You only charged at 5A but if you say charged at 20A then the voltage will have been higher on all points in that curve due to voltage drop inside the cell based on cell internal DC resistance. To be sure that cell is fully charged no matter the cell internal resistance and cell charge rate (within reason) then is good to have a voltage termination at least at 3.5V on that fastly increasing part of the curve at the end. For my SBMS I selected the 3.55V as a good end point for most cases and LiFePO4 batteries to be sure cell is full then charging completely stops as absorption and float are detrimental to Lithium batteries. Also the SOC is set to 100% in case some small errors accumulated since last charge as the SOC is calculated and not based on voltage (as you seen is just not possible to guess SOC from the voltage with LiFePO4).
Great video - I have intuitively known that - charging was complete at 3.4 volts since settling Voltage is around 3.35 - 3.37 V anyway and this confirms that fact. Thanks for posting .
How old are these people claiming to be first/second/third? Anyway, hopefully you now have a better understanding of voltage vs capacity......No need to charge to more than 13.325v for a 4s battery (resting voltage, ok?).
And now you know why using voltage targets to get to a known state-of-charge when charging at a very low C-rate (0.05C at 5A for a 100Ah cell) is nearly impossible without counting the actual watt-hours going into the battery. The curve will look a bit different when charging at 0.2C, and it will look different again when charging at 0.5C. Generally speaking, the higher the charge rate, the higher the voltage target you can set to get the same state-of-charge result. And, consequently, the lower the charge rate, the lower the voltage target you need to set to get the same result. And, as I've said in prior posts, charging at ultra-low C rates runs up against what is basically a no-win situation, because of two things. (1) You wind up having to set such a low voltage target that you hit up against that flat part of the curve and it literally becomes impossible to achieve any sort of reliable state of charge based on only measuring the voltage. And (2) At low voltage targets the battery's current acceptance can easily drop well below what even minimal solar panels could deliver, and you wind up leaving solar energy sitting on the table instead of going into the battery. Plus, as the cells age, those voltage levels all shift. What might work with very fine tuning on a new cell will stop working within a year, requiring re-tuning every year or two. So how does one deal with this problem? Generally speaking, by having a charge controller + BMS combination that is able to actually count the watt-hours going into or out of the battery and stop the charge when the appropriate capacity is reached (regardless of voltage). This way the charge controller can use a higher voltage (hence increasing the battery's current acceptance and using all the energy available from the solar panels), and the charge controller can count (net) watt-hours going into the battery and cut-off the charge based on that to avoid over-charging it. Either the charge controller will have a separate set of terminals for the load and an internal shunt, or the BMS will have a separate set of terminals for the load (and an internal shunt), or you D.I.Y. the shunt and one or the other has sensor terminals for an external shunt (usually requiring a calibration sequence). And then either the charge controller or the BMS with the appropriate feature can be programmed with a capacity limit. If the charger controller handles it then it simply cuts charging off. More likely though the BMS handles it and simply cuts off the charge controller (causing the charge controller's output voltage to instantly reach its target). Without counting watt-hours, the poor-man's solution is to set a voltage target on the low end and then set a non-zero absorption time at that target to try to achieve the desired state of charge. -Matt
The discharge graph is a standard for lifepo4 battery.The flat spot is where you want to keep your batteries at.Charge to just before the voltage shoots up and discharge to where it drops suddenly.I knew this as soon as I saw that graph and you Andy have done excellent work to teach this to people because the battery manufacturers do not tell you this.Its insane how they want you to charge up to 57V knowing it shortens the life of the battery.Thanks
How you you know when to stop charging/discharging at a specific state of charge during a flat part of the I-V curve? You do coulomb counting - use software to keep track of how much energy goes in vs out to estimate where the battery should be at between deeper cycles to re-calibrate the algorithm and reset cumulative error.
Yes, coulomb count is the only option you have, either with a separate shunt or BMS (with integrated shunt). The Victron Smart Shunt I'm using does exactly that. It also recalibrates when the battery is fully charged (parameters can be set individually).
@@OffGridGarageAustralia And exactly what I use on my battery too 👍. (Thanks Andy) It does get out of calibration though if it hasn’t synced at top of charge for a while.
The reason for charging to a high voltage of 3.65 is for speed. Generally you will charge to 3.65 and stop the charging at .05C (5 Amps). The voltage will sag over time to about 3.4v. Knowing that a fully charged cell sits at around 3.4 volts, You could just charge to that voltage until you hit 0 amps and have a fully charged cell. This however would take much longer to achieve the same result.
I am from INDIA , Namaste 👃👃Very help full video to understand LFP Cell Charge and discharge AH and volt profile, before we used volt meters they shows full voltage actually at that time battery drained to 90% after running 8 to 10 km. of E rickshaw Battery shut downs, after watching video we replaced volt meters with True capacity meters they measures each amp hour during charge and discharge thank you very much for making such a nice videos.
thank you you're the most intelligent review of this product I've seen after watching almost 60 videos everybody's trying to revive a damaged battery pack and the whole time I'm thinking it's a support module not a replacement for a dead cell battery you're the only person that actually tested full charge and that was an awesome video I learned a lot
We only need to charge past 3.4 to top balance. However if you don't bulk/absorption charge at a high enough v/cell you will slow the charging down a lot from like 70% . I had it today, I'd turned the MPPT down to 3.4/cell as we had week of sun and couldn't use the power, but just had 2 dull days I'm about to have a wet week, and this afternoon it was sunny so went from 3.4 to 3.5/cell and went to charging at full power from 70% to top off the pack, was throttling the power - I can see it happen as the MPPT basically just moves away from the max power point. In this test stopping at 1A I think you missed a couple of amp hours.
I agree! The reason to allow a higher voltage per cell is to harvest as much solar energy as possible, as fast as possible. Andy actually changed the cut off to 0,1A without telling us. It is visible while examining the end result.
It's only time you save, not capacity. I have another video coming soon where I show that. The difference between charging to 3.4V or 3.5V. Very interesting!
As @off grid and working hard points out, the sun is not always up. You want to harvest when available, and since the cells can take the charge, with as high current as possible hence higher per cell voltage! I’m fully aware of the capacity difference between 3.4 to 3.5, it is close to nothing. Been off-grid for 2 years with lifepo4 now and probably dove in to most caveats by now 🙈
So cool! I love the software and this charge-discharge too is very nice! That is alot for the 100ah! This clarifies the LiFePo4 discharge curve nicely! Thanks so much Andy! Cheers and sunny days bro 🌞 ⛅ 🌞
You always make informative and entertaining videos, Andy! Well done! You briefly mentioned the voltage discrepancy between your multimeter and the charger/discharger. Could you please show some details of the calibration features in this device? I'm very close to the "shut up and take my money!" stage with this device, and possibly the big brother. Lol.
I'm very impressed!!! This answers all my questions with my LiFePO4 batteries wooooh! I don't really watch videos more than 10 minutes but I did! I'm enjoying listening to you and learning at the same time. I already watched your other videos specially on charging the cell to 3.4v VS 3.5v amazing!! Watched it 4 times LOL :D
I bet this is not a new branded cell, but a refurbished one for cheap cause 10 Ah more is too much. Most likely it haad been a 120 Ah battery and to avoid any they simply sell those as 105 Ah. If things go wrong for then the customer will still be happy due to the expectation of 105 Ah And if he gets 10 Ah more then he will be happy,not complain and simply feel lucky and accept scratched terminals and so on. He would not a accept the condition of the batteries if he had bought that battery as a 120 Ah version.
Agree. According to specs that 100Ah battery weights more and is dimensionally larger than an EVE 105Ah battery. I assume the battery makers are deliberately derating the cells capacity specs so the battery still performs well during end of warranty date. Not to say derating specs is wrong, manufacturers have been doing this for a long long time. Andy thanks for another informative video. I learn much from you...
Thanks. 3.4V is a bit low unless you let it absorb for a very long time. If you can charge to 3.45V though, that would be ideal. 3.6V is not necessary and only for top balancing purposes. There is no additional capacity gain charging from 3.45V to 3.6V.
GREAT VIDEOS!!! I plan to see them all!! ...wish I found you earlier! I also got "better than expected" results for a brand new battery. But they dropped off somewhat after a few cycles...have you noticed this action?? THUMBS UP AND SUB. --daLe
SLA compared to LFP? Terrible. Used them for years in my caravan and then swapped. The LiFePO4 are FAR superior in EVERY respect. It’s like comparing a Model T to a Tesla….
I didn't get it, I've been looking for a good explanation about this, Charging and discharging concept. My charging device is for a simple household a, c, d and 9 volts. My question is, in what mode the batteries get charged?. For some people might be simple, not for a 68 year woman.
Sir please test a 32700 6Ah cell and let see at which voltage it reach the full charge and at which voltage it get fully discharged....sir please do have a test
Great video really interesting 👍 A coupla questions . . . Is it correct that you should try to keep the low point 20% and the high point to 80% . . . .amd if yes then how do you set up a system in a car to maintain those perameters? I have solar and ac charging setup but just trying to look after the battery as well as possible. Cheers Ben
What a brilliant tool allowing you to answer do we need to charge to higher volts and what does it give us? I’m pleased to say that I’ve now watched every one of your videos and eagerly awaiting more. Have liked all of them and subscribed too!
