Lithium Battery State Of Charge VS Voltage Drift. System Issues? Several Solutions
ฝัง
- เผยแพร่เมื่อ 18 พ.ค. 2024
- Lithium Battery State Of Charge VS Voltage Drift. System Issues? Several Solutions. OUR SOLAR EQUIPMENT......CLICK HERE: www.signaturesolar.com/?ref=c... COUPON CODE for $50 off: Countryliving12
HARVEST RIGHT FREEZE DRYER: affiliates.harvestright.com/1...
GROWERS SOLUTION: growerssolution.com/Countryli... DISCOUNT CODE: CountryLiving10
HENGEAR: hengear.com/?ref=countrylivin... Coupon Code: countryliving15
SEEDTIME: seedtime.us/jkmel4ukl4tonac09... ($5 in FREE seeds)
Solar and Electrical Equipment:
EG4 6000xp Split Phase Inverter: signaturesolar.com/eg4-6000xp...
Victron Multiplus 2: signaturesolar.com/victron-mu...
Victron Smartsolar MPPT: signaturesolar.com/victron-sm...
Victron Cerbo GX Comms Hub: signaturesolar.com/victron-en...
Victron Cerbo Touch 50: signaturesolar.com/victron-en...
Victron GX Touch Wall Mount: amzn.to/3ruaXR7
Victron Lynx Distributor: signaturesolar.com/victron-en...
Mini Rail Roof Mount: signaturesolar.com/asphalt-mi...
Solar panel rail clamp: signaturesolar.com/end-clamp-...
EG4 6500ex Off-Grid Inverter signaturesolar.com/eg4-6-5k-o...
EG4-LL LiFePo4 Batteries: signaturesolar.com/eg4-ll-lit...
Growatt SPF 5000 US Off Grid Inverter: signaturesolar.com/growatt-5k...
IMO PV Disconnect Box: signaturesolar.com/dc-disconn...
Nader 200Amp Breaker: signaturesolar.com/nader-dc-c...
Growatt Transformer: signaturesolar.com/growatt-5k...
12 position electrical sub panel: amzn.to/3DVKZpZ
4 gauge THHN wire: amzn.to/3pYrXuB
4/0 and 2/0 Battery Cable: amzn.to/3QXfZgy
Copper Battery Ring Terminals: amzn.to/3CLRK0y
70 amp double pole breaker: amzn.to/3AvMVWp
Safety switch: amzn.to/3FXCkHO
conduit as needed
Simpson Flip Toggle mounting hardware: amzn.to/47AFhtw
Ticonn IP67 Electrical Box: amzn.to/3YBFKHX
Dinkel Terminal Blocks: amzn.to/3QMq6Yj
DIN rails: amzn.to/3OxbpFS
Hardie backer board
Pike Industries Bus Bar: amzn.to/3R3bSzs
EG4 Brightmount Rack: signaturesolar.com/eg4-bright...
Parts list:
1 1/2" conduit
90° conduit elbows: amzn.to/3qL33Tm
conduit LB boxes: amzn.to/45SEWAR
conduit threaded connectors: amzn.to/3OTCbZc
lock nuts: amzn.to/3P3bCAO
thread protectors
Conduit weatherhead: amzn.to/3L61WnV
Conduit straps: amzn.to/45S6iHf
grounding rod: amzn.to/3LspxzF
ground rod clamp: amzn.to/47Yzi1O
sub panel: amzn.to/3DVKZpZ
AC breakers: amzn.to/3AvMVWp
Safety disconnect switch: amzn.to/3FXCkHO
Grounding bar: amzn.to/3LspDY3
Neutral Lugs: amzn.to/45U4Iog
Ring terminals: amzn.to/3EfLRGK
Ferrule Kit: amzn.to/3sCgh5p
PVC Electrical Glue: amzn.to/3OWoLvz
1/2" flexible conduit: amzn.to/45OyXNt
Liquid tight conduit connectors: amzn.to/45OyXNt
Distribution protection box: amzn.to/3P2t7Bq
Dinrail: amzn.to/3OxbpFS
Dinkel terminal blocks: amzn.to/3QMq6Yj
DC breakers: amzn.to/3L4iPQ0
Surge protection device: amzn.to/3qTDuPQ
Combiner box: amzn.to/3YBFKHX
Solar inline fuses: amzn.to/3PnRZVK
Flip toggle mounting hardware: amzn.to/47AFhtw
MC4 connectors: amzn.to/3qWiW9s
WIRE! : amzn.to/3pYrXuB and amzn.to/3QXfZgy
All links are affiliate links that we earn a commission from. There is no extra charge for you at all but it helps support our channel. - แนวปฏิบัติและการใช้ชีวิต
![[Full Episode] MasterChef Junior Thailand มาสเตอร์เชฟ จูเนียร์ ประเทศไทย Season 3 Episode 3](http://i.ytimg.com/vi/A1XXYbV048k/mqdefault.jpg)
I so appreciate you mentioning that the math doesn’t always add up and that the calculating can sometimes seem impossible. I have always appreciated how well you describe what is actually going on with the elements of the system. Your manner of communication makes it very easy for me to understand and I appreciate it BIG TIME!
