Balancer efficiency is around 95.6%. Look at the moment 16:30. Final energy is 3591 + 1538 + 3622 + 1516 = 10267 mAh. Initial energy is 5391 + 5344 = 10735mAh. Efficiency is 10267 / 10735 = 95.6%
The reason a active balancer is good to use to equalize cells is as they are under load or charging there is a variation in voltage per cell and this is where it equalizes the cells over time do same test but do a mild discharge and recharge cycle around 5 times I bet the cells are almost identical on capacity on theyre capacity on final test them individually and be prepared because this is where these active balancers shine.
Balancing only when they are nearly full or nearly empty. This is the reason why we usually use top balancing. When the cells are nearly full you can do balancing. Balancing is useless when you do the balancing in an early or mid state of charge or discharge.
@@OffGridGarageAustralia I told you ; drive with a raspberry only when you are on top or bottom balance ;) You have an input "run" on the balancer, with a little relay drived by the rasp when soc is >95% or SOC < 5% (or the cell voltage if you have), drive the activ balancer :) That's what i do (I have an arduino which communicate with my BMS and know the cell voltage for each cell), and work perfectly
enthusiast from Philippines here..great content..haven't thought of that with my 32650 cells from China..always looking forward with your next test Andy..keep it up!
That was an eye-opener, It's great to see the different chemistries act in different ways. Don't forget to test cells 2 & 4 to see if they were damaged from the 5C tests because this would greatly impact those last tests. Thanks
Brilliant use of the balancer to demonstrate LiFePo battery state of charge differences at a given voltage (in the flat range), additionally demonstrating that balancers only help at the top end of charge. BTW I charge my battery up almost full, then plug in the balancer for a few days. After that, I unplug it until I think I need to check and balance the cells again. Thanks for this video-very engaging!
Think about this another way. Without charging you recovered ~30% more capacity. And if you left it to discharge you would probably get close to 45% before a bms would trigger low cell voltage.
From what I know, I think this balancer is still the best available. The problem with this verse a "BMS" with this balancer you still need a high voltage and low voltage cut off system beyond your charge controller settings and your inverter settings. I am working this problem before I put my LiFePO4 cells into service, but just not there yet. Andy, I hope one of your goals near the end of all this is to tell your subscribers the very best way to keep your cells balanced while protecting them. :)
Great test Andy, definitely proves a point regarding the flat discharge curve and balancing. You proved exactly what you hypothesized prior to starting, thank you for making the effort. It seems other people can't listen and understand what you were trying to demonstrate.
Yeah, I got the same impression sometimes with people not understanding that the reason of doing such testing is to better understand this chemistry and learn from that. Such tests have nothing to do with real world scenarios. Same with the 5C discharge test, I mean, who does that in reality (except of RC users)?
Well the one issue is that the balancer said it was still working -- so I think for this test not enough time elapsed. Like the SOC, the balance delta is also a curve with time and voltage delta being the variables. I'm all for increasing the data we have on the topic, but ultimately I question the utility of such info as we're not using these battery types to sit around -- we need daily cycling. I can't really see a use case for wanting to utilize these cells in a long-term storage situation where neither use nor storage of such a pack is viable. A load or a charge on any such system should be enough for a balancer to truly balance the cells as the difference in resistance leads to a difference in voltage and gives the balancer more meat to chew on.
@@OffGridGarageAustralia Well, you've actually turned into probably my favourite channel for your enthusiasm, testing scenarios and regular video output. I don't know how you get the time and determination to undertake the lengthy tests and deliver video content at the same time, but for me, you have me enthralled, waiting for each episode and my thanks.
I just recently added an active balancer to one of my 8S packs (using 100Ah LiFePO4 cells). Both packs were the same model and same supplier but one of them is quite well matched, usually ends up about a 25mV spread after 3.5V absorption (for about an hour) using the 50mA passive balancer in the BMS. The other pack was almost 10 times that spread, the active balancer brought that down to 5mV. Good stuff!
Hi Andy, I did a test just the other way....solar charging 8 cells and discharging..without bms or balancer....as you may remember I have also 3 faulty cells...they dropped tremendously in voltage ...one even down to 0.165 volt after having disconnected all cells and let them rest for 2 days. I repeated the test with 8 cells ...all in good shape....and they seem to stay close together....at least from the voltage that stated no difference after 3 days and reaching around 3.3 voltage while charging. Combining your results with mine verified that using lifepos only in the flat part of the charge/discharge curve will end up with a total difference in capacity of the cells. So what Victron does going from time to time into the steep part of the charging curve in a high soc is best practice for the sake of lifetime of a battery. So the problem could be that during winter time you should have a keen eye on the battery and try to get it up to a higher soc by not dragging too much energy out of it. I think it is a good idea to put all cells parallel from time to time and top balance them as a maintenance job. With a battery built yourself one can do that....but how can I use the battery best in my car..‽..how accurate is the measurement of the soc in the car ? Limits up to 80% soc are recommended...I think it may be good to charge up to 100 % from time to time with a lower current....and avoid charging with hpc if you got enough time to charge.
Lithium phosphate batteries like most batteries will undergo self discharge if left standing long enough. Consequently, the cells at lower state of charge(SOC) will fall out of the flat region of the curve faster than the cells at higher SOC. Meaning with enough time, enough voltage differential will be created to allow the cells at HIGHER SOC to top up the cells at lower SOC until all cells arrive at essentially the same SOC. The beauty of active balancing is that cells which might fall below the safe limit of 2.2 or 2.5V are kept within the safe voltage limits by energy tapped from neighbouring cells with that have energy to spare (higher SOC) . This to me ensures better health or longevity for the entire pack. Together the entire pack stare within the safe zone for longer.
Really good job there Andy. I believe most of us learn a lot from this video including myself. 1 1) Now we know the LFP voltage does not represent the state of charge especially around the platoue of he curve 2) Now we know where is the weakness of using the AB. But most of us agree that, teh daily solar system will have charge and discharge where at some point, the voltage of the cell has to reach bottom and the top where AB could do its job efficiently. Apart from this, like my previous comment, i would really like to see how much current would AB for let say 3 days even tough the voltage has been balance. There has been few occasion where i found out that AB would continuously to use the current even the cell are balance to one extend drain the cell. Again, THANKS A LOT for GREAT VIDEO.
Interesting test Andy. I’d be curious to see what would happen if you did a discharge with them in series with the balancer connected. As the empty cells got into the step part of the curve the balancer should start working again.
that balancer actually have superb efficiency. they do not burn the excess power and shuttle it around. They are really worth the money and if you program it somehow to just do balancing near the top and bottom curve they pretty much spot on!
If you charge or discharge the balancer will have more difference in voltage and get them closer. And as cells get closer to discharged or charged they will balance better. These balancers I think also shunt at 3.75v. So just charging at 100mA with 3.76v per cell will get all cells back to the same.
@@OffGridGarageAustralia Sure. But in the real world you never just have batteries sitting and doing nothing. Even a partial charge or discharge will help them balance a lot.
The missing 5% is interesting. Is it mostly the inefficiency of the balancer as a cost of heat of resistance or is it the electrical cost of the monitor? A little of each? It seems rather small actually and considering how much work the balancer had to do it would seem to indicate that the electrical cost of using the balancer is probably quite small in real world application.
Take these numbers with a grain of salt. The cell monitor is not that accurate and shows slightly different numbers every time I connect it even the voltage has not changed actually. The test was more to show how much difference we have in SOC at almost the same voltage.
LiGFePO4 charge efficiency is apparently about 96% (found that in a google search with a 2014 document IIRC ). So that would account for the bulk with some likely smaller I^2R losses.
just calculate the resistive losses on each cell + wiring losses + mosfet switching losses + mosfet RDSon losses + capacitor ESR losses I think that 5% is reasonable
When the battery pack is in use, the voltage is often outside the flat area of the charge curve. Over time, the active balancer will balance the battery pack. I also see the actual purpose of the balancer in keeping the cells balanced and not balancing new, completely different cells.