@@OffGridGarageAustralia thanks to you. A very entertaining watch. I only hope that my channel launch can provide even half as entertaining content. Hope you have good success getting to the bottom of the panel issues. I’ll be watching :)
Im non techie. Im still confused after watching this video. So at how mamy percent should i use lifepo4 battery before i charge? And what % should i charge to??
As I already explained is full charging with Lifepo4 batteries at least from time to time beneficial to keep the spaces for the lithium ions open in the carbon material, otherwise they wither away when not used and then the remaining spaces will be jammed, which would lead to more expansion and degeneration. Also deep discharge is with lifepo4 batteries not as bad as for the "classic" Lithium batteries, cause in Lifepo4 is moving all the lithium from one to the other side, beside the classig lithium batteries, were only the half of lithium is moving and the other half is needed to keep the structure, where the lithium is docking on to it, when coming back.
What would be critical to point out is what max charge voltage and what min cut-off should we use to get most capacity while preserving battery life. What do you think about 3.45 and 3.0 V ?
Hi Andy, thanks for your test. One more wish, would you please also put the two CSV files on your homepage for download. Then we can have a look at the exact values at certain points. Thanks By the way, different "C" rates in the discharge will show different voltage values. My experience.
@@OffGridGarageAustralia ...wich brings a new question: What would that curve look like in cold or hot environment situations ? That would be extremely relevant. (especially for those living in a different climate) 🤔
To answer your question, you need to get the charging voltage above the knee to properly top balance the cells. In case you don't already know, you can change the title on the graph from "EB Tester Software" to whatever you like. I number my cells and change the title to the cell number so i know which cell was tested. You can also save an image of the graph that you could print and attach to the cell of you want. Having test data for each individual cell is important when you need to put cells in parallel. I have a battery with groups of 2 in parallel and 4 groups in series. Having individual cell data is important to make each group's capacity as close as possible to get the maximum capacity out of the battery.
Yeah, I saw this in the software. That makes sense to keep all the data for future reference. Well, you can only top balance the battery if there is no other current going in and out of the cells at that time. Otherwise your balance current may not be high enough to do its job.
@@OffGridGarageAustralia You don't need the current to stop flowing to balance. You just need to be above the knee. At that point your charge controller will be carrying the system load and your battery can be considered fully charged for all practical purposes. Make sure you have enough voltage difference between your balance voltage and your charge end voltage on your BMS so that the BMS doesn't interrupt the charging cycle during normal operation, especially if you're using a BMS that uses a relay or contactor for protection. Remember it's not the BMS's job to balance an unbalanced bank, only to maintain a balanced one. Assuming your cells were balanced at the start, you should only need a few mA to maintain balance as the cells degrade. If you're using an absorption voltage that's different from your float voltage, you can consider the absorption time as "balance time". Also remember that it's not necessary to achieve perfect balance during every charge cycle, everyday. It can be an iterative process and you only need to keep them close enough to prevent overcharging any one cell and having your BMS trip out on overvoltage. Don't worry about trying to achieve laboratory results using Chinese cells and electronics in the real world.
When the LiFePO batteries have been at rest for a few hours, I can always tell the difference between 'below 55% capacity' and 'above 70% capacity', because there is a 1% difference in open circuit voltage. This 55-70 bump in the curve is true for all LiFePO batteries I've tested. However, the bump will become harder to see during fast charge or discharge.
Hi Aaron, ur numbers sound pretty good to me without the test results, but Im curious after a year, do you still find/believe these numbers are still correct? & when u say u have calculated these numbers, is that from mathematical &/or theoretical calculations, or from personal experience of ur own battery setup maybe? Im curious to find Andy’s results for this, while those numbers sound ok, they may be subject to a number of changes over time is my guess, and probly the best way to fairly accurately know the SOC is by coulomb/watt counting, so if u know what has gone into ur battery, unless there is a problem, u should be able to get that or a percentage thereof back out. Cheers
Hi Andy. If you just cycled the battery between say 3.35 and 3.1V (never discharge lower or charge higher), what would the actual capacity be then? In terms of preventing deterioration (and therefore increasing lifetime cycles), that seems optimal from these first curves. It seems to me that you are just 'derating' the cells by 10 or 12%. Did I understand that correctly?
What does "Mr. Garage" say about a shallow discharge / charge cycle of an LiFePO4 battery? Example - A typical 2000WH SOGEN battery bank with an approximate 500 watt air conditioner load that is fed/charged with ~ 475 watts of solar. When the AC compressor cycles off, now fan only, (~40 watt load), the 475 watts are now charging the Sogen for ~ 5 minutes. Then after the AC compressor kicks back on, the use cycle is repeated. The Sogen display varies from about a 62% charge level when the AC compressor is on and the Sogen charge display increases to about 63-64% when the AC is "fan only". Does this usage pattern have any negative effects on the Sogen's LiFePO4 batteries???? Me AND all your other viewers may have a similar question about this "short cycle" use.... Thanks ahead of time for your experienced opinion....
Keeping the battery right in the middle of its SOC is perfect and there is nothing wrong with doing this. I would still recommend to fully charge that battery maybe once a year and do a top balancing of the cells as they will drift a bit over time.
Thank you very much for your reply. When new, the unit, (Pecron E2000 LFP), arrived at about 65% charge and I DID charge fully, then let it drain with ~ a 400 watt load until the BMS turned the unit off. This initial full cycle- 100% to 0% then full charge to 100% I believe was proper to "calibrate" the newly assembled cells in the total bank. With a potential hurricane that may arrive within days, I'll be sure to top off the unit for the expected power outages. Then every couple of months I'll run a FULL cycle for the BMS to fine tune the cells in the bank. @@OffGridGarageAustralia
Hi Andy, as I tested my batteries I did ccv with 3.4 V and 0,10 A as cut off. Some of my batteries did reach the cut off....and some stayed around 0,20 a for couple of days...and I stopped manually. I am sure my 16 cells are not all from the same producer....the company (BLS) I 've ordered via Ali Express seems to be a trader not a manufacturer. I will continue testing ...my batteries , divided in two blocks now been parallel connected for around 2 weeks. Now I will discharge down to 2.5 V cutoff and charg e then up to 3,65 V. I hope to find out witch cells are almost close from charge& discharge curve and so match best to build a batterie...It may take a while till I have results....looking forward to your next videos.....have fun experimenting...stay curious...;-)
I did the same test and once charging stopped at 3.4V, the absorption time took like 10h+ for this cell to reach 0.1A. I then fully discharged and found the cell having had 100% capacity, so it was fully charged at 3.4V with a loooong absorption time.
hi.i cant get my viltron dc charger to save the lifepo4 setting.its greyed out.ive got a new lifepo4 to install in my camper van,ive adjusted my renogy rover mppt to li setting any ideas anyone ?I had my engine running
The stepped curve as against smooth on the zoomed in part of the charge graph suggests an artifact associated with the electronics not the battery chemistry.
Thanks mate a nice test a test like this is helping me a lot with questions I have sure i am going to use it in my serie connection to 500 volt. no i do not have all the answers but battery behavior matters..
Very helpful, I am from Portugal and I am building small battery bank to support peaks on my solar installation. Only 2K battery 48V. My doubts was till wich level charge and discharge. Your videos are fantastic, keep going.
I have always questioned the "rule of thumb" figures thrown out for charging and discharging cutoff. This is the first instance I have seen of someone closely examining the charging and discharging curves. Of course, this is only the base data, I assume that you will try adjusting the cutoffs to see how much it impacts the capacity. Determining how various cutoff points affect the longevity of the cells might take longer than we have left in our own lives. Thank you for your efforts.
Thanks, Jeff. Yes, this is an older video where I started to find out what is actually happening with charging and discharging LiFePO4. I have many more videos on the channel, where I did all the testing and analysis of what you said in your comment. As I could not find any useful and consistent information online, I started my own experiments and shared the results here on the channel.
Just a thanks for your videos. Bought this, through your link, and am doing a cell test tonight. Wouldn't have got it started without your very thorough instructions, and google of course.
Es gibt aber auch noch ein Mittelding zwischen dem kleinen 20er und dem 40er, nämlich den EBD-A20H. Damit kannst Du auch 20A laden und er geht auch bis 30V.
Ancient man made fire by friction and force and and said : "I am man!". Today we make sparks made from the sun and say : "I am modern man !" We haven't changed that much but the technology sure has :) Cheers.
@@OffGridGarageAustralia nice - glad you liked it. I edited the original post because if misspelled change[d] - omitted the d. :) Clearly I cannot type very well.