You’re welcome. Glad I could help
Lipo4 12v battery has 4 cells , 24v has 8 cells , 48v has 16 cells.... so on and so on.
Look at each 4 cells as a glass of water , you can drink whats in the glass and no more.
You can fill it to the top , and its full.
Now connect multiple glasses together (100 glasses) by 1/4 inch straws .
Thats a lot of water but limited by the ability of the water to flow thru the straws.
Manufaturing of the straws has measused tolerance, not all the same.
Battery banks will never perform equally.
Best course or action to fix the problem.
Move up to bigger glasses , maybe a swimming pool and best , dig a lake.
Theres no replacement for displacement.
@@vinny143 nice addition 😁😇😁
I just prefer open loop. I run many systems this way, and it just works. Communication is just another thing to complicate and go wrong.
Same here. Voltage is pure. Furthermore, at least in my case, each battery ( I have 9 for a total of 135kWh ) will behave a bit different. As you head up or down...near the high or low... they'll equaliz. I do agree that voltage is not precise until you get near the top or bottom end of the range. However if you get the inverter capacity set up right...the inverter will try to give a soc that is based on actual power in and out of the pack.
I have Solark and SOK but the issues are the same. IMO voltage mode(you mentioned lead acid) is the only way to go, and you need to set the inverter to shut down before the BMS. SOC is a myth and only a rough estimate. it drifts off about 1% each cycle, and need to be at voltage limits to reset. So your 'drift' is because you never are allowing the SOC algorithm to know where it should be until it hit in your case a Real 0% which is determined by the voltage curve shoulder. Until it hits a shoulder the bms is guessing. instead, i watched the curves and set the float/absorb/low/shutoff based on where the shoulders are. once a month i allow the voltage to go a bit higher for a couple minutes (in solark , this is the equalize voltage) and this resets the soc in all the battery bms units. for the sok batteries, it seem the voltage might be a bit more. I start the generator at 50v. the bms will not reset at 56.5, but will at 57.5. the shoulder occurs for me reliably a 55.5 on the high side. also talk to Dexter at current connected about the myth of 20-80% for Lifepo. its ok to hit "100%" for short periods, and in fact its neccessary for the bms to know whats going on.
Interesting. Thanks. Good to know that this is across the board with different brands of equipment.
I will attempt another top balance with the Chargeverter to try to reset.
@@CountryLivingExperience As a minor point, I would not call this a 'top balance'. Top balance really adds energy to the battery. The extra voltage in this case isn't really resulting in extra energy going into the battery (only a very very minimal amount is involved). Instead, it's telling the BMS that because the voltage is rising rapidly despite very little current going in and therefore the 'SOC' is by definition at its full state. For example, you will notice this where one second the SOC reading is way off (e.g. 75% or so), then a second or 2 later it is at 100%. It is not that 25% of the stored energy magically appeared - instead that the BMS realized the gauge was off.
@@mm-zw1zc I understand. Thanks
Hey I had the very same issue and what I did to remedy it was to bring all the cells to a higher voltage which for mine was 56.0 volts then I let them balance for about 2 hours within 30 minutes they all balanced out. And each battery read 100 percent SOC even the one that was stuck at 80 percent. I then purchased a Victron Smart shunt which allowed to count for more of an accurate state of charge for the entire stack and I disabled comms and allowed solar assistant to use the victron shunt to measure SOC for the stack and it has been very accurate! As it counts every amp in and out. And it works with the 6000xp with solar assistant just need to get the raspberry pi or orange pi and a VE direct cable and then you can make changes and monitor from anywhere in the world!
