in the future balancers will be more accurate, but maybe not because it comes down to the science of these batteries. NMC seems to be ez pz to balance but LifePo4 is more complicated.
@Daniel Ardelian - I agree that time spent 100% SOC does reduce lifetime capacity. But I ask you ... Why do most manufactures specifically state .. Do not Float Charge at 4.2 volts per cell? Because ... 4.2 Volts is significantly above 100% SOC ...
At 53.2 you're only at about 80% soc so of course the cells are close together in voltage. It's when you get to the extremes (fully charged or near empty) that you will see an out of balance condition.
If there is imbalance then the cells are always out of balance, you just don't see it due to the flat voltage-to-SOC characteristic of LiFePo4 at mid-voltages.
You don't see it but it's there...one cell has more energy than another...Then when you discharge, the cell with lower SoC will hit the lower voltage limit first and this will limit the usable capacity of the whole bank.
I don't know how you know 53.2v is 80 percent charge? Did you have a smart shunt monitoring his battery all day long to measure the exact amp hours?
@@uhjyuff2095 A reliable test was done on these cells that yielded the following: 57.6=Full charge, 54v=99%, 53.6v=90%, 53.2v=80%, 52.8v=70%
@@SpeakerKevin lol
Why can't you use zener diodes to balance the cells? They come in 3.3V .
Though it really seems an ideal solution to have the right voltage of solar cells to charge each cell individually.
6x 0.55V cells in series, however many lots of six in parallel as you like.
You cannot balance any cells at 3.3V, that is far too low. You want to be in the part of the charge curve, where the voltage is rising, so 3.45V for LiFePO4 and around 4V for Li-ion. And this may be a different point for different batteries and also may have to be adjusted once the cells get older.
@@OffGridGarageAustralia yeah and I guess you'd need another diode, perhaps an LED in series with the Zener but opposite polarity. Which could increase the charging bypass voltage to 3.5, and stop the cell discharging through the Zener.
I'm no electronics whiz, but that seems to be the simple standard voltage regulator setup.
If an LED was used then you would get a nice indicator light telling you the cell is at voltage, when the Zener breaks down at the cutoff limit.
Just a thought.
Living on a boat and wanting stuff that can handle marine environments and be easily diagnosed and fixed from spares you have on hand.
I mean, its life or death if you can't keep vital systems running at sea.
I've got bought digital voltage readouts for each of my six cells. I think I can do alright piloting the vessel on electric propulsion by spade switching neighbor cells in a series of six from series to parallel for throttle and top and bottom balancing purposes while the solar and motor are connected to the bank.
I would be nice to have a backup for the voltage readouts incase they should malfunction. Though switching the cells to parallel would diagnose that quickly.
Complicated electronics at sea is nightmare city.
Maybe we need a solar charge controller with built in BMS!
SBMS0 and DSSR20 but not for 48V
Some BMS is out there will not allow the balancing function if the battery is not charging I had one like that I would rather prefer it to only balance above a certain pack voltage and above a set difference between cells but allow that anytime
With my current setup I use my victron shunt relay to turn on a separate balancing board when it reaches 100% state of charge and then I have set 1 hour absorption at 3.5 volts on mine
Why do you charge to 100% SOC ?
I have a 8s 24v EVE system. After doing my first balance at 3.55v (No need to go higher IMHO) my cells stay within 10/15mv from 3v to 3.4. I exclude both ends of the charge/discharge curve as usable because it is really negligible. If I were to need that extra juice, it means I did not size my system properly.
Hey Andy just catching up with you.. not sure if I can go that low effectively. Probably need more panels at this point. Certainly is good for longevity though at 3.35v...