I'm puzzled as to why Andy (or any other off grid tragic) needs to be doing all this in depth testing stuff. 1. The manufacturers of each battery factory research and provide the most efficient charge and discharge graph specs for their product. I've got mine. Why not just follow them? 2. They are deduced by tertiary-educated chemical-electro engineers, probably in touch with the best lithium chemists in the world when making their batteries. Some design batteries for space travel. Every product would be bench tested 100x more than Andy's work. 3. Despite Andy's months of garage testing, good manufacturers have teams of engineers working 5 days per week and millions of dollars of testing equipment. Why are we humble customers not just following their product spec sheets? Does Andy REALLY THINK he's smarter than all the technical people put together in (for instance) the Daly Dongguan BMS Electrics Company or Tianjin Lishen Battery Company, rated 5th in the world? These are young, cutting edge, PhD Chinese graduates, born into this era of hi-tech battery technology. 4. Won't each battery maker's product vary somewhat from the tests Andy does on his brands? I've got Lishen Cells, reputed to be one of the best. They make billions for e-Bikes. But I know there are many totally crap brands of RV LiFePo batteries out there with *recycled lithium* ... nominal 100ah but probably good for only 80; supposed to be 12.5kg of lithium, but really only 10kg. Surely those ones will perform WAY DIFFERENT to mine or Andy's. 5. I appreciate there are many lithium myths out there which need to be disproven, and many amateurs like myself thirsting for good information. But surely, "just follow your manufacturer's recommended spec sheets" is the easiest way to go. Ffs, Victron (for example) know their products better than anyone. Just explain them. No need to critique everything as somehow imperfect. Why all this perfectionist angst? Or is it just a curiosity neurosis hobby for retired old codgers in garages who think their battery or BMS maker "might have missed something".
Hi Andy! I have no previous experience with any BMS device. I am confused about the operation of the JK BMS device because I have mixed up the device: Situation 1) The measured capacity of the cells is 307Ah, but I originally wrote the BMS "Battery Capacity" value as 280Ah and I have also tried the values 307Ah and 300Ah for that point. It seems that changing this value completely messes up the calculations. A real example: When I set "Battery Capacity" to 280Ah and "Remain Battery" = 100% and "Remain Capacity" = 280Ah. Then I change "Battery Capacity" to 300Ah so "Remain Battery" = 69% and "Remain Capacity" = 193.2Ah. This seems completely counterintuitive! Situation 2) I charged the battery directly through its main terminals - that is, I did not put the negative wire of the charger in the "-P" port of the BMS device, but directly in the terminal of the battery. While doing this, I noticed that the voltages of the cells do increase, but the BMS does not increase the "Remain Battery" and "Remain Capacity" values at all, even if I charge with a 25A charger for 40 minutes. So the information about the actual booking amount is completely out of place?
Could there be a programming error or flaw in its software for Situation 1 that they never noticed?
But anyway, I'm asking for advice: - Can I safely forget about these readings like "Remain Battery" and "Remain Capacity" and just focus on monitoring things like "Ave. Cell Volt." value to see what the actual battery charge is? - Would it be wisest to use the battery in the interval where the pole voltage moves gently to decrease or increase - and not discharge it so empty, where the voltage drops quickly, and not charge it so full, where the voltage rises quickly? - Which voltage values would be the best for the OVP and OVPR and UVPR and UVP points of the cells?
Now if you had a time machine, you could repeat the test after a few years use and see how well the battery holds up with such accurate data to compare, as always, thanks for the great video.
Andy you say "10% more battery capacity" is impressive. No it's not. Every manufacturer of everything specs their product with a nominal capacity with variously hidden "headroom" or "safety factor" above and below what they want the customer to use. Eg, Crane cables are some 3x the Safe Working Load. If you tested a 1 tonne to break at 3 tonnes you should NOT be impressed it can lift 300% more than the label says. Conversely, if you found a battery which only gave 100% of its specs, you should in fact be UNIMPRESSED it is graded right on its edge.
Warning, highly controversial and potentially inflammatory opinions ahead... :) "If you're still using Fahrenheit" Sure, Celsius more popular, but does that really make it better? Why should our temp scale be related to the freezing and boiling point of water? I'm not sure why that matters. Shouldn't it be based on the human experience? 0 Fahrenheit is **generally** about as cold as it ever gets and 100 Fahrenheit is **generally** as hot as it ever gets. We get a much better resolution that way. thevane.gawker.com/fahrenheit-is-a-better-temperature-scale-than-celsius-1691707793#:~:text=Fahrenheit%20gives%20you%20almost%20double,than%20Saint%20Celsius%20ever%20could. (BTW, I really don't care how **YOU** measure your temperature. Do whatever makes you happy.)
Could always add extra decimal places for resolution. Not necessarily limited to integer values, that's just a decision of the thermometer makers and not inherent in the temperature scale -
@@OffGridGarageAustralia at 3.45V absorption, you almost fully charge the cells without taking them high up the upper knee. At 3.375V float, you can keep using power from solar after the battery has been charged. Anything lower, you discharge the battery and waste solar, any higher and you start damaging the battery. There’s a consensus that under 3.4V you can’t overcharge a cell no matter how long you absorb at that voltage, while in this setup it only lasts from fully charged to sunset. These numbers come up after a period of testing and everyone ends up at those. The graphs you create show the logic here as well. About balancing: in my opinion the BMS protections are a last resort fail safe. I never want to see them triggered. I balanced the cells manually and they maintain balance. I can charge single cells inline but didn’t need to yet.
You are finding exactly what I have found. No need to go below 3 volts, and above 3.4 volts, there's almost nothing outside that range. You are still limited by the weakest cell, but that is true no matter what charge and discharge profile you use. You can easily get 90% of capacity without placing any stress on the cells. I had found out using the bigger tester that my laptop battery won't last long enough to record a complete test, so for the last month I have been waiting on the parts for an 11th generation in NUC to be able to save the graphs. They just released it about two weeks ago, so I finished building it yesterday. Had to use a dummy HDMI plug so it can boot headless, and I just remote desktop to it. So I have only completed 2 recorded tests, despite having it for a month now (Eve cells), but have done lots of manual tests and was reaching the same conclusions as you about the time you did the videos on charging and float voltages. On a 280AH cell (measures 278.8) I am finding about 5 amp hours above 3.4, and 5 amp hours below 3.0. I think I like your 3.1v cutoff better.
Thank for sharing, John. A NUC is a good idea as it uses only a few watts of power running for so long. I've got my old Dell workshop computer running for several days now and with the cloudy weather at the moment, solar is barely enough to recharge what I take out with all my testing.
@@OffGridGarageAustralia Yes, since it uses a laptop CPU, memory and SSD, it is very good on power consumption (like a laptop). It even uses a 65 watt power supply just like a laptop. I did have to buy a 10 port USB hub, since I have a lot of programmable loads and power supplies to connect. Four of the 18 amp Riden power supplies, and four of the ET5410 electronic loads. This is for testing packs at higher rates, obviously.
Excellent Vid Andy setting up my workshp off grid: LiFeP04 48vdc 100Ahr with MPP Solar contro ller. From yor reasearch is the "best" to set the charge/discharge to cycle between 98 and 10% ? My place will be 15 to 20 degrees C at all Times year.
Hello, I just purchased two of the EcoFlow Delta Pro Battery systems along with the Double Voltage Hub to join the two EcoFlows together to provide 220 volts to my 10 circuit stand by electric panel. I also purchased two EcoWorthy Solar Tracker systems and mounted 4 of the 440Watt Magnus Solar Panels on each Solar tracker system. After watching a few of your very informative videos is: What should I set the charge limit to and what should set the discharge limit to? From What I learned from your videos I have set the Charge limit to 95% and the discharge limit to 5%. I want to get the most life from the EcoFlow Units and get the most energy from them. Thank You for taking time to educate the public.
Could be just the software and the resolution it uses to draw the graph. Also if you move the probes at terminals, it sometimes has better/worse connection and voltage goes up or down.
Andy you maybe interestedin the icharge x6 or x8 charger/discharger as it has the ability to use a bigger battery as the load, so you can do discharge tests but capture that energy in your bigger battery.
I looked into it but could not find any information that you could actually charge a battery with a higher voltage during the regen discharge? So I cannot discharge the 3.2V cell and charge this energy into the 50V battery bank, right?
@@OffGridGarageAustralia I keep forgetting your pack V, you would need to look at an X12 with a Volt input of 11-53V dc. But it will take the single cell output regen to wherever you choose to apply it. Not stopping smog, but cool
@@chriscord6524Yeah think you are right the x12 does have the better input voltage range as my x8 only goes up to 49v which isnt quite enought for a 48v pack at full charge.
Awesome! This proves all of my hours of watching my batteries and learning (The hard way!) I also determined it was not worth squeezing every little bit out of the batteries. My settings for my pack are 50v (3.125v/cell) at the bottom. and 56v (3.5v/cell) at the top. The top voltage in my opinion could be lowered to 3.4v/cell with very little loss of ampacity. Thanks for sharing this. Very interesting!
Thanks, Dan. Yeah, that is absolutely right and you have some good settings there for your battery. I've done a test with charging to 3.4V only. Interesting result. Video is coming soon.