Nice to see the total Wh stays close from start to end. So in the end it will work for voltages. It will always help when you perform this as maintanance every now and then (with active charging to for instance 90%) so the total capacity will stay close.
1) V=IR, so if the voltage difference is small, this type of balancer can only produce a small current. 2) If you were cycling the cells, the different voltage sag under load at different SoC might improve the usefulness of the balancer compared to this static test.
You want to obviously catch the cells when they're not in that flat portion, namely charging and discharging. I think the biggest advantage of an active balancer is to augment a balance charger that has a limited discharge capability (which it uses to balance). So for instance if a balance charger can only discharge 200ma in order to balance, it's faster and more efficient to let an active balancer do that work. Caveat: the voltage difference threshold that a balance charger starts balancing is usually less than the threshold of these active balancers.
I really believe that the testing you do helps the rest of us think critically as we endeavor to make the most reliable system choices. Where I am at is using Smart Lifepo4 Battery Protection Board 300 amp BMS 48V 20s for Bluetooth in 16S with a 5A Active Equalizer Capacitor Balancer for each 16 cell bank. Hoping for more maintenance free situation as I move way from the 16 T-105 Trojan batteries I am so tired of babysitting. My question is, how much do the difference between the two types of BMS used together is beneficial for the batteries or diminish the functionality of their intended process for their individual purpose? Does the residence balancing of the Smart BMS diminish the capacitance reactance transfer between batteries through the Active Equalizing Balancer? Does the Active Equalizing Balancer interfere with the representative information of each cell the Smart BMS is trying to provide?
Thanks for your comment and great question. I have turned off the balance function in my BMS and let the balancer do all the work. The BMS balances only with a very small current (>200mA). The BMS is only for safety reasons to catch high or low voltages and disconnect in such an emergency. You can also set the BMS balancing to assist the active balancer if you set the kick in voltage a bit higher to only start if voltages are getting really high. You'll be interested in my next video coming tomorrow 😉
@@OffGridGarageAustralia Extreme thanks 😊. I still want to see you let magic smoke out of something. I think it would be great to see other mistakes that we can make that can let the magic smoke out, that we all want to avoid. I think there will be enough views to more then pay for the damage of say a BMS hooked up wrong, or something else, safely 😉. Anyways you have quite a following.
Balancing is an approach to solve the issue arising when some batteries are full, or empty and we cannot use the capacity in non full/non empty ones. Active balancing is just another approach that achieves exactly the same goal, in a different way. The only point of concern I can see is about the maximum current they can produce in these extreme situations,to compensate for non balances batteries.
Loved the video Andy....a very good test to show the way a lithium battery works. I was reading the comments and i see a lot of others putting out ideas and i wonder how many of them have done or will do a test of their proposed idea? Lots of talk but little to back up the ideas. I dont mean to criticize i am making an observation. This is why i like watching your channel. Opinion is put to the test.
So if we could program a Raspberry Pi to turn on the balancer when the highest charged battery hits 3.55V, wouldn't it slow down the top charged batteries and allow the lower charged ones to catch up?
Great video as always Andy. I have just installed an active balancer in my pack only an hour ago as my cells were drifting a bit like you pack did only a few weeks ago. Where I think it will help is the end of the charge curve and the end of the discharge curve. I programmed my bms to stop charging if a cell reaches 3.35v however there is a big deviation of cell voltage when nearly full. This should help it to top balance and hopefully stop the bms from cutting out too early if I have a weaker cell whilst discharging. For a few quid I think it is worth it. Over the last hour my deviation was 0.14, after only an hour it is 0.10 and my pack is accepting a charge again.
I’m surprised no one has mentioned that you really need to re-test the capacity off the two tortured cells as, if their capacity has been substantially reduced, the percentages may be closer. Certainly though to actually balance they would need to be into the knee portion of the curve.
Do you know what is the maximum resting voltage (after 5 hours rest as for an example) you can get from these batteries that would indicate a 100% full one (because at the end the curve is not flat)? Or do the maximum always go back during resting at the end of the flat part of the curve ? (sorry for my English, it is not my native language).
Andy, it is obvious that the active balancer needs no controls, it only works when the voltages are substantially different at either end of charge or discharge. In fact with an active balancer if you have reliable charge and discharge voltage control cut -off the only purpose for a BMS is to monitor individual cells. So is the BMS really needed? Dan
So next test, a 3 month test(harsh winter conditions)- basicly charge all batteries to storage capacity around 66%(if I remember correctly), connect the active balancer, and see how much it eats up in passive-/sometimes-active-idle usage, by balancing an unused battery - aka: Should we in the cold dark north not leave the balancer connected, when there's no expected charge coming in.
If you don't charge or discharge and just 'park' your battery, there is no need to connect anything to it. Connect the balancer and BMS only if you charge/discharge. around 60% is the best SOC if you don't use the battery for a while, that's correct.
I second this! I am looking to put LifePO4 battery in our remote Northern cabin which sits empty for months at a time and several of those months below freezing.
@@OffGridGarageAustralia yes, but that means having to connect and disconnect the balancer for the seasons, it would be great if we could just leave it on (if I recall, there was a on/off jumper, I guess that would at least allow for remote controlling it) - but maybe this is more of a balancer issue, than a battery issue.
Valid that the balancer will not equilze when connected as you did. But if the battery is being cycled, cells 1 & 3 will hit the knee and dump the excess to 2&4 which will equalize them over several cycles. I think that this problem you point out is only at the mid charge point with static unbalanced cells but in real world cycles, the balancer will equilize as it cycles.
You should try now an equal test with the same starting conditions, but with the balancer connected while the full pack is being discharged all togheter, and see if the sum of the capacity is equal to the sum of two single full cells
Kirchhoff was right, but how about Mendel? Therefore these balancers seem to have a contact to be only switched on occasionally, i guess. But on the other Hand, what would be, if the balnacer balances all the time? You never know the SOC in the flat area of the curve, with or without balancer. And if you leave the flat area, it might be easier to keep the cells close to each other, where they need it? So for me, this test showed clearly the different SOC of cells with the same voltage, but seems not to have much practical relevance. What do you think about turning on the balancer in case of a deivation greater than 200mV in total or one ore more cells are above 3.48V? Will you retest the capacity of the tortured cells? And it is still summer here... My Polo showed 42°C today in Yambol. How is your battery rack going? Gut's Nächtle
Hi Andy... In same setup (1&3- full, 2&4- empty, active balancer connected), please perform a capacity test(charge with 14.0V & discharge upto 12.0V). That will be very interesting 🙂🙂🙂. Will wait for it.
Each balance leads would need small variable boost-converter so that the lead that charges empty cell would allways have like 10 mv more than the lead that is charged. That way the charge capasity could be pumped so that it would become closer to each other. In theory if that 10mv difference could reach to -1mv that would mean that now the lower charge cell would have turned to highter charge cell.
The capacitor pump balancer may not produce any additional overpotential for balancing. The results are similar to just paralleling cells and letting them passively balance, which will also not equalize cells. An active balancer with a DC-DC inductor boost circuit does provide the over-potential addition. So don't conclude your results happens to all active balancers. You also threw in some 'fog' with the two abused cells. You have a battery impedance meter. Why don't you use it? It gives informative numbers beween 20% and 85% state of charge to help show if you did significant damage to the 5C discharged cells. At 5C discharge and high cell current you most likely damaged the SEI layer and degraded the electrolyte. SEI will regrow on next few recharges but it consumes some of available lithium, reducing cell capacity and increasing cell Rs due to thickening of SEI layer. Electrolyte degradation will increase Rs. It will produce some coating of broken down chemicals in electrolyte that will coat the anode and cathode surface interface to electrolyte reducing lithium ion migration paths, increasing cell Rs..