Try not to balance you cells at all. You will be surprised how little will change to what you are doing now. You can take 3 years vacation, and when you come back, your cells will be still where they are today. I didn't absorb, float or whatever for almost a year with my pack now. My BMS would need a trigger voltage of 3,55V to burn 30mA away. None of my cells ever saw that voltage. The only time I had to remove a bit of energy from one cell group was, because it was a bit ahead from day one already. LiFePo4 and balancing, is the most overexaggerated topic ever :)
Andy, It looks like you got your system fine tuned or at least making friends with it.
Unfortunately, many of us don’t have the flexibility that your system has or it’s for a totally different use, but we all are fighting the same gremlins.
I was doing some tests on a quick turn around charge for 12v 280ah battery for the trolling motor in an aluminum fishing boat. Use all day, charge nightly at home and be ready early the next day. Not being able to control the absorption time (dumb charger) and not having the luxury of time, it gets charged at hard at 50 amps. The BMS simply doesn’t have the time to work the imbalance with the feeble 50 milliamp balance current. I started seeing too much imbalance and the Overkill BMS would end the charge because of a high cell and with it the balancing. Things would only get worse each cycle. I found the Heltec capacitive active equalizer/balancer. I use the 3S 4S 5A model for my 12 volt project and the 13S - 17S 5A model for my 48v systems. They are efficient and haven’t noticed any heat because it doesn’t waste energy through resistors but channels voltage from the higher cells to the lower ones. Standby current is negligible but it can be put into hibernation by adding a switch. They work together (parallel) to the BMS’s leads at the battery. My delta? Usually within .004 Volts 30 minutes after my charger finishes at 14.5 V and never varies more than .012 V during the charge. Right now it’s fluctuating between .001 and .002 delta at 70% charge and the equalizer has been off since it’s last been used fishing a few weeks ago. If it’s not going to be used right away we don’t charge until the day before since they don’t like being stored at full charges. The thing is a beast close to 4 times the useable power and 10lbs lighter than the 100ah AGM it replaced. More running voltage to the trolling motor = faster, far longer and may outlast the boat lol. Cheers
Sounds like you have a handle on it, so to speak lol. What if you had two batteries that way you could give your batteries the time if needed and then switching your batteries every other day instead. More cost of course, anyway just a thought.
@Karl Jensen - So, you are masking the symptoms of a *BAD* Cell ?
@@MrSummitville
No, try charging any group of top balanced, high ah (270-320) grade A, impedance matched cells with a 1/4 C rate and virtually no absorption/ balance time with the most common BMS’s (only .050 milliamp balance current). Remember most BMS’s only balance during the last part of the charge and when the charge is disconnected the balancing is stopped , but necessarily complete. Eventually you’ll end up with an imbalance as you get to full charge. This is a a common issue for the avid fishermen, golf cart, fork lift, applications that get high use but need more than mediocre balancing current. Another thing people don’t realize is temperature difference of cells. The cells in the core vs the ends have different rates of heat gain or loss, either self generated (charge/discharge) or environment. Temperature differences can effect voltage/capacity. Then there’s time. Theses things are new. How matched will these things be 5 years from now? But there’s an easy inexpensive cure or a bit of insurance.
@@SkypowerwithKarl Charging at C/4 is typical for a 280 Ah 48 Volt Off-Grid battery bank with a 3,500 Watt PV array. It is done every day. And if you need a 5 Amp Active Balancer then one of your cells is BAD or failing - ie it does not match the other cells. Also, do not use a balancer that stops balancing when charging stops. In your situation, you created the need, for a High Amp Active Balancer.
@@MrSummitville
Hello, balancing after it’s done charging is an active balancer. Most popular BMS’s for internal battery/case use don’t have a large balancer because they can’t get rid of heat (resistance). Capacitive equalizers (balancer) are the future, very low heat. They may even open the doors for B grade non top balanced systems. Keep in mind that a solar system isn’t constantly disconnected and the battery won’t see a charge till it’s depleted. Solar is a constant give and take where it highly likely to have long intervals of absorption available.