Hi, I have 8 Pylontech 2.4 kWh 48v 50Ah 15 cell modules and your videos clearly demonstrate just how safe they play it with the recommended charge/discharge cut off voltages, 53.3v charge cutoff and 47.2 v discharge cutoff which gives a cell voltage of 3.5533v maximum charge voltage and 3.1466 minimum discharge voltage. Is it possible to calculate the number of complete cycles by dividing the total kWh hours drawn from the battery bank ie 2,530kWh by the battery bank capacity ie 19.2kWh? Or do I need to access this data by connecting a lap top to the master battery data port?
Yes, these voltages you mentioned are totally save to use. You cannot just divide the kWh by the capacity of the bank. That would give you a rough idea but usually you don't do full charges from 0-100%. If you do 10 cycles from 50%-60%-50% SOC it is far softer than doing one from 0-100%. It all depends where you do your cucles (near the peak at bottom or top) and what the DoD is.
Awesome tester, any chance you can do a 5 Amp discharge test on the 100AH lifepo4 battery, and share the csv. I would like to see the voltages during the discharge as they will be higher than discharging at 20 Amps. I'll sponsor the beer's for the test 👍
Absolutely, Colin. I'll put your request on the list of future tests. Will come soon and I'll make a short video about it. Would be interesting to see how much more capacity we will get out at a lower discharge rate.
An interesting test would be how the battery reacts to both high charge and high discharge for short periods. So near the bottom and you have a high discharge how much sag and where it's soc returns to. Same with charge.
I just got my "new" battery cells from AliExpress and was obviously ripped off. Could you show the "resistance" test of your small (20A) battery tester? I own the same and would like to know which values to expect in new batteries. Depending on which current I use the result is between 3 and 10 mR, which is is way too high in my opinion. But maybe this tester doesn't work fine-grained enough.
I have a request Andy, when mentioning the temperatures of your equipment could you put down in the corner what that is in Fahrenheit? Celsius doesn't mean anything to me as I'm in the US. Thanks so much for your videos, I've learned so much from you.
Hi, thx for the video - have same device for testing..since the EBC-A20 offers also just to monitor cells - will it work with an external device for charging/discharging and if so, how many "external" amps it could coop with? the 5A loading and even 20A discharge is a pain for 310Ah cells...
Brilliant, however I wonder if you could take your viewers through the whole process of setting up the tester and software as I also would like to obtain the big brother of the tester you have used in this video. Thanks, Ken in England
I am planning to build a 12v battery, only have a dc dc converter, wondering if this would be good enough to charge or donyou need something that stops ats the current drops?
A DC to DC converter will work if setup correctly. I have used one many times to charge a 12V lithium battery. Important is that the battery still has it's own BMS as protection device. Set the DC converter to 13.8V and let the battery charge until the current is almost 0A (or under 5A for a 100Ah 12V battery). It's then charged 99.5%. Remove the inverter afterwards.
The results are good for general guidelines but not much more than that. Results will be different between different lots or different manufacturers 100Ah batteries and will probably change for used batteries.
No surprise that at a 0.05C discharge rate that you get better than 100AH. Are you happy with the smaller cell or do you regret not getting an Eve LF280AH cell for testing? Nice work
Where can i find copies of the data (charge/discharge) that you generated. The curves were quite interesting. Seeing what the cycle degrade looks like would be interesting. Why put in more energy than what you take out---but, this probably changes so that's probably why they charge to higher voltage points. This was interesting. Thanks
Great video, very useful! I would love to see you doing the same test on one of the EVE280 cells! I am in the middle of top balancing my 16 280’s. All in parallell with 10A, they have stayed at 3,35v all weekend now. 😂 Starting my setup with Multiplus and other charge controllers very soon, and I am very uncertain how high to set the cut of voltage. As mentioned in the comments, balancing can only be done at the high (or low) end. But in the other hand, I do not want ro stress the cells. I guess I will end up with a fairly low cut of voltage (3,4?) and then adjust that to 3,65 in between, if/when the cells get out of balance. I really love your videos, it makes my day. Keep exploring! 😊
One test that I'd like to see with these cells that I can't find anywhere is... Can we boost up the voltage with a step-up module to 24v, 36v or 48v for ex. for an ebike battery using only 1 or 2 cells. So 3.2 or 6.4v boosted to 36v with the stepup module in order to make a small simple battery with few connections like the 18650 lithium cells. Would love to know if this is possible to do.
Hi CK, while anything is technically possible, this idea would require a very hefty booster circuit to achieve the results ur looking for on an ebike & that would probly get pricey. Remember v=ir, so this 100Ah 3.2V Cell has roughly 320 watts, so say ur ebike motor is 32v and uses 10A, that’s 320 watts right? Not really when u go thru a soft grassy patch or up a hill, but also remember ur 3.2v cell now has to put out at least 100A continuous. I like ur idea for simplicity, but I just dont think it would be very viable the way electronics are presently. Cheers
Don't worry about voltage. Worry about high State-Of-Charge (SOC), THAT is what causes damage to Li-Ion cells with graphite anode (including LiFePo4). And another observation: Charge with 5 Amps to 3.4 volts, no absorption, gives you about 98% SOC Charge with 20 Amps to 3.4 volts, no absorption, gives you about 70 - 80% SOC, a bit unpredictable. Later edit: It is true that using higher voltages increases the risk that the battery would charge to 100% or maybe even overcharge a bit.
You only charged at 5A but if you say charged at 20A then the voltage will have been higher on all points in that curve due to voltage drop inside the cell based on cell internal DC resistance.
To be sure that cell is fully charged no matter the cell internal resistance and cell charge rate (within reason) then is good to have a voltage termination at least at 3.5V on that fastly increasing part of the curve at the end. For my SBMS I selected the 3.55V as a good end point for most cases and LiFePO4 batteries to be sure cell is full then charging completely stops as absorption and float are detrimental to Lithium batteries. Also the SOC is set to 100% in case some small errors accumulated since last charge as the SOC is calculated and not based on voltage (as you seen is just not possible to guess SOC from the voltage with LiFePO4).
Thanks, the 20A test is coming very soon.
I was looking for your explanation/comment here Dacian :)) excellent observations...
Thumbs up and subscribed! No one is breaking it down like this on youtube, thanks!
Thanks for the sub!
Great video - I have intuitively known that - charging was complete at 3.4 volts since settling Voltage is around 3.35 - 3.37 V anyway and this confirms that fact. Thanks for posting .
Cool, thanks
How old are these people claiming to be first/second/third? Anyway, hopefully you now have a better understanding of voltage vs capacity......No need to charge to more than 13.325v for a 4s battery (resting voltage, ok?).
Well, they are as old as people mentioning them 😂😂😂
And now you know why using voltage targets to get to a known state-of-charge when charging at a very low C-rate (0.05C at 5A for a 100Ah cell) is nearly impossible without counting the actual watt-hours going into the battery. The curve will look a bit different when charging at 0.2C, and it will look different again when charging at 0.5C.
Generally speaking, the higher the charge rate, the higher the voltage target you can set to get the same state-of-charge result. And, consequently, the lower the charge rate, the lower the voltage target you need to set to get the same result. And, as I've said in prior posts, charging at ultra-low C rates runs up against what is basically a no-win situation, because of two things. (1) You wind up having to set such a low voltage target that you hit up against that flat part of the curve and it literally becomes impossible to achieve any sort of reliable state of charge based on only measuring the voltage. And (2) At low voltage targets the battery's current acceptance can easily drop well below what even minimal solar panels could deliver, and you wind up leaving solar energy sitting on the table instead of going into the battery.
Plus, as the cells age, those voltage levels all shift. What might work with very fine tuning on a new cell will stop working within a year, requiring re-tuning every year or two.
So how does one deal with this problem? Generally speaking, by having a charge controller + BMS combination that is able to actually count the watt-hours going into or out of the battery and stop the charge when the appropriate capacity is reached (regardless of voltage). This way the charge controller can use a higher voltage (hence increasing the battery's current acceptance and using all the energy available from the solar panels), and the charge controller can count (net) watt-hours going into the battery and cut-off the charge based on that to avoid over-charging it.
Either the charge controller will have a separate set of terminals for the load and an internal shunt, or the BMS will have a separate set of terminals for the load (and an internal shunt), or you D.I.Y. the shunt and one or the other has sensor terminals for an external shunt (usually requiring a calibration sequence). And then either the charge controller or the BMS with the appropriate feature can be programmed with a capacity limit. If the charger controller handles it then it simply cuts charging off. More likely though the BMS handles it and simply cuts off the charge controller (causing the charge controller's output voltage to instantly reach its target).
Without counting watt-hours, the poor-man's solution is to set a voltage target on the low end and then set a non-zero absorption time at that target to try to achieve the desired state of charge.
-Matt
Thanks Matt. good comment.
The discharge graph is a standard for lifepo4 battery.The flat spot is where you want to keep your batteries at.Charge to just before the voltage shoots up and discharge to where it drops suddenly.I knew this as soon as I saw that graph and you Andy have done excellent work to teach this to people because the battery manufacturers do not tell you this.Its insane how they want you to charge up to 57V knowing it shortens the life of the battery.Thanks
How you you know when to stop charging/discharging at a specific state of charge during a flat part of the I-V curve? You do coulomb counting - use software to keep track of how much energy goes in vs out to estimate where the battery should be at between deeper cycles to re-calibrate the algorithm and reset cumulative error.