I did that type of top balance in the past and it didn't make a difference. I may give it a shot again.
Glad you got that more accurate reading with the SmartShunt and glad to hear it works with the 6000xp. I'll look into the raspberry and orange things but I don't really know what those are.
Thanks 👍
Great timing for this video. I have dual eg46500ex spit phase and the bms comms was kicking my inverters to grid even though I don’t have grid connected. SBU was used. Once I switched to USER and set proper charge parameters it solved the issue. I got the EG4 HUB and found all 6 batteries were fine cells were only .01 off. Its still a mystery what caused the issue. I have been researching the voltage drift and trying to get more info. Im going to solve the problem and move to 6000xp. Your channel has helped me in my solar journey thank you for sharing.
Cool. Glad you are finding solutions to the issues you are facing. Glad we could help as well.
Seems like they used a bms from a battery type that is full at 57 volts or used those parameters. My lead carbon bank is 100% at 51.6v my soc shows 69% still charges proper, last time I checked all 16 batteries all within .1v with no bms or equalization possible, -30c charge np. 48 kw weighs 2000 lbs more than liPO, as long as it works from day1, having a failure is rare.
I recomend to measure voltage on inverter not SOC. Proper voltage is what is key to protect batteries. Than add a device that measures energy in and out of battery for SOC, it can be separate, or connected over RS485
I appreciate your analysis very good. Question how did you determine which jumper cable was bad? I have 12 batteries but it only reports 600 amps and I think that should be 1200 thanks great job
You’re welcome. It was trial and error with the cables….process of elimination.
Great video, not hearing much coverage of this issue that seems to happen with all these Lifepo4 batteries. I have 4 chins 48v 100AH tied in parallel and I get one battery showing 82% the next showing 84, 88,90. Same issue voaltage is the same across the units. This is total voltage though. When they reach 100 some of them snap in and say 100. Sometimes I wait until a few hours after dark after the bank has rested and use the XiaoXiangElectric app which is free to reset the capacity. Are EG4 Chinese made as well? I Bert the app can connect to that BMS to reset. Thanks for covering this.
You're welcome.
Yes, the EG4 are Chinese made. Hope you can get yours worked out.
Your charging system needs to top out at 56.0V and be held there for some length of time so your BMS can complete cell balancing. Seems to me that might be your problem.
Itll never be fixed.
Degrigation starts at manufaturing completion.
BMS is just trying to patch the holes in a sinking ship.
Thats why lithium cells have a dated life cycle.
@@vinny143 No, this isn't how LiFePO4 works. The cells and batteries are fine. The charging parameters or the BMS parameters were just misprogrammed. They will rebalance with a few days worth of cycling with proper settings.
--
In anycase, if balancing is the goal, the charge target should be set to 3.55V/cell (56.8V) and held there for roughly 2 hours per charging cycle. The packs will be in balance again within a week.
56.0V (3.50V/cell) will work only if the BMS's in the batteries are using reasonable settings for the float start voltage... which is a bit hit or miss so if one can't verify the BMS settings, use a 56.8V (3.55V/cell) hold for balancing.
If one can verify and/or adjust the BMS configuration in all the batteries, one can fine-tune everything and use lower voltage targets (such as using 3.45V/cell for the target and start balancing at 3.40V/cell). But... unfortunately most people who try to mess with BMS and charger settings do so with an incomplete understanding of how the cells and batteries work, and wind up massively unbalancing their batteries instead.
-Matt
I used Solar Assistant with the Victron shunt. When the Victron read 70%, I had SA change my MPP unit from Inverting to Charge/Pass Through and then back to Inverting when it got to 90%.
If SA can control the EG4 units, you could do the same. The trick is to have SA read the Victron Shunt for the battery (and not the EG4)
Thanks for the heads up
note that if you are using a shunt to measure, you will need to account for inefficiencies. The round trip is not 100% efficient, so the amount you put in must be more than you pull out by a few percent. you still need to watch voltages.
Here are the Victron shunt settings for LiFePO4.
NOTE: Some of Victron's suggestions in their manual for LiFePO4 are INCORRECT. I have listed the correct settings below.