Hola Andy tienes que hacer un video con las diferentes formas de hacer un top balance ya que es un proceso que mucha gente no sabe y la importancia de esto depende de que el usuario disfrute de sus baterias sin complicaciones
Yes, I' have the balancer on the big battery since I installed it a few videos back. A balancer and a BMS can absolutely run in parallel. Why do you think they can not?
@@OffGridGarageAustralia Because the bms also balances as well as protects the cells. I thought they might conflict somehow. Iam not that knowledgeable in this arena so I ask before I just do something. I have learned a lot from you and your channel. Many thanks
@@OffGridGarageAustralia I was glad to see you post this... It would be great if you could reiterate this a few dozen more times as there are many channels that treat them as an OR instead of an AND. Looking forward to more tests. Also, you can probably get your hands on some LTO batteries? I would love to see you introduce the World to them.
I thought it was cummon knowlegde that balancing and equalising can only be done at high SOC or low SOC, preferably while charging. After seeing your video from the balancer on the big battery i ordered my 3 21S equalisers. to assist my BMS to balance these huge cells. without them it took 4 or 5 topups (days) to equalise all the cells on top, but they will go out of balance every time i drain them down to 15% or lower. Because i use used cells, this is expected. new cells would not need that much balancing. REALY looking forward of wiring these up though... 54 wires to extend and cablemanage ... yihaa
You should get yourself a current probe and an oscilloscope. Those balancers do not work with a constand flowing DC current, but they handle the capacitors as intermediate storages for energy, and pull/push charge energy to and from the cells. Measuring with a DC current probe does not give you the correct picture. The balancer operates around 5 KHz swapping energy.
Got your reply but it does not show here due to this fantastic TH-cam bug: Knut H. Ødegaard replied: "Off-Grid Garage Hantek have some good and not too expensive probes!" Thanks, I'll have a look at this.
Ok answer me this: When Andy's measuring the amps going through the balancer wires, some are positive amps, some are negative amps, depending on which cell is connected. How does that clampmeter know which way the current is flowing? It's a single wire!
The direction of the magnetic field surrounding the cable is dependant in the direction the the current flows through the cable. A 10s explanation: th-cam.com/video/9p3t9NOfCtA/w-d-xo.html
Very very nice Video. Thnak you ♥️ But what will happpen if u charge them now with the equalizer connected. Do they have the same capacity then? Especially with a charge current of 1C and 2C...
Andi danke danke... Ich hatte das schon vermutet. Meine Tests verliefen ähnlich... Wenn aber dann eine Zelle von den Volt her gesehen, zusammenbricht. Dann lädt der Balancer quasi gleich wieder nach. So lange deine Entnahme nicht zu hoch ist, ist das auch nur ein kleineres Problem... Steigt der Verbrauch, kommt der Balancer nicht hinterher. Der wirklich wichtige Part , liegt meiner Meinung nach zwischen 0 und 15 Prozent und dann zwischen 90 und 100 etwa... Da geht richtig die Post ab... Pete von Australien hat da mal nen schönen Stresstest gemacht.. Kennst du seinen Kanal?
Not with a balancer and LiFePO4 cells, no. When they meet in the flat part of the curve, there is not enough voltage difference to keep pushing current and the balancer will turn of. The cells will remain that far apart as I showed in the video. They are (voltage) balanced but not equalized.
Interesting. What I would like to know is how it would behave in a real world scenario where the battery is cycled between 11.6V (2.9 per cell) and 14V (3.5 per cell) several times. This would resemble the real usage of a solar system and would include parts of the steeper areas of the curve.
Well, I've got the big battery connected to the 16s balancer since I installed it and it works a treat! Haven't seen the cells been apart for more than 5-6mV since then.
@@OffGridGarageAustralia Same for me. But there is this guy who claims that equal voltage is not equal SOC.😎 So the question is whether this is also true in a "real" environment.
@@russellwilson5246 Not really. To do so I would have to test the capacity of some of the cells in my real battery. And this would render it useless for several days. Andy has the perfect vehicle to do the test with these small batteries without the need to use a production system.
could an active balancer create a problem when you slowly discharge your batteries it would be doing a bottom balance. let's say you have a weak cell that can only hold about 80% as the ones it's in series with. then if the Sun comes out and you quickly charge to full could the active balancer top balance fast enough to keep up with the imbalance it created?
Yes, that is the exact problem with the active balancer. It does not care and will just balance at any time if there is a voltage difference. I will have a video coming this week which explains that again in all details. I did a test over two weeks with one battery connected to a balancer all the time and the other one not. The result is undoubtable.
Hi sir discharge all the 4 cells . Then charge full and connect the balancer. The capacity must be similar. Because balances only balance voltage. Not the state of charge. To charge cell the voltage should be higher .
@@rattusfinkus yeah, that is great. I'm getting such a balanced as well soon, comes with Bluetooth and lets us change parameters and settings. Gives us more flexibility and we can then actually see how it works and which feels very charged of discharged.
@@OffGridGarageAustralia What I have also used it for was monitoring the cells in my 2p16s pack. I would check to see which pairs of cells were always at a high voltage and which pairs were always low. I then swapped a single cell out of the pairs with high and low pairs. Over a period of a few months I did this twice and now my pairs have a lower voltage difference.
But leaving it hooked up all the time is still a great idea.. Because when you are using load, you will charge and discharge the battery as well.. When you are at the bottom and the top of the voltage for the battery the balancer WILL work. When you are in the middle, it does nothing. Overall though it should both automatically top and bottom balance the cells when you are charging or discharging the battery. So in a solar example.. You will likely get down low enough that it helps balance just before the sun starts charging. And when you grt to the top of charge it will also allow top balancing. TLDR.. It hurts nothing when it is not working in the middle.. But it helps on both ends of the charge and discharge cycle. I'll bet you thr cells will equalize and stay equal if you put that setup through 5 to 10 cycles with the balancer installed than doing the same with the balancer not installed. That is the test you should do. Take your 60/30 set up.. Do 5 charge /discharge and full charge cycles with the balancer and go back down to the middle of voltage. Then check each cell capacity. Do the same test with no balancer. I'll bet you have better results on the balanced battery.
Active balancers are only useful if you need to use mismatched cell. They will help when a lower capacity cell has a higher/lower voltage on charge/discharge so the BMS will not shut off charge/discharge as fast.
I have the muller energy 250a and the app is different than your newer video. There is a setting called “Balance Settings” what voltage should I set the Equalization Voltage. Default is set at 3300 mv. Any recommendations
If only they made some other type of electronic device you could attach to batteries of various configurations that would actually balance the state of charge, not just the cell voltage. It could also address other critical operations, like preventing you from charging the battery with too much current, discharging too much current, or charging when the cells are too cold. Such a unit would manage all of these functions and we could call it a..."Battery Management System", "BMS", for short. :)
Aye but with mismatched cells it will divert charge into the higher capacity cell near the end of charging as the curve becomes more steep. On Discharge it will divert power into lower capacity cells as the curve steepens. You’ll still get more power out of a mismatched battery with an active balancer no?