Heya as longer you "work/test's" with these batteries you better you will untherstand how they work and what are the best settings
I run 2 banks, and allow 2 days for internal balancing (1 day minimum) before use, charged to 3.4v per cell, average. More ammo for your 2 bank solution!
Why are your cells getting so far out-of-balance ?
@@MrSummitville They aren't. It takes 1-2 days to run a bank down, and less time to charge.
Andy, with victron system what do you think about 18s configuration?
I´m using 18S Strings 3 in Parallel 280Ah so far i have no negative experienced, i can balance at 3.53-3.54 per cell (64V pack voltage), i have used cells so i have to rebalance every time i get to 10% or 100%. My BMS (batrium) cant keep up so i´ll be instaling active balancers soon aswell.
PS, i have a Victron 3 fase Multiplus 5000VA system
Hey OGG i seem to have a goofy Shunt also. I think the voltages and current readings i get are close but not exact. I have Valance Batteries (2) and i have no clue what each cell reads, I haven't been able to get my laptop and batteries to communicate. Do you think the BMS in those batteries are smart enough to regulate the voltage cells safely and properly on its own? BTW today is the first day my Victron Charge controller stayed on Float all day. Ive never seen that before. Ran my AC and for about an hour and the Float light never left Float mode.
I do to
Puh, I'm not alone... 😊
@11:47 - You drew a GREEN line and said, "This is where my balancing *STARTS* " Actually, the balancing started when the Cell Voltage first went *ABOVE* 3.351 Volts, during BULK Charging. How can your balancing "start" so late - when the voltage is actually FALLING below 3.351 volts per your graph? Doesn't your balancing actually terminate ( it does not begin ) when the cell voltage falls below 3.351 volts?
Question that you may have already answered... When you "float" at 53.6 I've noticed on my shunt/monitor that if there is a load, it momentarily uses battery and then the solar charge controller adds power until it is pretty much 0... perhaps 0-4W in either direction... If the load is larger and you remove it, momentarily it seems to charge the battery before backing off/etc. Is that a bad thing? Seems like at one point, the thought was to bulk charge and then essentially set the float low enough that it doesn't re-trigger any charging unless there is a decent drain on the battery.
Perhaps on a related note, the unit i'm using is a Growatt unit. I notice that the CC charging up to whatever voltage completes after which it "absorbs" for a few minutes at CV. As soon as that finishes, it goes into FC and floats at my desired float charge. The odd thing (i'm thinking this is strange) is that the battery meter shows that even with no load and inverter off, perhaps up to 200W starts being drained from the battery and tapers off until it reaches the float voltage. So it appears on those, that the unit actively discharges from higher voltage down to the float voltage. (FYI, my 24v system is using 26.8 as float and 29.2 as constant current charge per battery manufacturers recommendation - when i had CC charge to 27.8 they advised me it was too low as their BMS would never balance the battery).
I do not agree that your charge controller is doing this => "... the unit *ACTIVELY* discharges from higher voltage down to the float voltage ..."
Why don't you just use Tail current of say... 2 Amps instead of Absorption time? That way your batteries will be charged exactly the same every time.
Probably because it would run for too long at the higher voltage and endanger weaker cells during that period. His goal with the 10 minutes was only to trigger his SOC reset to occur, not for charging purposes.
@@fredio54 #1 He is charging at a modest voltage of 55.2 so endangering cells is not an issue. #2 On a sunny day and the charge amperage is high, he will hit the absorption voltage much quicker so the charging will stop prematurely compared to an overcast day which will have a much longer charge time at a lower amperage. If you stop charging a battery at 20 or 30 Amps compared to stopping at 2 to 5 Amps (at the same voltage), you will be at a lot lower state of charge when you're done. Very inconsistent.
@@SpeakerKevin #1 I beg to differ, danger is defined by the following simple math: 55.2 - (3.25 * 15) = 6.45V the voltage one cell will get to if the others have not become saturated and started to exceed nominal voltage. That's fried/toast territory. #2 Not true if he maintains *any* voltage higher than nominal for an extended period as proven on this channel rather conclusively.