The coulomb efficiency is very high so:
Peukert exponent: 1.05
Charge efficiency factor: 99%
Battery Capacity: (as appropriate. e.g. 100Ah for a 100Ah battery)
Typical settings for resetting the SOC tracking to 100% properly and for the display:
Charged voltage: 55.2V (3.45V/cell)
Tail Current: 2%
Charged detection time: 3m
Discharge Floor: 10% (very typical, but people have their own ideas)
TTG averaging period: I recommend 5m.
And most people use 0.10A for the current threshold.
Battery SOC on reset: Typically use "Set to 100%". There is no "right" setting for this.
-Matt
@@junkerzn7312Agree. The 'right' way IMO is to know how your system performs by looking at the voltage 'shoulder' or where the voltage starts to change rapidly relative to the same current flows. These points are IMO what 'define' 0% and 100%. In my experience, the 100% is pretty easy to define based on this. The 0% is a lot harder because the shoulder isn't as sharp and is more affected by things like temperature, but for my setup I call this at when I get to 50.0V when drawing about 1a/battery (which is my typical load). I could go a bit lower here, but start getting nervous that a bigger load could come on and drive it lower more quickly.
✌️😜 Like you said Brother, awesome real-world info on a working system 😁💒 God Bless us All 🇺🇸🙏✝️
I have the same V1 of the EG4 LL batteries and have a similar, but different issue that I have previously reported; I have batteries reporting 100% when the voltage is down in the 48 volt range (not 100%); I have to top them off periodically with an EG4 charger or eventually the inverter will trip an alert for low voltage and run a recharge cycle from a grid connection. The alert clears when the battery is fully recharged.
You are one of the few that have commented that understand the drift. Mine were doing the same. I think, like you said, it is a periodic top balancing issue. I may have to top balance mine again to pull the SOC and voltage back into alignment.
@@CountryLivingExperience - I'm still using the Growatt inverters also, so I think that confirms that it is specifically battery firmware and NOT the inverter hardware/software issues. I have NOT noticed the drift yet on the "production system", which is 100% off grid with NO grid connections, consists of (5) EG4 LL batteries, (2) growatt 5000 watt, 240 volt inverters, each with their own split phase transformer and about 4000 watts of solar panels. My "test system" consists of (1) growatt 120 volt, 3000 watt inverter with (1) EG4 LL battery and NO solar. This is grid tied and setup to function as a UPS. This system has the reported drift and about every 3 months, it throws a low voltage battery light error and goes through a charge cycle (from the grid) and tops it off without intervention (if I don't manually do it with an EG4 charger). I have thought of using the victron shunt like you, even if it is only to confirm the drift by seeing over time that the batteries are not being fully topped off (wattage discrepancies on the victron shunt).
I may just kick mine back into lead acid mode so I can set the voltage parameters instead of using soc to determine low cutoff. The soc from the batteries with comms enabled just seems too unreliable. I may also have to do another top balance.
when your system shut down, how did yu get it fired back up? That is almost a black start right?
It was almost a black start. There were a few that had about 15% left. It took a long time for it to charge though. About two days.
With lipo4 cells all power is delivered at a very flat discharge curve.This is probably the differance as the BMS counts amps in and out not very accurate.This does not mean battery is at different states of charge just the estimate is out????
Yep, that is what I was talking about.
Hey, thanks for this video. I'm dealing with a persistent '90% +/- charge on my meager 12v system. This helped me look closer at the probablye cause.
Probable cause... stupid YouStup won't let me correct.
You’re welcome
I am facing the same problem from the day i bought my second battery and put the 2 in parallel.
Another problem i have though is that the 2 batteries are charging/discharging at different rates. The difference can even be 3 or 4 Amps for a total load of 30Amps. Which is a lot! Have you ever measured charge/discharge rates on each battery to see if they have different rates?
I can see the charge/discharge rates on the screens of each battery. The battery bms should put the load where it is needed most......that is if you have communications set up between the batteries.
@@CountryLivingExperience BMS is not always accurate in Amp counting. But lets say BMS is the same for each battery . Do you always see 10Amps on each battery for a 90A load?
@@Theo2482 I know that. I don't monitor mine that closely for amperage draw on each individual battery.