Darn that voltage drop that you get after after you fully charge and the drop under load a li battery-that's why balancing should be happening with everything going in the charging direction (battery under charge)=still waiting for the another test on cells based on the capacity you found, 5ah that you did not discharge at 5C based on the capacity of 7.2ah
I have the cheap little $12 I think balancer for my 4s 10p headway pack. It does keep them close but not perfect. Most I've seen it work was 1 amp even. Frustrating part is wondering how efficient they really are and your right, do I really need to leave it on? And like you proved am I losing ah per cell? I check them daily, they are very close. Maybe I'll unplug it for a day and see. Anyway thanks for your excellent content. Very informative. I'm very new to lithium just one of those annoying car audio nuts.
oh Andy .... you have already proved you cant balance in the mid voltage range. Results are no surprise, you know to balance it has to be above the top "knee" (or below the bottom knee). When i saw "equalisation" I thought that you were going to do the balancing on a once a month equalisation voltage rather than your usual moderate - and use that old FLA MPPT setting in conjunction with the balancer - you did suggest this before. By the way, summer here in UK, I'm on summer MPPT setting 3.35/cell and I'm floating from 66% charge again due to the very flat voltage curves. But at least the bank isnt sitting at high voltage all day - still more than enough power to live and power everything i need too.
very interesting...you do good videos..thank you...i wonder if we conected them in parelel would the same thing happen or would it balance voltage and energy? i learn from another of your videos that absorption time counts in terms of energy...i wonder if the parelel method might give more absorption time becaus it does not switch off..also perhaps time with a parelell method might bring the energy into balance?
I'm guessing for LiFePo only charging the cells all the way up (to the upper limit of the BMS), then putting the balancer on. Rinse, wash, repeat for wildly divergent SOCs.
You can solder those tabs, just rough them up with sandpaper first. I think If you left the balancer connected until the balance currents equalled zero, the cells would be very close in charge at that point. Because of internal resistance, when there is balance current flowing the voltages will not correctly correspond to charge level. Instead of proving that balancing these cells does not work, I think you just proved that you didn't balance them for long enough to complete the task.
FWIW, Energy (Whrs) is not linear proportional to Ahrs. SOC is not representative for energy content. Do the same test measuring Whrs.. The result should be slightly different.
Let's now start with 2 fully charged and 2 empty, leave the balancer connected and then let this combination as a complete 4S 12v battery do a full charge and then a full discharge cycle. This will tell us how the balancer works In a Real life senario.🤓
Great video Andy,congratulation!This video came to meet the problem I am facing now. I received the 32 302 amps batteries which i was waiting for and started the top balancing procedure in three steps, 3.4,3.5 and 3.65 volts. I connected all 32 batteries in parallel and I got a huge bank of 3.2 volts and 9600 amps. In the first phase I charge the batteries with 3.4 volts and the bank sucks a current of approx. 15 amps. Assuming the batteries came with 50% SOC, this first phase will take about 12 days. Please advise me what to do, wait until the battery bank reaches 3.4 volts and the current close to 0 for the first step, or charge them directly with 3.65 volts. The source I use can charge a maximum of 60 amps.
If they are all in parallel, you can just set the charger to 3.6V and let it connected until the amps are almost zero. That is a good start for connecting them in series afterwards.
@@OffGridGarageAustralia Thank you Andy!I saw on some sites from the guys in the marines that they use this procedure, in three steps, but it takes too long, so I will listen to your advice. I thought that this procedure, in three steps is better for battery health.
Please Andy help me with an opinion or an idea. I connected the 32x302 Ah cells in parallel 4 days ago. Until yesterday I charged them at a voltage of 3.4 volts, and the voltage did not increase more than 3.32 volts from 3.29 as they had when I received them. Yesterday I increased the voltage from the source to 3.6 volts, the current consumed from the source is around 28-30 amps and the voltage is around 3,324 volts.I am worried and I'm afraid there is something wrong with the batteries because the voltage does not increase. Maybe because the battery bank is very large, 3.2 volts / 9600 amps, this top balancing procedure will take a very long time?And it is possible to calculate this time?
@@mihaitaiosub totally normal😊 If you charge 32 cells with 302Ah in parallel with only 30A. That will take: 32x302Ah/30A=322h charging time or close to 2 weeks. Set the charger to 3.6V and let it charge....
@@OffGridGarageAustralia Thanks for the reply. Starting from the fact that maybe they came loaded at 50%, maybe I'm lucky to last only a week. I'm worried that the voltage doesn't increase ay all.
Hi there. Interesting test. However, not fair for the active balancer. Your test could bring the capacity closer together given time. Try your test again with the active balancer connected to all the cells. IE. During charging the 4 cells in series, the active balancer keeps the SOC of all cells together. During discharge, Again the active balancer keeps the cells SOC together during discharge. One thing that I have noticed in your testing is that you are not allowing the battery to become fully charged. The batteries equalize best when they are charged to saturation of 3.55 volts or 3.60 volts. Be aware that your DOD of 2.5 volts isn't doing your batteries any favors. My experience is that if you keep your batteries at the lower cut off point that your performance will decrease more aggressive over time. If you choose a voltage closer to 3.0 volts your batteries will last longer. I'm using active balancing here on Li-Ion cells. My system is running at 28.8 volts max upper and 24.5 volts lower max on discharge. Keep smiling 😃
However this does not reflect a realistic use case. In reality you'll be using your batteries in the 20-80% range. This use/load will cause the lower charged cells to drop in voltage bellow the higher charged cells, thus the active balancer will have enough voltage difference to do it's job. Granted it might take longer, but in the end you'll eventually get equalized SOC.
Not a surprise. Energy != Voltage Power != Voltage Current != Voltage Flat voltage curve means you're not measuring anything but a near-meaningless voltage number. It's nothing but emergency cut off information at top or bottom.
![NISAMANEE - TADA (ท้าดา) [ OFFICIAL MV ]](http://i.ytimg.com/vi/hcQHEPMmJpU/mqdefault.jpg)
Balancer efficiency is around 95.6%. Look at the moment 16:30. Final energy is 3591 + 1538 + 3622 + 1516 = 10267 mAh. Initial energy is 5391 + 5344 = 10735mAh. Efficiency is 10267 / 10735 = 95.6%
The reason a active balancer is good to use to equalize cells is as they are under load or charging there is a variation in voltage per cell and this is where it equalizes the cells over time do same test but do a mild discharge and recharge cycle around 5 times I bet the cells are almost identical on capacity on theyre capacity on final test them individually and be prepared because this is where these active balancers shine.
Great suggestion.
Balancing only when they are nearly full or nearly empty. This is the reason why we usually use top balancing. When the cells are nearly full you can do balancing. Balancing is useless when you do the balancing in an early or mid state of charge or discharge.
Yep, exactly.
@@OffGridGarageAustralia I told you ; drive with a raspberry only when you are on top or bottom balance ;) You have an input "run" on the balancer, with a little relay drived by the rasp when soc is >95% or SOC < 5% (or the cell voltage if you have), drive the activ balancer :) That's what i do (I have an arduino which communicate with my BMS and know the cell voltage for each cell), and work perfectly
enthusiast from Philippines here..great content..haven't thought of that with my 32650 cells from China..always looking forward with your next test Andy..keep it up!
Thank you!
That was an eye-opener, It's great to see the different chemistries act in different ways. Don't forget to test cells 2 & 4 to see if they were damaged from the 5C tests because this would greatly impact those last tests. Thanks
Brilliant use of the balancer to demonstrate LiFePo battery state of charge differences at a given voltage (in the flat range), additionally demonstrating that balancers only help at the top end of charge. BTW I charge my battery up almost full, then plug in the balancer for a few days. After that, I unplug it until I think I need to check and balance the cells again. Thanks for this video-very engaging!
Thank you Stephen. I'm glad it was useful and you got the purpose of the video right away 😉
Your enthusiasm makes my day better...
Think about this another way. Without charging you recovered ~30% more capacity. And if you left it to discharge you would probably get close to 45% before a bms would trigger low cell voltage.
From what I know, I think this balancer is still the best available. The problem with this verse a "BMS" with this balancer you still need a high voltage and low voltage cut off system beyond your charge controller settings and your inverter settings. I am working this problem before I put my LiFePO4 cells into service, but just not there yet. Andy, I hope one of your goals near the end of all this is to tell your subscribers the very best way to keep your cells balanced while protecting them. :)
Great test Andy, definitely proves a point regarding the flat discharge curve and balancing. You proved exactly what you hypothesized prior to starting, thank you for making the effort. It seems other people can't listen and understand what you were trying to demonstrate.