@@SpeakerKevin To be perfectly clear, the voltage for a 16s pack you have to charge at to be 100% sure to not overcharge any cells is defined by (3.25 * 16) + 0.4 = 52.4 - any higher is at risk of damage without appropriate balancing techniques and equipment.
@@fredio54 His delta voltage between cells is a mere .005V. Plus he has a fully programmable BMS that will shut down the charging if any cell gets outside of a safe voltage.
Can you help me ??
Hey Andy, there are 2 cells in 48v pack that are too high and shuts off the BMS. One video of yours showed you lowering a cell with light bulbs or something. How did you do that?
I just use some car 12V light bulbs and connect them to the cells which are getting too high to discharge them a bit until all other cells have caught up.
Passive Balancing is slow & for Low Capacity cells. Typically below 2A to 10A for large capacity cells, Transfers from Hi cells to Lo Cells.
3.65V per cell is the "Max Allowable" for LFP. They always settle to 3.55 within an hour and that IS the 100% point in reality. Remember Nominal os 3.200V per cell and full working power-curve is 3.000-3.400.
EVE SPECS INFO:
4.2) Standard Charge
The standard charge means charging the cell with charge current 0.5CA and constant voltage 3.65V at (25±2)℃, 0.05C cut off.
4.3) Standard Discharge
The standard discharge means discharging the cell with a discharge current 0.5CA and cutoff voltage 2.5V at (25±2) ℃. If required, the battery can be discharged at 1.0CA constant current to a cutoff voltage of 2.5V.
EVE 280AH Cell: (280 X 0.05C) = 14A
HERE are Settings used with a Midnite Classic Solar Charge Controller for LFP Charge Profile. A QNBBM Active Balancer installed on all LFP Packs with the BMS.
All equipment MUST BE Voltage Corrected & Calibrated (VERY IMPORTANT)
- Divide Values X2 for 12V. Multiply X2 for 48V.
Absorb: 28.2 for 15 minutes (3.525vpc) (some call this boost)
Equalize: OFF
Float 27.9V (3.4875vpc)
MIn Volts: 22.0 (2.750vpc)
Max Volts: 28.7 (3.5875vpc)
Rebulk Voltage: 27.7 (3.4625vpc)
End Amps: 14A (*1)
This get's the bank charged to full with high amps (Constant Current) and then float (Constant Voltage) tops off so the cells are on average between 3.475-3.500. I am running 7/24/365 so float is used up by the Inverter + provides whatever the packs will take to top off.
(*1): End Amps is calculated from the {Highest AH Battery Pack} in a Bank. IE: 200AH X 0.05 = 10A 280AH X 0.05 = 14A.
** Coulumbic Efficiency for LFP is 99%
WHY are your cells getting so out-of-balance that you actually need 2 Amps of Balance Current ?
I charge to 100% , turn off charging until 90% at the Same day and only then recharge. It is not good or healthy to keep the cells above 3.45V for a long time. But i think victron does not allow such a setting Based on SOC
Why are you charging above 90% SOC ?
Does your BMS balance on float ? certainly my Daly BMS will only balance while the battery is charging. Having said that I agree with your voltage settings. I believe your BMS software has the option to balance when not charging. It would be interesting to see that in action.
Yes, it will start on the set balancing turn on voltage when the voltage is going UP and will continu as long as the voltage is above that and current is being delivered. I think the approach Andy takes is not working, but he will have to find out himself.
@@HansKeesom The key phrase there is "current being delivered". On float there is very little being delivered, only enough to maintain battery losses, so balance would not work.
@@muddy11111 Balancing during Float Charge does *NOT* not work on your BMS. Other BMS may balance as long as a Cell voltage is above a setpoint ...