@@CountryLivingExperience If there is a different amperage draw on each individual battery then the SOC issue should be partly because of that. Let's say 2A draw difference can make a big change in SOC across the individual batteries over time.
thanks for the info...I still run lead acid as weight is not an issue for a home grid, can "overcharge"with less risk, have outside battery bank in below 0, and to me it seems lead acid can be more of a "dumb" system. I get annoyed with my smart phone and the ideal of firmware hassle is just a big turn off...really hope we get a different commercial chemistry soon that doesn't require a computer software engineer to get it to charge...anyway thanks again for reminding me that lifepo4 is beyond my comfort level....what I don't get is why don't your batteries even out as the 51v nom should be across all series cells with close to even internal resistance? I guess I just don't understand the BMS systems at all.
I totally understand your point of view. I wish systems were still a little more “dumb”. Sometimes the cells even out in the middle of the charging/discharge cycles in the middle of the curve.
old timer - god bless guys like you , im 68 but living in 2024 ....❤❤❤❤
@@overthetopcargotrailerjerr2278 I'm a young 45 and I miss some of the old electric mechanical systems I worked with in the early 2000's for HVAC. I could "see" the diagram and tell how it worked with the switching. I guess it is easier to let the board do all the thinking and just swap boards when they get fried but it is a pain when you need to wait for the week to get the board shipped.
Ok, a few things here.
* First and foremost, the shunt calculates the state of charge ONLY by counting the coulombs in and out of the battery. Its counting current over time, basically. The only time a shunt (in the BMS or the external shunt) EVER uses voltage is when it is determining whether the SOC should be reset to 100% or 0%, depending on how it is setup.
This means that for LiFePO4 you generally want to charge the battery to full. This is true for EVs with LFP batteries too, by the way. It just isn't possible to track the SOC exactly.
If you let the SOC drift for too long, it will basically turn into noise... completely garbage and not tell you the actual state of charge. Whether you use a Victron shunt or depend on the BMS, doesn't matter. You solve this by charging the battery to the appropriate target voltage at least every once in a while (like once every few weeks at worst). Most people just charge to the appropriate target voltage as their go-to configuration.
The correct charge target voltage for those EG4's is 3.55V/cell = 56.8V, with a 30-minute to 2-hour hold or current-tail setting.
* For LiFePO4 voltage is almost completely irrelevant with regards to figuring out the SOC until the actual state of charge is roughly over 95% or under roughly 5%. Those are the only times where the voltage gives a good indication... only when the battery is almost full or almost empty. I'll talk more about that later.
* Any current going into the battery will push up the voltage roughly scaled to the current, and any current going out of the battery will pull down the voltage also roughly scaled to the current. Thus, unless the current is zero, you can't determine ANYTHING from looking at the battery voltage.
* LiFePO4 cells will naturally float when not loaded or charging, and they will do it at slightly different voltages. Because there are 16 cells in series inside a 48V battery, this means that individual batteries can actually vary by up to around 0.5V, sometimes even more, and still all be at the same state of charge.
* TOP balancing: All you do for the EG4 batteries is charge the battery to 3.50V/cell (56V) or higher. Best to charge to 3.55V/cell to ensure that the BMS's start balancing., Proper BMS configurations only start balancing once cell voltages are (roughly) above 3.45V/cell because trying to balance cells below that voltage does not work... will take them out of balance instead of putting them into balance. But if you do not know what the BMS's configuration is, charge to 3.55V to catch the edge cases and ensure that balancing happens.
* BULK, voltage targets. For LiFePO4, any voltage (at zero current) >= 3.45V/cell is 100% SOC. Most people set their bulk target voltage to 3.50V/cell (56.0V) or 3.55V/cell (56.8V). The BMS itself disconnects at 3.65V/cell (58.4V). NEVER SET YOUR CHARGE TARGETS TO THE BMS DISCONNECT VOLTAGE. You don't want the batteries constantly disconnecting as they get full. It is completely unnecessary to charge the battery that high, it will be at 100% SOC well below that voltage.
Typically for charging you set the bulk target and either set a current tail or a fixed time period to hold the target. For a fixed time period, use 30 minutes to 2 hours. Equalization is not used, turn it off. LiFePO4 doesn't need equalization at all.