Yeah, I got the same impression sometimes with people not understanding that the reason of doing such testing is to better understand this chemistry and learn from that. Such tests have nothing to do with real world scenarios. Same with the 5C discharge test, I mean, who does that in reality (except of RC users)?
Well the one issue is that the balancer said it was still working -- so I think for this test not enough time elapsed. Like the SOC, the balance delta is also a curve with time and voltage delta being the variables. I'm all for increasing the data we have on the topic, but ultimately I question the utility of such info as we're not using these battery types to sit around -- we need daily cycling. I can't really see a use case for wanting to utilize these cells in a long-term storage situation where neither use nor storage of such a pack is viable. A load or a charge on any such system should be enough for a balancer to truly balance the cells as the difference in resistance leads to a difference in voltage and gives the balancer more meat to chew on.
@@OffGridGarageAustralia Well, you've actually turned into probably my favourite channel for your enthusiasm, testing scenarios and regular video output. I don't know how you get the time and determination to undertake the lengthy tests and deliver video content at the same time, but for me, you have me enthralled, waiting for each episode and my thanks.
@@MrHantu666 great feedback, thank you very much. I'll go and shave my legs now and have a 🍺
😉
I just recently added an active balancer to one of my 8S packs (using 100Ah LiFePO4 cells). Both packs were the same model and same supplier but one of them is quite well matched, usually ends up about a 25mV spread after 3.5V absorption (for about an hour) using the 50mA passive balancer in the BMS. The other pack was almost 10 times that spread, the active balancer brought that down to 5mV. Good stuff!
Hi Andy, I did a test just the other way....solar charging 8 cells and discharging..without bms or balancer....as you may remember I have also 3 faulty cells...they dropped tremendously in voltage ...one even down to 0.165 volt after having disconnected all cells and let them rest for 2 days. I repeated the test with 8 cells ...all in good shape....and they seem to stay close together....at least from the voltage that stated no difference after 3 days and reaching around 3.3 voltage while charging. Combining your results with mine verified that using lifepos only in the flat part of the charge/discharge curve will end up with a total difference in capacity of the cells. So what Victron does going from time to time into the steep part of the charging curve in a high soc is best practice for the sake of lifetime of a battery.
So the problem could be that during winter time you should have a keen eye on the battery and try to get it up to a higher soc by not dragging too much energy out of it.
I think it is a good idea to put all cells parallel from time to time and top balance them as a maintenance job. With a battery built yourself one can do that....but how can I use the battery best in my car..‽..how accurate is the measurement of the soc in the car ? Limits up to 80% soc are recommended...I think it may be good to charge up to 100 % from time to time with a lower current....and avoid charging with hpc if you got enough time to charge.
Great work sir You have my respect
Lithium phosphate batteries like most batteries will undergo self discharge if left standing long enough. Consequently, the cells at lower state of charge(SOC) will fall out of the flat region of the curve faster than the cells at higher SOC. Meaning with enough time, enough voltage differential will be created to allow the cells at HIGHER SOC to top up the cells at lower SOC until all cells arrive at essentially the same SOC. The beauty of active balancing is that cells which might fall below the safe limit of 2.2 or 2.5V are kept within the safe voltage limits by energy tapped from neighbouring cells with that have energy to spare (higher SOC) . This to me ensures better health or longevity for the entire pack. Together the entire pack stare within the safe zone for longer.
Self discharge in LFP cells is very very low. It's 6 times less than with other chemistries. Almost negligible.
Really good job there Andy. I believe most of us learn a lot from this video including myself. 1
1) Now we know the LFP voltage does not represent the state of charge especially around the platoue of he curve
2) Now we know where is the weakness of using the AB. But most of us agree that, teh daily solar system will have charge and discharge where at some point, the voltage of the cell has to reach bottom and the top where AB could do its job efficiently.
Apart from this, like my previous comment, i would really like to see how much current would AB for let say 3 days even tough the voltage has been balance. There has been few occasion where i found out that AB would continuously to use the current even the cell are balance to one extend drain the cell.
Again, THANKS A LOT for GREAT VIDEO.
Interesting test Andy. I’d be curious to see what would happen if you did a discharge with them in series with the balancer connected. As the empty cells got into the step part of the curve the balancer should start working again.
that balancer actually have superb efficiency. they do not burn the excess power and shuttle it around. They are really worth the money and if you program it somehow to just do balancing near the top and bottom curve they pretty much spot on!
Great Test Andy! 🤘
Thanks! 👍
If you charge or discharge the balancer will have more difference in voltage and get them closer. And as cells get closer to discharged or charged they will balance better. These balancers I think also shunt at 3.75v. So just charging at 100mA with 3.76v per cell will get all cells back to the same.
That's not why I did the test...
@@OffGridGarageAustralia Sure. But in the real world you never just have batteries sitting and doing nothing. Even a partial charge or discharge will help them balance a lot.
The missing 5% is interesting. Is it mostly the inefficiency of the balancer as a cost of heat of resistance or is it the electrical cost of the monitor? A little of each? It seems rather small actually and considering how much work the balancer had to do it would seem to indicate that the electrical cost of using the balancer is probably quite small in real world application.
Take these numbers with a grain of salt. The cell monitor is not that accurate and shows slightly different numbers every time I connect it even the voltage has not changed actually. The test was more to show how much difference we have in SOC at almost the same voltage.
LiGFePO4 charge efficiency is apparently about 96% (found that in a google search with a 2014 document IIRC ). So that would account for the bulk with some likely smaller I^2R losses.
just calculate the resistive losses on each cell + wiring losses + mosfet switching losses + mosfet RDSon losses + capacitor ESR losses I think that 5% is reasonable
When the battery pack is in use, the voltage is often outside the flat area of the charge curve. Over time, the active balancer will balance the battery pack. I also see the actual purpose of the balancer in keeping the cells balanced and not balancing new, completely different cells.
Nice to see the total Wh stays close from start to end. So in the end it will work for voltages. It will always help when you perform this as maintanance every now and then (with active charging to for instance 90%) so the total capacity will stay close.
Excellent presentation. Now i understand why balancers don't work with LiFePo4 cells!
This is very similar to my Nickel Iron cells a very flat voltage curve until you get to the end.
All Lithium chemistry batteries are like this.
Your experiment is always interesting to me. Good work.
1) V=IR, so if the voltage difference is small, this type of balancer can only produce a small current. 2) If you were cycling the cells, the different voltage sag under load at different SoC might improve the usefulness of the balancer compared to this static test.
now put the pack on charge and see if the balancer keeps the the same when the charger is at top to bring the cap up together.
You want to obviously catch the cells when they're not in that flat portion, namely charging and discharging. I think the biggest advantage of an active balancer is to augment a balance charger that has a limited discharge capability (which it uses to balance). So for instance if a balance charger can only discharge 200ma in order to balance, it's faster and more efficient to let an active balancer do that work. Caveat: the voltage difference threshold that a balance charger starts balancing is usually less than the threshold of these active balancers.
Very nice test!!! thanks from Brazil!!!
Thanks for watching!
I really believe that the testing you do helps the rest of us think critically as we endeavor to make the most reliable system choices.
Where I am at is using Smart Lifepo4 Battery Protection Board 300 amp BMS 48V 20s for Bluetooth in 16S with a 5A Active Equalizer Capacitor Balancer for each 16 cell bank. Hoping for more maintenance free situation as I move way from the 16 T-105 Trojan batteries I am so tired of babysitting.
My question is, how much do the difference between the two types of BMS used together is beneficial for the batteries or diminish the functionality of their intended process for their individual purpose? Does the residence balancing of the Smart BMS diminish the capacitance reactance transfer between batteries through the Active Equalizing Balancer? Does the Active Equalizing Balancer interfere with the representative information of each cell the Smart BMS is trying to provide?