Yes, coulomb count is the only option you have, either with a separate shunt or BMS (with integrated shunt).
The Victron Smart Shunt I'm using does exactly that. It also recalibrates when the battery is fully charged (parameters can be set individually).
@@OffGridGarageAustralia And exactly what I use on my battery too 👍. (Thanks Andy)
It does get out of calibration though if it hasn’t synced at top of charge for a while.
First !
Awesome video! Thats a pretty neat charger/discharger and like the way it graphs everything out!!!
Thanks 👍
The reason for charging to a high voltage of 3.65 is for speed. Generally you will charge to 3.65 and stop the charging at .05C (5 Amps). The voltage will sag over time to about 3.4v. Knowing that a fully charged cell sits at around 3.4 volts, You could just charge to that voltage until you hit 0 amps and have a fully charged cell. This however would take much longer to achieve the same result.
Yeah, I figured that now with doing more testing. The absorption time at a lower voltage is taking forever.
I am from INDIA , Namaste 👃👃Very help full video to understand LFP Cell Charge and discharge AH and volt profile, before we used volt meters they shows full voltage actually at that time battery drained to 90% after running 8 to 10 km. of E rickshaw Battery shut downs, after watching video we replaced volt meters with True capacity meters they measures each amp hour during charge and discharge thank you very much for making such a nice videos.
thank you you're the most intelligent review of this product I've seen after watching almost 60 videos everybody's trying to revive a damaged battery pack and the whole time I'm thinking it's a support module not a replacement for a dead cell battery you're the only person that actually tested full charge and that was an awesome video I learned a lot
We only need to charge past 3.4 to top balance.
However if you don't bulk/absorption charge at a high enough v/cell you will slow the charging down a lot from like 70% .
I had it today, I'd turned the MPPT down to 3.4/cell as we had week of sun and couldn't use the power, but just had 2 dull days I'm about to have a wet week, and this afternoon it was sunny so went from 3.4 to 3.5/cell and went to charging at full power from 70% to top off the pack, was throttling the power - I can see it happen as the MPPT basically just moves away from the max power point.
In this test stopping at 1A I think you missed a couple of amp hours.
I agree!
The reason to allow a higher voltage per cell is to harvest as much solar energy as possible, as fast as possible.
Andy actually changed the cut off to 0,1A without telling us. It is visible while examining the end result.
It's only time you save, not capacity. I have another video coming soon where I show that. The difference between charging to 3.4V or 3.5V. Very interesting!
@@OffGridGarageAustralia yes Andy, but the sun is not out all the time ;) especially here in the UK. So do not want to miss out on photons!
As @off grid and working hard points out, the sun is not always up. You want to harvest when available, and since the cells can take the charge, with as high current as possible hence higher per cell voltage!
I’m fully aware of the capacity difference between 3.4 to 3.5, it is close to nothing. Been off-grid for 2 years with lifepo4 now and probably dove in to most caveats by now 🙈
@@Mr.X3D The time limit is the factor here for charging 'higher' and therefore faster, I agree with you.
So cool! I love the software and this charge-discharge too is very nice! That is alot for the 100ah! This clarifies the LiFePo4 discharge curve nicely! Thanks so much Andy! Cheers and sunny days bro 🌞 ⛅ 🌞
Thanks a lot for your kind words, David.
You always make informative and entertaining videos, Andy! Well done!
You briefly mentioned the voltage discrepancy between your multimeter and the charger/discharger. Could you please show some details of the calibration features in this device?
I'm very close to the "shut up and take my money!" stage with this device, and possibly the big brother. Lol.
A good test would be to compare the time it takes for a charge to 3.65 at 5 amp cut-off vs a charge to 3.4 volts with 0 amp cut-off.
Very good. I'll put this on my list.
I'm very impressed!!! This answers all my questions with my LiFePO4 batteries wooooh! I don't really watch videos more than 10 minutes but I did! I'm enjoying listening to you and learning at the same time. I already watched your other videos specially on charging the cell to 3.4v VS 3.5v amazing!! Watched it 4 times LOL :D
Oh, wow, thank you very much, Michael. Glad the information if of use for you. And, welcome to the channel!
I bet this is not a new branded cell, but a refurbished one for cheap cause 10 Ah more is too much.
Most likely it haad been a 120 Ah battery and to avoid any they simply sell those as 105 Ah.
If things go wrong for then the customer will still be happy due to the expectation of 105 Ah
And if he gets 10 Ah more then he will be happy,not complain and simply feel lucky and accept scratched terminals and so on. He would not a accept the condition of the batteries if he had bought that battery as a 120 Ah version.
Agree.
According to specs that 100Ah battery weights more and is dimensionally larger than an EVE 105Ah battery.
I assume the battery makers are deliberately derating the cells capacity specs so the battery still performs well during end of warranty date. Not to say derating specs is wrong, manufacturers have been doing this for a long long time.
Andy thanks for another informative video. I learn much from you...
wrng hugh catiy les full chard diszçh bater ited chàre suchage bgrrr bsrr lobger it kadt
Thanks for the amazing video. Is it enough if we charge lifepo4 battery upto 3.4V or we need to charge upto 3.6V?
Thanks. 3.4V is a bit low unless you let it absorb for a very long time. If you can charge to 3.45V though, that would be ideal. 3.6V is not necessary and only for top balancing purposes. There is no additional capacity gain charging from 3.45V to 3.6V.
@@OffGridGarageAustralia Thanks for the reply sir
Hi from INDIA
You are respectful in knowledge and experience and age.
Myself Teacher in govt school of Delhi Thanking you.
GREAT VIDEOS!!! I plan to see them all!! ...wish I found you earlier!
I also got "better than expected" results for a brand new battery. But they dropped off somewhat after a few cycles...have you noticed this action??
THUMBS UP AND SUB.
--daLe
Would be interested in seeing how sealed lead acid batteries compare with LiFePo?
I have only one sealed battery here and it's not working great any more. I'll get another one and do some testing.
SLA compared to LFP?
Terrible. Used them for years in my caravan and then swapped. The LiFePO4 are FAR superior in EVERY respect. It’s like comparing a Model T to a Tesla….
Thank you Andy for the capacity test using the tester. I'm also looking to buy a similar tester.
Wait for the bigger one in one of the next videos. Then make a decision.
@@OffGridGarageAustralia Yup, waiting. Thanks Andy!
I didn't get it, I've been looking for a good explanation about this, Charging and discharging concept.
My charging device is for a simple household a, c, d and 9 volts.
My question is, in what mode the batteries get charged?. For some people might be simple, not for a 68 year woman.
Sir please test a 32700 6Ah cell and let see at which voltage it reach the full charge and at which voltage it get fully discharged....sir please do have a test
Honestly you should have pushed the cell more. 20amp discharge and 5 amp charge isn't jack. 50/50 would have been more realistic.
15:10 Did you say that the ZKE-Tech EBC-A40L can't test a 4s battery, only single cells? But the smaller ZKE-Tech EBC-A20 can?
Great video Andy! I really enjoy your channel. Thanks for being so perseverant in doing your testing and analysis!
Great video really interesting 👍
A coupla questions . . .
Is it correct that you should try to keep the low point 20% and the high point to 80% . . . .amd if yes then how do you set up a system in a car to maintain those perameters?
I have solar and ac charging setup but just trying to look after the battery as well as possible.
Cheers Ben
this is why you need to top end balance a battery, This tester would be ideal to test my Yinglong LTO cell
What a brilliant tool allowing you to answer do we need to charge to higher volts and what does it give us?
I’m pleased to say that I’ve now watched every one of your videos and eagerly awaiting more.
Have liked all of them and subscribed too!
You have watched every video??? Wow, thank you so much. I haven't even watch all of them 😂
Thanks so much!
@@OffGridGarageAustralia thanks to you. A very entertaining watch. I only hope that my channel launch can provide even half as entertaining content.
Hope you have good success getting to the bottom of the panel issues. I’ll be watching :)
Im non techie. Im still confused after watching this video. So at how mamy percent should i use lifepo4 battery before i charge? And what % should i charge to??
Definitely a nice and capable tester/evaluator for LiFeSO4 battery! I would want one if I felt I would use it more than once. 👍🤪
Totally agree!
As I already explained is full charging with Lifepo4 batteries at least from time to time beneficial to keep the spaces for the lithium ions open in the carbon material, otherwise they wither away when not used and then the remaining spaces will be jammed, which would lead to more expansion and degeneration.
Also deep discharge is with lifepo4 batteries not as bad as for the "classic" Lithium batteries, cause in Lifepo4 is moving all the lithium from one to the other side, beside the classig lithium batteries, were only the half of lithium is moving and the other half is needed to keep the structure, where the lithium is docking on to it, when coming back.
What would be critical to point out is what max charge voltage and what min cut-off should we use to get most capacity while preserving battery life. What do you think about 3.45 and 3.0 V ?