Hi there i follow you with great interest in the U.K.At the moment i run a 15kw powerwall made of laptop batteries and very nice to .But i want to go for brand new batteries so i watch your youtube every episode.Can i ask one thing.In the summer in uk(HA HA) my batteries are fully charged by 10 oclock in morning(30amp).MYy question do you bring your batteries up to float conditions so the battery can absorb all day at low amps.I am still trying to get this float thing lithium ion rise to voltage and the amps drop off a cliff the last 10% takes about 1 hour then .1amp forever.Is this the same with lifep04??.I love the flat curve(charge/discharge) my inverter can do 60v so i am planning to run 17s @ 3.45V So the float will be about 9hrs???
Bulk and Absorb charge to 90% SOC and then you can try Float Charging the LiFePO4 battery bank at approx 3.35 Volts per cell or whatever voltage pushes ~0.1 amp into a 100AH cell ...
...and now everyone is changing their values...
Everyone should use voltage that work best for them. It is Capacity (amp-hours) vs # of Cycles vs Calendar Aging trade-off fro everybody. Nobody should blindly change their voltage just because this *might* work for Andy. This is *NOT* an off-grid home ...
@@MrSummitville I could not agree more with you.
You worry too much - go have a beer :)
I think you are onto something here, but I would (for science purposes) run it like you said for 2-3 months, and then charge it to 3,6V per Cell and see how far they drift apart.
Just to see what happens.
Why does he need to charge to 3.6 Volts? Do you balance your car tires at 150 MPH? Nobody cares that their tires would be unbalanced at 150 MPH.
@@MrSummitville To make sure that the cells dont drift apahrt SOC wise. I mean you would not notice it by the cell voltage in the middle of the charge curve.
Why don't you take Jeremy's advice and only charge to 54.4V (3.4V/cell). You could use a longer Absorption time for balancing, or better yet, set the tail current at 1-2 Amps. This should give you a resting voltage of around 53.9V/3.37V.
I am now confused about the balancing process. I have a Daly BMS and it only shows that it is in balancing mode when the BMS is in charge mode. Once the BMS is switched off because it reached the designated charge voltage, there is No balancing. So I wonder if balancing is not only to reduce individual cell voltages or does it also increase those cell voltages which are too low? Danke
The way I understand BMS is it always lowers the highest cell down to the lower ones voltage via mosfet. It will continue to do that until all of the cells are charged to your set point.
How can the BMS balance your cells with say 30mA when you still charge your cells with 5A at the same time. That does not make any sense...
Only an active balance will decrease voltage of cells which are too high and transfer this energy to cells with lower voltage. Normal BMS don't do that, they just burn off the energy through a resistor.
I would imagine that different manufactures configure their balancing act many different ways. I’m also willing to bet that some do a piss poor job at system logic and can only come up with a sub-par product. I would consider that Daly BMS a shitty design if it only balances (or tries to balance) while charging only. It will never achieve its purpose in my view if that is the case. I agree with Andy, this makes no sense. Must be a design flaw.
@@Do_the_Dishes what if it is actively balancing and instead of depleting high voltage cells, it is supplying the lower cells with additional charge current?
@@OffGridGarageAustralia An Active Balancer might be able to reduce the voltage of one cell, if and only if, the Active Balancer can transfer *more* the 5 Amps of charging current, many (most?) do not. And if it is a Switched Capacitor type of Active Balancer and it is rated at "5 Amps", then it will probably only average 2.5 amps transfer between the highest & lowest cell.
BUT ... Cell Balancing, whether Active or Passive, should *only* occur during the end of the Absorb Charge or during Float Charge when the charge amps are LOW ...
i dont think scc are that smart from what i have scene the batterys only offers current if the solar is not producing enogth power to run the load so then takes power from the battery untill solar comes into sun again and can then power the load and the charge the battery this will happen with out a charge controller
🐸
i myself like active balancers as these will start working any time the battery is 30mv diference and takes power from the higher battery and dump it into the lower battery i feel this is less wastefull on power then dump load balancers
A 30mv difference between adjacent cells or 30 mv across the whole pack? What Mfr & model of A.B. does 30mv across the whole pack? Andy's battery bank only has a 5mv difference across the whole pack, and it only operates near the end of the charge cycle = very limited time.