Thanks for your comment and great question. I have turned off the balance function in my BMS and let the balancer do all the work. The BMS balances only with a very small current (>200mA). The BMS is only for safety reasons to catch high or low voltages and disconnect in such an emergency.
You can also set the BMS balancing to assist the active balancer if you set the kick in voltage a bit higher to only start if voltages are getting really high.
You'll be interested in my next video coming tomorrow 😉
@@OffGridGarageAustralia Extreme thanks 😊. I still want to see you let magic smoke out of something. I think it would be great to see other mistakes that we can make that can let the magic smoke out, that we all want to avoid. I think there will be enough views to more then pay for the damage of say a BMS hooked up wrong, or something else, safely 😉. Anyways you have quite a following.
@@xr680r Thank you. I'll see what I can do 😏
Balancing is an approach to solve the issue arising when some batteries are full, or empty and we cannot use the capacity in non full/non empty ones. Active balancing is just another approach that achieves exactly the same goal, in a different way. The only point of concern I can see is about the maximum current they can produce in these extreme situations,to compensate for non balances batteries.
Loved the video Andy....a very good test to show the way a lithium battery works. I was reading the comments and i see a lot of others putting out ideas and i wonder how many of them have done or will do a test of their proposed idea? Lots of talk but little to back up the ideas. I dont mean to criticize i am making an observation. This is why i like watching your channel. Opinion is put to the test.
Great point! Thank you.
So if we could program a Raspberry Pi to turn on the balancer when the highest charged battery hits 3.55V, wouldn't it slow down the top charged batteries and allow the lower charged ones to catch up?
Great video as always Andy. I have just installed an active balancer in my pack only an hour ago as my cells were drifting a bit like you pack did only a few weeks ago. Where I think it will help is the end of the charge curve and the end of the discharge curve. I programmed my bms to stop charging if a cell reaches 3.35v however there is a big deviation of cell voltage when nearly full. This should help it to top balance and hopefully stop the bms from cutting out too early if I have a weaker cell whilst discharging. For a few quid I think it is worth it. Over the last hour my deviation was 0.14, after only an hour it is 0.10 and my pack is accepting a charge again.
I’m surprised no one has mentioned that you really need to re-test the capacity off the two tortured cells as, if their capacity has been substantially reduced, the percentages may be closer. Certainly though to actually balance they would need to be into the knee portion of the curve.
This video is coming tomorrow at this time 😉
Do you know what is the maximum resting voltage (after 5 hours rest as for an example) you can get from these batteries that would indicate a 100% full one (because at the end the curve is not flat)? Or do the maximum always go back during resting at the end of the flat part of the curve ? (sorry for my English, it is not my native language).
Andy, it is obvious that the active balancer needs no controls, it only works when the voltages are substantially different at either end of charge or discharge. In fact with an active balancer if you have reliable charge and discharge voltage control cut -off the only purpose for a BMS is to monitor individual cells. So is the BMS really needed?
Dan
So next test, a 3 month test(harsh winter conditions)- basicly charge all batteries to storage capacity around 66%(if I remember correctly), connect the active balancer, and see how much it eats up in passive-/sometimes-active-idle usage, by balancing an unused battery - aka: Should we in the cold dark north not leave the balancer connected, when there's no expected charge coming in.
If you don't charge or discharge and just 'park' your battery, there is no need to connect anything to it. Connect the balancer and BMS only if you charge/discharge.
around 60% is the best SOC if you don't use the battery for a while, that's correct.
I second this! I am looking to put LifePO4 battery in our remote Northern cabin which sits empty for months at a time and several of those months below freezing.
@@OffGridGarageAustralia yes, but that means having to connect and disconnect the balancer for the seasons, it would be great if we could just leave it on (if I recall, there was a on/off jumper, I guess that would at least allow for remote controlling it) - but maybe this is more of a balancer issue, than a battery issue.
Valid that the balancer will not equilze when connected as you did. But if the battery is being cycled, cells 1 & 3 will hit the knee and dump the excess to 2&4 which will equalize them over several cycles. I think that this problem you point out is only at the mid charge point with static unbalanced cells but in real world cycles, the balancer will equilize as it cycles.
Nice one Andy! 👍
Thanks 👍
Andy I think you should try the same test as 1 12v battery but leave the balancer on and then check capacity
I certainly can.
You should try now an equal test with the same starting conditions, but with the balancer connected while the full pack is being discharged all togheter, and see if the sum of the capacity is equal to the sum of two single full cells
Kirchhoff was right, but how about Mendel?
Therefore these balancers seem to have a contact to be only switched on occasionally, i guess.
But on the other Hand, what would be, if the balnacer balances all the time? You never know the SOC in the flat area of the curve, with or without balancer. And if you leave the flat area, it might be easier to keep the cells close to each other, where they need it?
So for me, this test showed clearly the different SOC of cells with the same voltage, but seems not to have much practical relevance.
What do you think about turning on the balancer in case of a deivation greater than 200mV in total or one ore more cells are above 3.48V?
Will you retest the capacity of the tortured cells?
And it is still summer here... My Polo showed 42°C today in Yambol.
How is your battery rack going?
Gut's Nächtle
Video suggestion: What are the different types of “lithium” chemistries and how they differ?
That's on the way. Thank you.
@@OffGridGarageAustralia Yes, that will be a good one!
Excellent video as always Andy, I was wondering if I really needed to top balance my new 280Ah cells, this video shows that the answer is 100% yes 😁
Hi Andy...
In same setup (1&3- full, 2&4- empty, active balancer connected), please perform a capacity test(charge with 14.0V & discharge upto 12.0V).
That will be very interesting 🙂🙂🙂. Will wait for it.
Thanks Andy, Very helpfull
Each balance leads would need small variable boost-converter so that the lead that charges empty cell would allways have like 10 mv more than the lead that is charged. That way the charge capasity could be pumped so that it would become closer to each other. In theory if that 10mv difference could reach to -1mv that would mean that now the lower charge cell would have turned to highter charge cell.
Great content 👌 Thank you !
Thank you, Imad.
The capacitor pump balancer may not produce any additional overpotential for balancing. The results are similar to just paralleling cells and letting them passively balance, which will also not equalize cells. An active balancer with a DC-DC inductor boost circuit does provide the over-potential addition. So don't conclude your results happens to all active balancers.
You also threw in some 'fog' with the two abused cells.
You have a battery impedance meter. Why don't you use it? It gives informative numbers beween 20% and 85% state of charge to help show if you did significant damage to the 5C discharged cells. At 5C discharge and high cell current you most likely damaged the SEI layer and degraded the electrolyte.
SEI will regrow on next few recharges but it consumes some of available lithium, reducing cell capacity and increasing cell Rs due to thickening of SEI layer. Electrolyte degradation will increase Rs. It will produce some coating of broken down chemicals in electrolyte that will coat the anode and cathode surface interface to electrolyte reducing lithium ion migration paths, increasing cell Rs..
Been waiting all vid for what an equalizer does
Hola Andy tienes que hacer un video con las diferentes formas de hacer un top balance ya que es un proceso que mucha gente no sabe y la importancia de esto depende de que el usuario disfrute de sus baterias sin complicaciones
for holes in strip a hole punch works greats thats for doing belts or some used for putting holes in paper
Hi Andy great video any chance of doing video on your solar power output from your panels to see the daily and monthly outputs through the seasons 🤔
You are drilling those tabs? Guess they don't sell paper punches in Australia.
Not with 6mm, no.
@@OffGridGarageAustralia Mine are roughly 1/4 inch (6mm). How big are they in Australia?
Andi, you mentioned that you have a balancer on your large battery pack. Do you ? And can you run the balancer and a bms attached at the same time.
Yes, I' have the balancer on the big battery since I installed it a few videos back. A balancer and a BMS can absolutely run in parallel. Why do you think they can not?