Yes, 3.45V as bulk/absorption voltage and 3.0V as cut off for the inverter are perfect.
Hi Andy, thanks for your test. One more wish, would you please also put the two CSV files on your homepage for download. Then we can have a look at the exact values at certain points. Thanks
By the way, different "C" rates in the discharge will show different voltage values. My experience.
Thanks Mel, I have now shared these .csv files as well.
Yeah the curves are only for this chemistry under these specific conditions.
@@OffGridGarageAustralia
...wich brings a new question:
What would that curve look like in cold or hot environment situations ?
That would be extremely relevant.
(especially for those living in a different climate)
🤔
To answer your question, you need to get the charging voltage above the knee to properly top balance the cells. In case you don't already know, you can change the title on the graph from "EB Tester Software" to whatever you like. I number my cells and change the title to the cell number so i know which cell was tested. You can also save an image of the graph that you could print and attach to the cell of you want. Having test data for each individual cell is important when you need to put cells in parallel. I have a battery with groups of 2 in parallel and 4 groups in series. Having individual cell data is important to make each group's capacity as close as possible to get the maximum capacity out of the battery.
Yeah, I saw this in the software. That makes sense to keep all the data for future reference.
Well, you can only top balance the battery if there is no other current going in and out of the cells at that time. Otherwise your balance current may not be high enough to do its job.
@@OffGridGarageAustralia You don't need the current to stop flowing to balance. You just need to be above the knee. At that point your charge controller will be carrying the system load and your battery can be considered fully charged for all practical purposes. Make sure you have enough voltage difference between your balance voltage and your charge end voltage on your BMS so that the BMS doesn't interrupt the charging cycle during normal operation, especially if you're using a BMS that uses a relay or contactor for protection. Remember it's not the BMS's job to balance an unbalanced bank, only to maintain a balanced one. Assuming your cells were balanced at the start, you should only need a few mA to maintain balance as the cells degrade. If you're using an absorption voltage that's different from your float voltage, you can consider the absorption time as "balance time". Also remember that it's not necessary to achieve perfect balance during every charge cycle, everyday. It can be an iterative process and you only need to keep them close enough to prevent overcharging any one cell and having your BMS trip out on overvoltage. Don't worry about trying to achieve laboratory results using Chinese cells and electronics in the real world.
When the LiFePO batteries have been at rest for a few hours, I can always tell the difference between 'below 55% capacity' and 'above 70% capacity', because there is a 1% difference in open circuit voltage. This 55-70 bump in the curve is true for all LiFePO batteries I've tested. However, the bump will become harder to see during fast charge or discharge.
Thanks for the info.
Lithium ion or lifepo4 which gives good power even at 10% battery charge left (for e scooter)
Sir plz mention discharge battery equipment name and model.I need to Discharge carbon zinc D type battery.
I am looking forward to seeing if Andy agrees with me that 3.23 and 3.32 fits the 80% and 20 % depth of discharge rates I have calculated
I'll do the test!
Hi Aaron, ur numbers sound pretty good to me without the test results, but Im curious after a year, do you still find/believe these numbers are still correct? & when u say u have calculated these numbers, is that from mathematical &/or theoretical calculations, or from personal experience of ur own battery setup maybe?
Im curious to find Andy’s results for this, while those numbers sound ok, they may be subject to a number of changes over time is my guess, and probly the best way to fairly accurately know the SOC is by coulomb/watt counting, so if u know what has gone into ur battery, unless there is a problem, u should be able to get that or a percentage thereof back out. Cheers
Hi Andy. If you just cycled the battery between say 3.35 and 3.1V (never discharge lower or charge higher), what would the actual capacity be then? In terms of preventing deterioration (and therefore increasing lifetime cycles), that seems optimal from these first curves. It seems to me that you are just 'derating' the cells by 10 or 12%. Did I understand that correctly?
The problem is charging to 3.35V only, you might a different SOC every time. This is still in that flat area of the curve. I'll do the test ;)
What does "Mr. Garage" say about a shallow discharge / charge cycle of an LiFePO4 battery? Example - A typical 2000WH SOGEN battery bank with an approximate 500 watt air conditioner load that is fed/charged with ~ 475 watts of solar. When the AC compressor cycles off, now fan only, (~40 watt load), the 475 watts are now charging the Sogen for ~ 5 minutes. Then after the AC compressor kicks back on, the use cycle is repeated. The Sogen display varies from about a 62% charge level when the AC compressor is on and the Sogen charge display increases to about 63-64% when the AC is "fan only". Does this usage pattern have any negative effects on the Sogen's LiFePO4 batteries???? Me AND all your other viewers may have a similar question about this "short cycle" use.... Thanks ahead of time for your experienced opinion....
Keeping the battery right in the middle of its SOC is perfect and there is nothing wrong with doing this.
I would still recommend to fully charge that battery maybe once a year and do a top balancing of the cells as they will drift a bit over time.
Thank you very much for your reply. When new, the unit, (Pecron E2000 LFP), arrived at about 65% charge and I DID charge fully, then let it drain with ~ a 400 watt load until the BMS turned the unit off. This initial full cycle- 100% to 0% then full charge to 100% I believe was proper to "calibrate" the newly assembled cells in the total bank. With a potential hurricane that may arrive within days, I'll be sure to top off the unit for the expected power outages. Then every couple of months I'll run a FULL cycle for the BMS to fine tune the cells in the bank. @@OffGridGarageAustralia
Hi Andy, as I tested my batteries I did ccv with 3.4 V and 0,10 A as cut off. Some of my batteries did reach the cut off....and some stayed around 0,20 a for couple of days...and I stopped manually. I am sure my 16 cells are not all from the same producer....the company (BLS) I 've ordered via Ali Express seems to be a trader not a manufacturer. I will continue testing ...my batteries , divided in two blocks now been parallel connected for around 2 weeks. Now I will discharge down to 2.5 V cutoff and charg e then up to 3,65 V. I hope to find out witch cells are almost close from charge& discharge curve and so match best to build a batterie...It may take a while till I have results....looking forward to your next videos.....have fun experimenting...stay curious...;-)
I did the same test and once charging stopped at 3.4V, the absorption time took like 10h+ for this cell to reach 0.1A. I then fully discharged and found the cell having had 100% capacity, so it was fully charged at 3.4V with a loooong absorption time.
hi.i cant get my viltron dc charger to save the lifepo4 setting.its greyed out.ive got a new lifepo4 to install in my camper van,ive adjusted my renogy rover mppt to li setting any ideas anyone ?I had my engine running
Can I store a 200amp hour lifepro at 100%? If not what can I use to discharge to 50%?
Sehr schöner, informativer Beitrag, danke!
Aber Du bist nicht wirklich in Australien geboren, oder?
Ne, bin nicht von hier, nur umgezogen 😉
Such mal nach 'Living Down Under' hier auf RUclips, dann findeste meinen anderen deutschen Kanal.
@@OffGridGarageAustralia Danke, auch ein sehr interessanter Kanal!
The stepped curve as against smooth on the zoomed in part of the charge graph suggests an artifact associated with the electronics not the battery chemistry.
I believe that GX 20 connectors on battery tester are rated up to 15A max...
Thanks mate a nice test a test like this is helping me a lot with questions I have
sure i am going to use it in my serie connection to 500 volt.
no i do not have all the answers but battery behavior matters..
Have you ever used that tester with EEL battery cells? That’s what I have ordered hopefully they are good cells?
Very helpful, I am from Portugal and I am building small battery bank to support peaks on my solar installation. Only 2K battery 48V. My doubts was till wich level charge and discharge. Your videos are fantastic, keep going.
Thank you very much, Abel and greetings to Portugal. Beautiful country, been there many times.
I have always questioned the "rule of thumb" figures thrown out for charging and discharging cutoff. This is the first instance I have seen of someone closely examining the charging and discharging curves. Of course, this is only the base data, I assume that you will try adjusting the cutoffs to see how much it impacts the capacity. Determining how various cutoff points affect the longevity of the cells might take longer than we have left in our own lives. Thank you for your efforts.
Thanks, Jeff. Yes, this is an older video where I started to find out what is actually happening with charging and discharging LiFePO4. I have many more videos on the channel, where I did all the testing and analysis of what you said in your comment. As I could not find any useful and consistent information online, I started my own experiments and shared the results here on the channel.
Thanks very good insights in chemistry!
Thanks Edward.
Just a thanks for your videos. Bought this, through your link, and am doing a cell test tonight. Wouldn't have got it started without your very thorough instructions, and google of course.
Charge the cell with 3.4v cutoff and then check the capacity to see if your idea is correct
Es gibt aber auch noch ein Mittelding zwischen dem kleinen 20er und dem 40er, nämlich den EBD-A20H.
Damit kannst Du auch 20A laden und er geht auch bis 30V.
Great video Andy. This is really useful information.
Thank you, Bruce!
Ancient man made fire by friction and force and and said : "I am man!". Today we make sparks made from the sun and say : "I am modern man !" We haven't changed that much but the technology sure has :) Cheers.
I print this out and it goes in the shed!