@@MrSummitville active balancer work all the time no matter what voltage andy's cells are still knew i suspect as they get older they will start to drift more
the active work between cells if one is 30mv lower then over this could even be 10mv im just going off the one i have it will start taking power from the higher cells and putting it into the lower charged cell and thay will trigger no matter the point of charge it just sees the set voltage diference and starts to balance
Hi Andi
Wie immer super Video
Schau dir mal das Baterium BMS an bis 7 A mit Kühlung und 2 A ohne Kühlung und swipe mal durch das Video th-cam.com/video/m2MuEpuMSiw/w-d-xo.html
Grüsse von der Schweiz
Danke Dir Daniel!
Weiss ich, habe ich mir schon alles angeschaut. Die sind ja gleich hier um die Ecke. Super teuer und keine Sicherheitsfunktionen. Hmmm, komme da immer noch nicht dran...
balancing is/should not discharge a cell that is high, it is/should bypass it partially with 0.2 A. it start that at 2,7 to have enough time to do this. The balancing turn on voltage is when the voltage is going UP, not when going down, so I fail to see how your idea at 12:02 is gonna work.
At the end you have failed logic. You moved the location of what you consider 100% charged, but that does not mean you have the same capacity/energy of wh's stored. It's like 95% is the new 100%, hurrah.
If you have less drift, that does not mean the balancing is working just as good, it just need to do less work. So that is in the case of this batteries a good reason to go for a lower top voltage. So instead of reducing the absorption time, it would have made more sense to lower the absorption voltage.
Starting the Cell Balance process at 2.7 volts is way too low ...
Hold the f up where u been you been missing for a few days
So, what is "wrong" with enabling / starting the Cell Balancing process during the Bulk Charge mode, when any cell goes ABOVE the 3.351 Volt set point ? Nothing is wrong, you just "claim" it wrong. In fact, it is good thing to enable / start the Cell Balancing as soon as any cell goes above 3.351 volts in Bulk Mode, to obtain *maximum* Balance Time. So, I do not see WHY you think is better to *DELAY* the start of Cell Balancing, long after the cell voltage has raised *ABOVE* 3.351 volts. This video, your graph & your "logic" make no sense what-so-ever ...
Hard to say what is right and what is wrong. If someone can explain, what does it mean, if several cells run up during charge and go lower at rest or discharge...
I think, becose of the slightly difference of internal resistance, and/or capacity of cells during charge or discharge using I=U/R we can explain the deviation during charge higer then 3,45. In my case, I always have 3 cells, which always runs up on 0.020v at 45A and 27,8v ( after 3,4v until 3.55v) and daly bms slightly slow down there run-up. And when charge is on finishing line, voltege increased above 28,2 and current starts to go down my Neey balancer starts the job, that 3 quiqly goes down under 3,45, and other 5 cells that was lower on 0.020v becomes higer on 0.050v and goes up to 3,630v! at high battery voltage and decreasing current. So that 3 cells, which looks higer becomse lower and look like not fully charge... So, if would set start voltage at 3,35v for Neey 4A balancer, it will discharge that 3cells, and in the end the they become not fully charged, and it will look like 5 cells at 3,612 and 3cells are still at 3,437v. I described my real situation. And for balancing I leave it in float at 3,588v for several hours to get deviation at 0.002v and then tern of. Iven ufter that, after rest, that 3 cells looks lower on 0,002v...
Maybe nex time I'll do experiment and try to charge to 3,600v all battery and then separately that 3 cells one by one with automatic 3,65v Lifepo4 charger to 3,65 and some absorbsion. Maybe that 3 cells become fully charged. But maybe in this case they could run up much more quiker and bms and balancer would discarge tham, and again.... 😞