@@OffGridGarageAustralia Because the bms also balances as well as protects the cells. I thought they might conflict somehow. Iam not that knowledgeable in this arena so I ask before I just do something. I have learned a lot from you and your channel. Many thanks
I have two of them coming, just found out they are arriving on the 30th. Your answer is just in time.😀
@@OffGridGarageAustralia I was glad to see you post this... It would be great if you could reiterate this a few dozen more times as there are many channels that treat them as an OR instead of an AND.
Looking forward to more tests. Also, you can probably get your hands on some LTO batteries? I would love to see you introduce the World to them.
I thought it was cummon knowlegde that balancing and equalising can only be done at high SOC or low SOC, preferably while charging.
After seeing your video from the balancer on the big battery i ordered my 3 21S equalisers. to assist my BMS to balance these huge cells.
without them it took 4 or 5 topups (days) to equalise all the cells on top, but they will go out of balance every time i drain them down to 15% or lower.
Because i use used cells, this is expected. new cells would not need that much balancing.
REALY looking forward of wiring these up though... 54 wires to extend and cablemanage ... yihaa
Yeah, let us know how you go with that. I'm interested to hear your experience with the used cells and balancer.
@@OffGridGarageAustralia Will do Andy, thanks for sharing your journey to enlightenment :P
You should get yourself a current probe and an oscilloscope. Those balancers do not work with a constand flowing DC current, but they handle the capacitors as intermediate storages for energy, and pull/push charge energy to and from the cells. Measuring with a DC current probe does not give you the correct picture. The balancer operates around 5 KHz swapping energy.
I actually have an oscilloscope... but I don't have a current probe for it...
Got your reply but it does not show here due to this fantastic TH-cam bug:
Knut H. Ødegaard replied: "Off-Grid Garage Hantek have some good and not too expensive probes!"
Thanks, I'll have a look at this.
Ok answer me this: When Andy's measuring the amps going through the balancer wires, some are positive amps, some are negative amps, depending on which cell is connected. How does that clampmeter know which way the current is flowing? It's a single wire!
The direction of the magnetic field surrounding the cable is dependant in the direction the the current flows through the cable.
A 10s explanation: th-cam.com/video/9p3t9NOfCtA/w-d-xo.html
Oh interesting! I had no idea a clampmeter could do that. Thank you!
It would be nice to see what would happen when cycling the battery.
Keep up the great work.
I have one question can you tell me how much those sells actually weigh
Very very nice Video. Thnak you ♥️
But what will happpen if u charge them now with the equalizer connected. Do they have the same capacity then? Especially with a charge current of 1C and 2C...
Very interesting vid, thanks.
Thank you.
Andi danke danke...
Ich hatte das schon vermutet.
Meine Tests verliefen ähnlich... Wenn aber dann eine Zelle von den Volt her gesehen, zusammenbricht.
Dann lädt der Balancer quasi gleich wieder nach.
So lange deine Entnahme nicht zu hoch ist, ist das auch nur ein kleineres Problem...
Steigt der Verbrauch, kommt der Balancer nicht hinterher.
Der wirklich wichtige Part , liegt meiner Meinung nach zwischen 0 und 15 Prozent und dann zwischen 90 und 100 etwa...
Da geht richtig die Post ab...
Pete von Australien hat da mal nen schönen Stresstest gemacht.. Kennst du seinen Kanal?
So would two full cells to one flat cell be more likely to balance capacities?
Not with a balancer and LiFePO4 cells, no. When they meet in the flat part of the curve, there is not enough voltage difference to keep pushing current and the balancer will turn of. The cells will remain that far apart as I showed in the video. They are (voltage) balanced but not equalized.
Interesting. What I would like to know is how it would behave in a real world scenario where the battery is cycled between 11.6V (2.9 per cell) and 14V (3.5 per cell) several times. This would resemble the real usage of a solar system and would include parts of the steeper areas of the curve.
Well, I've got the big battery connected to the 16s balancer since I installed it and it works a treat! Haven't seen the cells been apart for more than 5-6mV since then.
@@OffGridGarageAustralia Same for me. But there is this guy who claims that equal voltage is not equal SOC.😎
So the question is whether this is also true in a "real" environment.
@@keyem4504 ha ha i get your humor...you and eddie have the "same for me" situations...you too can answer the question you ask...its a good question.
@@russellwilson5246 Not really. To do so I would have to test the capacity of some of the cells in my real battery. And this would render it useless for several days. Andy has the perfect vehicle to do the test with these small batteries without the need to use a production system.
Good Sr. which Active balance do you recommend for a 4S lipo4 battery configuration cell capacity is 305 amp each.
Very interesting, thank you.
could an active balancer create a problem when you slowly discharge your batteries it would be doing a bottom balance. let's say you have a weak cell that can only hold about 80% as the ones it's in series with. then if the Sun comes out and you quickly charge to full could the active balancer top balance fast enough to keep up with the imbalance it created?
Yes, that is the exact problem with the active balancer. It does not care and will just balance at any time if there is a voltage difference. I will have a video coming this week which explains that again in all details. I did a test over two weeks with one battery connected to a balancer all the time and the other one not. The result is undoubtable.
@@OffGridGarageAustralia thank you. I always look forward to your videos
Hi sir discharge all the 4 cells .
Then charge full and connect the balancer.
The capacity must be similar.
Because balances only balance voltage.
Not the state of charge.
To charge cell the voltage should be higher .
Yes 👍 🔌 🔋 ⚡
I have may balancer set to turn on at a higher voltage difference say 30mv, this seems to stop the balancing mid curve.
How did you do that, what triggers it?
@@OffGridGarageAustralia my balancer allows me to select the trigger volt difference
@@rattusfinkus yeah, that is great. I'm getting such a balanced as well soon, comes with Bluetooth and lets us change parameters and settings. Gives us more flexibility and we can then actually see how it works and which feels very charged of discharged.
@@OffGridGarageAustralia What I have also used it for was monitoring the cells in my 2p16s pack. I would check to see which pairs of cells were always at a high voltage and which pairs were always low. I then swapped a single cell out of the pairs with high and low pairs. Over a period of a few months I did this twice and now my pairs have a lower voltage difference.
@@rattusfinkus Yes, that's a god solution, Colin. Thanks for sharing. You optimise your battery pack this way and it will be perfectly matched.
Thanks Andy ... very informative :-)
But leaving it hooked up all the time is still a great idea.. Because when you are using load, you will charge and discharge the battery as well.. When you are at the bottom and the top of the voltage for the battery the balancer WILL work. When you are in the middle, it does nothing. Overall though it should both automatically top and bottom balance the cells when you are charging or discharging the battery. So in a solar example.. You will likely get down low enough that it helps balance just before the sun starts charging. And when you grt to the top of charge it will also allow top balancing.
TLDR.. It hurts nothing when it is not working in the middle.. But it helps on both ends of the charge and discharge cycle.
I'll bet you thr cells will equalize and stay equal if you put that setup through 5 to 10 cycles with the balancer installed than doing the same with the balancer not installed. That is the test you should do. Take your 60/30 set up.. Do 5 charge /discharge and full charge cycles with the balancer and go back down to the middle of voltage. Then check each cell capacity. Do the same test with no balancer. I'll bet you have better results on the balanced battery.
Yes, I have it connected all the time now and it works perfectly fine. I'm slowly charging up the battery again to see what it does at the top.
hello andy, since B2 and B4 are discharged to 0% shouldn't you discharge it to 0% again to check capacity transferred to it?
In this SOC, 66%, 29%, 67%, 29% is the top balancing can make all of them in the same state of charge?
Active balancers are only useful if you need to use mismatched cell. They will help when a lower capacity cell has a higher/lower voltage on charge/discharge so the BMS will not shut off charge/discharge as fast.