* FLOAT voltage. For LiFePO4, the ONLY valid float voltage range is 3.35V/cell to 3.375V/cell. That is 53.6V to 54.0V. Once charging is complete and the battery is at 100% SOC, the charger will stop pushing current and allow the battery to drop to its float voltage before applying any load support (holding the battery at the float voltage, if you still have solar available to do that).
The battery dropping to the float voltage DOES NOT CONSUME MORE THAN 0.5% OF THE SOC. The battery will STILL BE AT 99%+ SOC EVEN AFTER DROPPING TO THE FLOAT VOLTAGE.
* When discharging, the battery will generally drop into the 3.3V/cell (52.8V) range and for most of the discharge cycle it will be between 3.2V/cell and 3.3V/cell. At 3.0V/cell (48.0V) the battery is 95% discharged... SOC will only be roughly 5%.
Most people cutoff loads at 3.0V/cell (48.0V). There is no need to stress the cells by letting the battery drop to the BMS low cutoff voltage which is typically 2.5V/cell (40.0V).
--
Voltage Drift: There is no significant voltage drift for LiFePO4 with age. That isn't how LiFePO4 aging works. LiFePO4 cells simply lose capacity with age.... fewer amp-hours can be stuffed in and taken out of the battery. That's it. And with well taken-care-of batteries it takes a very, very, very long time to lose any capacity after the 2nd year. Usually some capacity is lost in the first 2 years (usually less than 5%). After that, capacity generally is lost only very very very slowly.
There are people with 15-year-old LiFePO4 batteries that are still pulling 95%+ of their original capacity.
--
OUT OF BALANCE CELLS
If you do not charge the battery sufficiently to trigger internal battery balancing by the BMSs the cells will slowly go out of balance. When cells go out of balance, the battery will begin disconnecting at lower and lower voltages while charging, and at higher and higher voltages while discharging.
To fix this, always charge to the proper charge target and hold the battery there. Just increasing the hold time to 2 hours and letting the battery cycle/operate normally for a week will do the job. You don't have to do anything special to fix this problem. The battery's BMS will fix it.
If this is happening to you, then you are using incorrect charging settings and these are both preventing the SOC tracking from working properly and also preventing the battery's BMS from properly balancing the cells.
--
COMMS
Comms are not really necessary for controlling even large battery packs, if they are in good shape. Comms should only be actively used to reduce charging current when it is detected that one or more batteries have disconnected, so as not to over-current the remaining batteries.
The only other use for comms is for display purposes only.
SOC values should NEVER be used for any servo, active management, or relay control operations. Never, ever. Always use voltage (typically relegated to voltages = 3.50V where it is obvious that the SOC is low or high).
-Matt
That is a lot to digest. Thank you for sharing.
It sounds like you are saying I need to abandon comms and set the 6000xp to lead acid mode where I can manipulate the charging parameters (voltage)? As far as I know, I have no control over those when I am in the Lithium Ion mode because it is relying solely on the battery bms'.
In your reply you talked about not mid-ranging the batteries. What did you mean by that? Did you mean that you should not continually charge/discharge the batts within a certain soc? Say taking the batts down to 45% then charging them to 80% then discharging, etc and so on?
You had me confused on that one because one cannot predict the amount of sun charging on an off-grid system. I never know if I am going to get to "100%" with some partly cloudy days and conversely how low I will get with numerous stormy days.
@@CountryLivingExperience Not trying to hold the battery in the middle of the SOC range to lengthen its life-span. It is totally unnecessary to do that for LiFePO4 due to the float relaxation of the voltage. Even for NMC one can charge to 70% without almost no damage to the battery.
(p.s. for anyone who owns a LFP EV its the same thing... don't be afraid to charge to 100%, but if you are really worried then still charge to 100% and just drop it down to 98% after fully charging by driving a little, before letting it sit).
Not reaching 100% regularly will not hurt the LFP battery at all, but it will cause the SOC tracking to more or less turn into gibberish over a month or two of cycling without hitting 100%. Also, the battery won't be able to balance its cells but cells in a well-balanced battery takes a lot longer to go out of balance... usually 4+ months.
So if you want the SOC to be accurate, you do want to charge to 100% every once in a while. But don't worry about not being able to do it very often across winter (for example). It won't really hurt the battery and even the cell balancing won't matter a whole lot since you aren't trying to utilize the battery's full capacity across winter anyway.