@@OffGridGarageAustralia nice - glad you liked it. I edited the original post because if misspelled change[d] - omitted the d. :) Clearly I cannot type very well.
I'm puzzled as to why Andy (or any other off grid tragic) needs to be doing all this in depth testing stuff.
1. The manufacturers of each battery factory research and provide the most efficient charge and discharge graph specs for their product. I've got mine. Why not just follow them?
2. They are deduced by tertiary-educated chemical-electro engineers, probably in touch with the best lithium chemists in the world when making their batteries. Some design batteries for space travel. Every product would be bench tested 100x more than Andy's work.
3. Despite Andy's months of garage testing, good manufacturers have teams of engineers working 5 days per week and millions of dollars of testing equipment. Why are we humble customers not just following their product spec sheets? Does Andy REALLY THINK he's smarter than all the technical people put together in (for instance) the Daly Dongguan BMS Electrics Company or Tianjin Lishen Battery Company, rated 5th in the world? These are young, cutting edge, PhD Chinese graduates, born into this era of hi-tech battery technology.
4. Won't each battery maker's product vary somewhat from the tests Andy does on his brands? I've got Lishen Cells, reputed to be one of the best. They make billions for e-Bikes. But I know there are many totally crap brands of RV LiFePo batteries out there with *recycled lithium* ... nominal 100ah but probably good for only 80; supposed to be 12.5kg of lithium, but really only 10kg. Surely those ones will perform WAY DIFFERENT to mine or Andy's.
5. I appreciate there are many lithium myths out there which need to be disproven, and many amateurs like myself thirsting for good information. But surely, "just follow your manufacturer's recommended spec sheets" is the easiest way to go. Ffs, Victron (for example) know their products better than anyone. Just explain them. No need to critique everything as somehow imperfect.
Why all this perfectionist angst? Or is it just a curiosity neurosis hobby for retired old codgers in garages who think their battery or BMS maker "might have missed something".
"Trust, but verify" 🙃
Scored an EBC-A10H for $60. Yess!
Yes Andy.
More tests, please.
More to come!
Hi Andy! I have no previous experience with any BMS device. I am confused about the operation of the JK BMS device because I have mixed up the device:
Situation 1) The measured capacity of the cells is 307Ah, but I originally wrote the BMS "Battery Capacity" value as 280Ah and I have also tried the values 307Ah and 300Ah for that point. It seems that changing this value completely messes up the calculations. A real example:
When I set "Battery Capacity" to 280Ah and "Remain Battery" = 100% and "Remain Capacity" = 280Ah. Then I change "Battery Capacity" to 300Ah so "Remain Battery" = 69% and "Remain Capacity" = 193.2Ah. This seems completely counterintuitive!
Situation 2) I charged the battery directly through its main terminals - that is, I did not put the negative wire of the charger in the "-P" port of the BMS device, but directly in the terminal of the battery. While doing this, I noticed that the voltages of the cells do increase, but the BMS does not increase the "Remain Battery" and "Remain Capacity" values at all, even if I charge with a 25A charger for 40 minutes. So the information about the actual booking amount is completely out of place?
Could there be a programming error or flaw in its software for Situation 1 that they never noticed?
But anyway, I'm asking for advice:
- Can I safely forget about these readings like "Remain Battery" and "Remain Capacity" and just focus on monitoring things like "Ave. Cell Volt." value to see what the actual battery charge is?
- Would it be wisest to use the battery in the interval where the pole voltage moves gently to decrease or increase - and not discharge it so empty, where the voltage drops quickly, and not charge it so full, where the voltage rises quickly?
- Which voltage values would be the best for the OVP and OVPR and UVPR and UVP points of the cells?
Thank you!
Now if you had a time machine, you could repeat the test after a few years use and see how well the battery holds up with such accurate data to compare, as always, thanks for the great video.
Thanks Dick. With the next batch of batteries I intend to keep a record of the graph and data for each cell, so I can refer back at any time.
chef had batery gebrrtor ôerr bsn siar iwer bsnk he fud isvyest we dud njt ned gert jñveoled
Andy you say "10% more battery capacity" is impressive. No it's not. Every manufacturer of everything specs their product with a nominal capacity with variously hidden "headroom" or "safety factor" above and below what they want the customer to use. Eg, Crane cables are some 3x the Safe Working Load. If you tested a 1 tonne to break at 3 tonnes you should NOT be impressed it can lift 300% more than the label says. Conversely, if you found a battery which only gave 100% of its specs, you should in fact be UNIMPRESSED it is graded right on its edge.
Warning, highly controversial and potentially inflammatory opinions ahead... :)
"If you're still using Fahrenheit"
Sure, Celsius more popular, but does that really make it better? Why should our temp scale be related to the freezing and boiling point of water? I'm not sure why that matters. Shouldn't it be based on the human experience? 0 Fahrenheit is **generally** about as cold as it ever gets and 100 Fahrenheit is **generally** as hot as it ever gets. We get a much better resolution that way.
thevane.gawker.com/fahrenheit-is-a-better-temperature-scale-than-celsius-1691707793#:~:text=Fahrenheit%20gives%20you%20almost%20double,than%20Saint%20Celsius%20ever%20could.
(BTW, I really don't care how **YOU** measure your temperature. Do whatever makes you happy.)
Could always add extra decimal places for resolution. Not necessarily limited to integer values, that's just a decision of the thermometer makers and not inherent in the temperature scale -
For my Winston cells, I prefer 3.45V absorption voltage and 3.375V float settings for the Victron MPPT
Same here. I have the 260Ah Winston cells (12V pack) and use the same numbers 👍
Thanks, Nick. What's the reason for that? When do you balance?
@@OffGridGarageAustralia at 3.45V absorption, you almost fully charge the cells without taking them high up the upper knee. At 3.375V float, you can keep using power from solar after the battery has been charged. Anything lower, you discharge the battery and waste solar, any higher and you start damaging the battery. There’s a consensus that under 3.4V you can’t overcharge a cell no matter how long you absorb at that voltage, while in this setup it only lasts from fully charged to sunset. These numbers come up after a period of testing and everyone ends up at those. The graphs you create show the logic here as well.
About balancing: in my opinion the BMS protections are a last resort fail safe. I never want to see them triggered. I balanced the cells manually and they maintain balance. I can charge single cells inline but didn’t need to yet.
You are finding exactly what I have found. No need to go below 3 volts, and above 3.4 volts, there's almost nothing outside that range. You are still limited by the weakest cell, but that is true no matter what charge and discharge profile you use. You can easily get 90% of capacity without placing any stress on the cells. I had found out using the bigger tester that my laptop battery won't last long enough to record a complete test, so for the last month I have been waiting on the parts for an 11th generation in NUC to be able to save the graphs. They just released it about two weeks ago, so I finished building it yesterday. Had to use a dummy HDMI plug so it can boot headless, and I just remote desktop to it. So I have only completed 2 recorded tests, despite having it for a month now (Eve cells), but have done lots of manual tests and was reaching the same conclusions as you about the time you did the videos on charging and float voltages. On a 280AH cell (measures 278.8) I am finding about 5 amp hours above 3.4, and 5 amp hours below 3.0. I think I like your 3.1v cutoff better.
Thank for sharing, John. A NUC is a good idea as it uses only a few watts of power running for so long. I've got my old Dell workshop computer running for several days now and with the cloudy weather at the moment, solar is barely enough to recharge what I take out with all my testing.
@@OffGridGarageAustralia Yes, since it uses a laptop CPU, memory and SSD, it is very good on power consumption (like a laptop). It even uses a 65 watt power supply just like a laptop. I did have to buy a 10 port USB hub, since I have a lot of programmable loads and power supplies to connect. Four of the 18 amp Riden power supplies, and four of the ET5410 electronic loads. This is for testing packs at higher rates, obviously.
@@john_in_phoenix how abt a RSPI4 instead of a NUC ?
@@cgmarch2359 Hard to run Windows software on an arm based Linux system.
Excellent Vid Andy
setting up my workshp off grid: LiFeP04 48vdc 100Ahr with MPP Solar contro ller.
From yor reasearch is the "best" to set the charge/discharge to cycle between 98 and 10% ?
My place will be 15 to 20 degrees C at all Times year.
Hello, I just purchased two of the EcoFlow Delta Pro Battery systems along with the Double Voltage Hub to join the two EcoFlows together to provide 220 volts to my 10 circuit stand by electric panel. I also purchased two EcoWorthy Solar Tracker systems and mounted 4 of the 440Watt Magnus Solar Panels on each Solar tracker system. After watching a few of your very informative videos is: What should I set the charge limit to and what should set the discharge limit to? From What I learned from your videos I have set the Charge limit to 95% and the discharge limit to 5%. I want to get the most life from the EcoFlow Units and get the most energy from them. Thank You for taking time to educate the public.
Nice video, cool tool ,
I am wondering what causes the small steps on the flat part of the charge curve. Chemistry in action?
Could be just the software and the resolution it uses to draw the graph. Also if you move the probes at terminals, it sometimes has better/worse connection and voltage goes up or down.