No. They are useful for permanent connected series banks, such as salvaged batteries.
Thank you sir
Very clear
You're most welcome
I have the muller energy 250a and the app is different than your newer video. There is a setting called “Balance Settings” what voltage should I set the Equalization Voltage. Default is set at 3300 mv. Any recommendations
The efficiency off the active balancer is awesome, do the math, with a top bleeding balancer you will lose 50% with the same test.
Yes, they are pretty good and do a great job on these high capacity cells.
If only they made some other type of electronic device you could attach to batteries of various configurations that would actually balance the state of charge, not just the cell voltage. It could also address other critical operations, like preventing you from charging the battery with too much current, discharging too much current, or charging when the cells are too cold. Such a unit would manage all of these functions and we could call it a..."Battery Management System", "BMS", for short. :)
Would you use an active balancer or a BMS for a 4s battery pack that sees 1000 amps of current draw?
Aye but with mismatched cells it will divert charge into the higher capacity cell near the end of charging as the curve becomes more steep.
On Discharge it will divert power into lower capacity cells as the curve steepens. You’ll still get more power out of a mismatched battery with an active balancer no?
Yes, you will get more capacity. But it is not much. Don't expect 10% or so, it is more like 1-2%.
Darn that voltage drop that you get after after you fully charge and the drop under load a li battery-that's why balancing should be happening with everything going in the charging direction (battery under charge)=still waiting for the another test on cells based on the capacity you found, 5ah that you did not discharge at 5C based on the capacity of 7.2ah
I have the cheap little $12 I think balancer for my 4s 10p headway pack. It does keep them close but not perfect. Most I've seen it work was 1 amp even. Frustrating part is wondering how efficient they really are and your right, do I really need to leave it on? And like you proved am I losing ah per cell? I check them daily, they are very close. Maybe I'll unplug it for a day and see. Anyway thanks for your excellent content. Very informative. I'm very new to lithium just one of those annoying car audio nuts.
oh Andy .... you have already proved you cant balance in the mid voltage range. Results are no surprise, you know to balance it has to be above the top "knee" (or below the bottom knee). When i saw "equalisation" I thought that you were going to do the balancing on a once a month equalisation voltage rather than your usual moderate - and use that old FLA MPPT setting in conjunction with the balancer - you did suggest this before.
By the way, summer here in UK, I'm on summer MPPT setting 3.35/cell and I'm floating from 66% charge again due to the very flat voltage curves. But at least the bank isnt sitting at high voltage all day - still more than enough power to live and power everything i need too.
I wanted to know the spread of SOC at one particular voltage.
Thumbs up and subscribed!
Awesome thank you!
very interesting...you do good videos..thank you...i wonder if we conected them in parelel would the same thing happen or would it balance voltage and energy? i learn from another of your videos that absorption time counts in terms of energy...i wonder if the parelel method might give more absorption time becaus it does not switch off..also perhaps time with a parelell method might bring the energy into balance?
Interesting test. Good results. So what does balance SOC?
I'm guessing for LiFePo only charging the cells all the way up (to the upper limit of the BMS), then putting the balancer on. Rinse, wash, repeat for wildly divergent SOCs.
O thanks for this test.
You can solder those tabs, just rough them up with sandpaper first.
I think If you left the balancer connected until the balance currents equalled zero, the cells would be very close in charge at that point.
Because of internal resistance, when there is balance current flowing the voltages will not correctly correspond to charge level.
Instead of proving that balancing these cells does not work, I think you just proved that you didn't balance them for long enough to complete the task.
Hello! Great video :) I really appreciate 😊 do you know if I can use exact the same balancer on 4 lifepo4 280 Ah battery?
FWIW, Energy (Whrs) is not linear proportional to Ahrs. SOC is not representative for energy content.
Do the same test measuring Whrs.. The result should be slightly different.
Let's now start with 2 fully charged and 2 empty, leave the balancer connected and then let this combination as a complete 4S 12v battery do a full charge and then a full discharge cycle. This will tell us how the balancer works In a Real life senario.🤓
Great video Andy,congratulation!This video came to meet the problem I am facing now. I received the 32 302 amps batteries which i was waiting for and started the top balancing procedure in three steps, 3.4,3.5 and 3.65 volts. I connected all 32 batteries in parallel and I got a huge bank of 3.2 volts and 9600 amps. In the first phase I charge the batteries with 3.4 volts and the bank sucks a current of approx. 15 amps. Assuming the batteries came with 50% SOC, this first phase will take about 12 days.
Please advise me what to do, wait until the battery bank reaches 3.4 volts and the current close to 0 for the first step, or charge them directly with 3.65 volts. The source I use can charge a maximum of 60 amps.
If they are all in parallel, you can just set the charger to 3.6V and let it connected until the amps are almost zero. That is a good start for connecting them in series afterwards.
@@OffGridGarageAustralia Thank you Andy!I saw on some sites from the guys in the marines that they use this procedure, in three steps, but it takes too long, so I will listen to your advice. I thought that this procedure, in three steps is better for battery health.
Please Andy help me with an opinion or an idea. I connected the 32x302 Ah cells in parallel 4 days ago. Until yesterday I charged them at a voltage of 3.4 volts, and the voltage did not increase more than 3.32 volts from 3.29 as they had when I received them. Yesterday I increased the voltage from the source to 3.6 volts, the current consumed from the source is around 28-30 amps and the voltage is around 3,324 volts.I am worried and I'm afraid there is something wrong with the batteries because the voltage does not increase. Maybe because the battery bank is very large, 3.2 volts / 9600 amps, this top balancing procedure will take a very long time?And it is possible to calculate this time?
@@mihaitaiosub totally normal😊
If you charge 32 cells with 302Ah in parallel with only 30A. That will take: 32x302Ah/30A=322h charging time or close to 2 weeks. Set the charger to 3.6V and let it charge....
@@OffGridGarageAustralia Thanks for the reply. Starting from the fact that maybe they came loaded at 50%, maybe I'm lucky to last only a week. I'm worried that the voltage doesn't increase ay all.
Is there another offgrid garage now, since you've started saying "in Australia" every time?
Could be, I don't know. quite a few people have asked which country I'm in, so...
So Americans don't confuse this with the Austrian show - Off Grid Lederhosen
a hole punch for paper would probably work well for those tabs
Hi there. Interesting test. However, not fair for the active balancer. Your test could bring the capacity closer together given time. Try your test again with the active balancer connected to all the cells. IE. During charging the 4 cells in series, the active balancer keeps the SOC of all cells together. During discharge, Again the active balancer keeps the cells SOC together during discharge. One thing that I have noticed in your testing is that you are not allowing the battery to become fully charged. The batteries equalize best when they are charged to saturation of 3.55 volts or 3.60 volts. Be aware that your DOD of 2.5 volts isn't doing your batteries any favors. My experience is that if you keep your batteries at the lower cut off point that your performance will decrease more aggressive over time. If you choose a voltage closer to 3.0 volts your batteries will last longer. I'm using active balancing here on Li-Ion cells. My system is running at 28.8 volts max upper and 24.5 volts lower max on discharge. Keep smiling 😃
However this does not reflect a realistic use case. In reality you'll be using your batteries in the 20-80% range. This use/load will cause the lower charged cells to drop in voltage bellow the higher charged cells, thus the active balancer will have enough voltage difference to do it's job. Granted it might take longer, but in the end you'll eventually get equalized SOC.
No, it's a test. Like the 5C discharge test.
@@OffGridGarageAustralia Well I can't argue with that :) Love the video's
Not a surprise.
Energy != Voltage
Power != Voltage
Current != Voltage
Flat voltage curve means you're not measuring anything but a near-meaningless voltage number. It's nothing but emergency cut off information at top or bottom.
will these batterys work for a bicycle battery?
Yes, absolutely, they are perfect for that. 16 in series will make the perfect 48V e-bike battery.