This does inform you on how much "hold time" at the charge target you want. My recommended range is 30 minutes to 2 hours. If you rarely get the batteries to 100%, use a 2-hour hold. If you often get the batteries to 100%, use a 30 minute hold. I usually do not use a current-tail at all... I disable the current-tail function because I want there to be a deterministic amount of time for cell balancing on those occasions where the battery is able to be charged to full.
-Matt
@@CountryLivingExperience Its totally up to you but I don't use COMMs in setups where my maximum charge/discharge power is
Thanks Matt. I appreciate the conversation and information.
I would like to see a comment from Signature Solar. To verify what’s happening and why. Is now your battery’s out of warranty because you have gone below 20%?
This is not an issue with sig solar per se but most, if not all, lifepo4 batteries. This happen with SOK, battleborn, Lion, etc.
They are not out of warranty because of a one off event. If I had them set to 0% and purposefully ran them to 0%, then that would negate the warranty.
Signiture solar can not predict the variable of circuit components.
Nice video.. what is the red ball above the one inverter? Did you do a video on it? :) and why is it only above one inverter?
It is one of those powder filled fire extinguishers that blows open when flames hit it. I only purchased one for the space.
Do you have a link for that product
Sure. Here you go.....amzn.to/3yrjAPI
SoC is just unreliable but as you say, the LifePo4 chemistry has such a flat curve you can't determine the 70% mid-range Soc by voltage alone. I run an 18650 INR chemistry (3.0v to 4.2v/cell) and use voltage alone for my management and ignore SoC. Back to LifePo4.... it may be possible to ignore the mid-range SoC operationally (who cares about 40% vs 45% Soc) and concentrate on max charge and min cut-off... and do it by voltage. The groups of batteries in parallel, because they're in parallel, will be the same voltage (definition of parallel'ed batteries) so you can treat the overall powerall as a single voltage point at the inverters and use the lead-acid or user defined settings for min/max operational on/off. This makes the SoC info only and will allow operations between voltage min/max voltage - which is wide enough for effective/safe operation. I think a min (in the relm of 3.0v/cell or 48v) / max (in the relm of 3.4v/cell or 54.4v) is wide enough you can maintain
Yep. If I do it by voltage, I would need to run it in Lead Acid mode but I am not sure f that shows soc or not. But I could set a low cut off voltage.
Most of the batteries are fine but I have a few outliers.
The concern is at night when I don't have incoming PV watts and I am asleep. This is critical if there is a shutoff for low soc (whether correct or incorrect).
I'll do a follow up sometime. I think by moving the batteries with soc that fluctuates too much to the end of the stack, I may be able to mitigate incorrect readings. Only time will tell.
I'll second this, though use the LiFePO4 ranges I put in my main posting. Basically, nobody should ever try to mid-range the SOC for LiFePO4 battery packs. It's a huge mistake to do that and will only cause the packs to go out of balance... which actually stresses them MORE rather than less because some of the individual cells will wind up at the cut-off (low or high).
Unless the temperatures are really cold, one can optimize the charging parameters and the BMS settings to turn on balancing at 3.40V/cell and then use a 3.45V/cell (55.2V) 100% SOC target. But to be honest, most people mess up the configurations trying to do stuff like that so I don't recommend those settings to DIYers..
It is far easier to just use a 3.55V/cell target (56.8V), which won't hurt the battery at all because it just isn't going to be held very long before you allow the battery to drop to its float. 56.8V virtually guarantees that the BMS will balance the cells properly with whatever its default configuration is.
The most critical setting for LiFePO4 is the float setting, which MUST be 3.35V-3.375V/cell, no exceptions. This voltage range will hold the SOC at 99%+ (after charging to full and letting it drop), and de-stress the cells at the same time, without current sloshing causing any aging issues.
So basically you charge to the target, hold for the current tail (or for a fixed period of time, 30 minutes to 2 hours), THEN you let it drop to the float. Most of the time the battery is going to be at the float or discharging lower at 0.2C or less, which are the least stressful conditions for LiFePO4.