Andy you maybe interestedin the icharge x6 or x8 charger/discharger as it has the ability to use a bigger battery as the load, so you can do discharge tests but capture that energy in your bigger battery.
Yes yes but the new s6 is much more muchier
I looked into it but could not find any information that you could actually charge a battery with a higher voltage during the regen discharge? So I cannot discharge the 3.2V cell and charge this energy into the 50V battery bank, right?
@@OffGridGarageAustralia I keep forgetting your pack V, you would need to look at an X12 with a Volt input of 11-53V dc. But it will take the single cell output regen to wherever you choose to apply it. Not stopping smog, but cool
@@chriscord6524Yeah think you are right the x12 does have the better input voltage range as my x8 only goes up to 49v which isnt quite enought for a 48v pack at full charge.
@@OffGridGarageAustralia maybe a boost converter could be worked into it to enable some recovery of the otherwise wasted energy from running tests.
Awesome! This proves all of my hours of watching my batteries and learning (The hard way!) I also determined it was not worth squeezing every little bit out of the batteries. My settings for my pack are 50v (3.125v/cell) at the bottom. and 56v (3.5v/cell) at the top. The top voltage in my opinion could be lowered to 3.4v/cell with very little loss of ampacity. Thanks for sharing this. Very interesting!
Thanks, Dan. Yeah, that is absolutely right and you have some good settings there for your battery.
I've done a test with charging to 3.4V only. Interesting result. Video is coming soon.
Hi, I have 8 Pylontech 2.4 kWh 48v 50Ah 15 cell modules and your videos clearly demonstrate just how safe they play it with the recommended charge/discharge cut off voltages, 53.3v charge cutoff and 47.2 v discharge cutoff which gives a cell voltage of 3.5533v maximum charge voltage and 3.1466 minimum discharge voltage.
Is it possible to calculate the number of complete cycles by dividing the total kWh hours drawn from the battery bank ie 2,530kWh by the battery bank capacity ie 19.2kWh? Or do I need to access this data by connecting a lap top to the master battery data port?
Yes, these voltages you mentioned are totally save to use.
You cannot just divide the kWh by the capacity of the bank. That would give you a rough idea but usually you don't do full charges from 0-100%. If you do 10 cycles from 50%-60%-50% SOC it is far softer than doing one from 0-100%. It all depends where you do your cucles (near the peak at bottom or top) and what the DoD is.
that ßhoukd lsst oong t7me
That is a really useful tool! A very good way to test and learn how the chemistry works, keep up the awesome content!
Thank you! Cheers!
I have learned so much from you. Thank you so much. I am gaining the knowledge and confidence to build my own 280ah lipo 24v bank!
Great to hear, Tim. That is fantastic!
I guess we charge to 3.65 because that will be the highest cell voltage in the pack. other cells may still be at 3.45..
It's saving time to charge to higher voltage, that's the only reason. Absorption will take longer if we set the cut-off voltage lower.
Awesome tester, any chance you can do a 5 Amp discharge test on the 100AH lifepo4 battery, and share the csv.
I would like to see the voltages during the discharge as they will be higher than discharging at 20 Amps.
I'll sponsor the beer's for the test 👍
Absolutely, Colin. I'll put your request on the list of future tests. Will come soon and I'll make a short video about it. Would be interesting to see how much more capacity we will get out at a lower discharge rate.
Thank You
An interesting test would be how the battery reacts to both high charge and high discharge for short periods. So near the bottom and you have a high discharge how much sag and where it's soc returns to. Same with charge.
Thanks Jason, I'll wait for the bigger tester for this test.
Great video but darn it! Now I need to buy this tester, it is all you fault and the Frog.
Bud, wait for one of the next videos when I show the gib one 😉
I just got my "new" battery cells from AliExpress and was obviously ripped off. Could you show the "resistance" test of your small (20A) battery tester? I own the same and would like to know which values to expect in new batteries. Depending on which current I use the result is between 3 and 10 mR, which is is way too high in my opinion. But maybe this tester doesn't work fine-grained enough.
I have a request Andy, when mentioning the temperatures of your equipment could you put down in the corner what that is in Fahrenheit? Celsius doesn't mean anything to me as I'm in the US. Thanks so much for your videos, I've learned so much from you.
Yeah, I mostly do that and most specs are in °C so...
0 - freeze point
15 - cold
25 - ok
35 - hot (human body is 37)
45+ very hot
Hi, thx for the video - have same device for testing..since the EBC-A20 offers also just to monitor cells - will it work with an external device for charging/discharging and if so, how many "external" amps it could coop with? the 5A loading and even 20A discharge is a pain for 310Ah cells...
I clicked the like button 5 times, but it only registers once... Doh.
You need to try harder Paul!!!
Or I mention aluminium busbars again and how great they are 😂
Brilliant, however I wonder if you could take your viewers through the whole process of setting up the tester and software as I also would like to obtain the big brother of the tester you have used in this video. Thanks, Ken in England
Andy, It's you! How things have changed after the Outlander! Best wishes, John Poxon
I am planning to build a 12v battery, only have a dc dc converter, wondering if this would be good enough to charge or donyou need something that stops ats the current drops?
A DC to DC converter will work if setup correctly. I have used one many times to charge a 12V lithium battery. Important is that the battery still has it's own BMS as protection device. Set the DC converter to 13.8V and let the battery charge until the current is almost 0A (or under 5A for a 100Ah 12V battery). It's then charged 99.5%. Remove the inverter afterwards.
Your right there is no need to charge at such high voltages. It is only pounding your battery. Returns in amp/hr for life taken is not justified.
Which model battery are you testing?
this tester will work on 24v100amp ready made battery thank you andy i ordered should get mon oct 4 dhl please let me know keep up the great work
how is the discharge voltage variation vs time of a best quality battery and a poor quality battery when the same load is connected?
Lots of videos here on the channel showing that! Also a mix of different cells capacities in series in one pack.
thanks for this insightful video, bud
How do you calculate the cut off voltage?
I'm thinking of getting one of these testers to test the 24v batteries we use at work.
That depends on your battery and chemistry if lead acid, li-ion or LiFePO4.
The results are good for general guidelines but not much more than that. Results will be different between different lots or different manufacturers 100Ah batteries and will probably change for used batteries.
im sure you already know this by now and im sure others in the coments have said curves and capasity will change with temprature.
No surprise that at a 0.05C discharge rate that you get better than 100AH. Are you happy with the smaller cell or do you regret not getting an Eve LF280AH cell for testing? Nice work
He discharged at 20A, that's 2C, no?
@@RickinICT No, 100A would be 1C. 20A= 0,2C 5A=0,05C
Discharge was 0.2C charge was 0.05C. And Andy kept saying that's why he has ordered the bigger tester, to do 0.2C charge as well.
Where can i find copies of the data (charge/discharge) that you generated. The curves were quite interesting. Seeing what the cycle degrade looks like would be interesting. Why put in more energy than what you take out---but, this probably changes so that's probably why they charge to higher voltage points. This was interesting. Thanks
Great video, very useful! I would love to see you doing the same test on one of the EVE280 cells! I am in the middle of top balancing my 16 280’s. All in parallell with 10A, they have stayed at 3,35v all weekend now. 😂 Starting my setup with Multiplus and other charge controllers very soon, and I am very uncertain how high to set the cut of voltage. As mentioned in the comments, balancing can only be done at the high (or low) end. But in the other hand, I do not want ro stress the cells. I guess I will end up with a fairly low cut of voltage (3,4?) and then adjust that to 3,65 in between, if/when the cells get out of balance.
I really love your videos, it makes my day. Keep exploring! 😊
Thank you Lars. I've got the bigger tester now and more batteries coming in, so stay tuned.
One test that I'd like to see with these cells that I can't find anywhere is... Can we boost up the voltage with a step-up module to 24v, 36v or 48v for ex. for an ebike battery using only 1 or 2 cells. So 3.2 or 6.4v boosted to 36v with the stepup module in order to make a small simple battery with few connections like the 18650 lithium cells.
Would love to know if this is possible to do.
Hi CK, while anything is technically possible, this idea would require a very hefty booster circuit to achieve the results ur looking for on an ebike & that would probly get pricey. Remember v=ir, so this 100Ah 3.2V Cell has roughly 320 watts, so say ur ebike motor is 32v and uses 10A, that’s 320 watts right? Not really when u go thru a soft grassy patch or up a hill, but also remember ur 3.2v cell now has to put out at least 100A continuous. I like ur idea for simplicity, but I just dont think it would be very viable the way electronics are presently. Cheers
@@evil17 alright, thank you very much for your answer 👍
Don't worry about voltage. Worry about high State-Of-Charge (SOC), THAT is what causes damage to Li-Ion cells with graphite anode (including LiFePo4).
And another observation:
Charge with 5 Amps to 3.4 volts, no absorption, gives you about 98% SOC
Charge with 20 Amps to 3.4 volts, no absorption, gives you about 70 - 80% SOC, a bit unpredictable.
Later edit: It is true that using higher voltages increases the risk that the battery would charge to 100% or maybe even overcharge a bit.