--
Now, going for a mid-range SOC... there's a lot of theory there, and theory is correct for NMC, NCA, Sodium, and numerous other chemistries. But not for LiFePO4. For LiFePO4 the cells work a bit differently. Even for LiFePO4, yes 100% SOC will put a little more wear on the cells, but since the voltage can drop all the way down to 3.35V/cell from the charge target of 3.55V/cell without losing any SOC, this de-stress condition makes a huge difference on the effect.
Because of this, nobody should be worried about trying to mid-range the SOC. ALWAYS have your charging target be 100% (typically between 3.45V/cell and 3.55V/cell depending on the BMS configuration). Any voltage >= 3.45V/cell at 0 current is 100% SOC at nominal temperatures (you need to go higher if it is cold), but that dratted BMS configuration might not deal with the situation properly. Hence 3.55V/cell (56.8V) is the go-to charging target that should be used.
The other thing to remember is that high temperatures are a major factor on battery calendar aging. Calendar aging for LiFePO4 is a function almost exclusively of temperature. So if you can put the batteries in an environment within 20-30C or so, and only charge/discharge them at roughly 0.2C most of the time, LiFePO4 will last a very, very, very long time. Longer than 15 years, without losing very much capacity.
-Matt
you are over thinking it. The BMS if not programmed right can in fact cause cells to get way out of balance. doing a top charge and balance with BMS setup correctly will get them all within .005 within all cells. This is why I built my own and use JK BMS so I have full control. if your charge voltage is 55.2 then balance should start at 3.45 and stop at 3.40 This allows BMS to take from the fully charged cell and give to all others. It took couple weeks for mine to top balance themselves this way. Yes 3.45 not 3.65 since that gives you about 10 more energy at the cost of over working your cells which causes swelling and heat issues which leads to lithium breakdown. I also use all in inverters instead of small MPPT controller like Victron because when my packs hit top charge they turn off and my inverters hold any solar voltage till I can use it and small MPPT units are not good at that and will burn up.
Degrigation will happen regardless on how you use the batteries.
@@vinny143 Yes but how fast. Mine are 8000 cycles to 80% I truly do not expect them to last that many cycles. 19 months since turned on and started to use I have 87 cycles used. 1,218,000 watts put in and out of each battery pack so far. At this rate if they can do it I will be long gone so I will never know how long they truly last.
@dc1544 lipo4 have been out On the market long enough to know a good estimate.
When the charged particles pass from one electrode to another during charge or discharge , the electrode , by what i understand , has the same structure as carbon fiber , starts to form microscopic cracks that the charged electrolytes get trapped in.
This is your degrigation.
If you charge to 3.45v per cell (13.8v on a 12v battery) and dont go below 3.22v per cell , the battery will last longer.
Most cycle life is measured from peak charge to peak discharge.
By adding more capacity (battery) , you will lessen the abouse on them and get a longer lasting battery system.
Most batteries meause life cycle "full charge to full discharge "...
And even at that rate , you'll still have advertised 80% capacity.
Its not like they are useless.
My system has 3 times the amount of watt hour storage , over what i use.
Battery back up , at a minimum , must run your appliances for 18 hours. (3pm -9am).
Longer the further north you are.
You can double the solar panels on an inverter by placing a junction switch between panel n charger.
This way , on a cloudy day , you'll still have solar coming in.
I need to post a video of my lab , maybe this weekend.
8000 cycles ÷ 365 days in a year = 22 years
@@vinny143 I agree with almost all you said except My battery banks are big enough that it takes 5-7 days for 1 cycle. Which I can not wait to really know how long these will last, will they just degrade inside after say 20 years or can they last 50 years?
@@dc1544 im going to research components of a battery
Thats why you need a , watt / amp meter....
And a note book.
I taught electrical theory / math, and you must absolutely use math to calculate capacity.
If a 100ah battery uses 100ah in 10 hours = 10 amps per hour.
If after 10 hours , your usage meter reads 50ah available, you used 50%.
Keep a notebook.... or plug it up and forget about it.
Warning lights will go off (idiot lights) to let you know to call someone more qualified to work on it.
Also , if you need 1500 calories a day to survive , and you use 1600 calories,, you need to recaculate a system.
These all in one systems are fantastic , absolutely fantastic , but your leaving yourself open for failure.
When one part of that all in one system fails (not including multiple batteries) , it all fails.
Having multiple inverters , charge controllers, battery banks , relay switches and an assortment of various other components , is the only way to protect yourself during a grid down or self sufficient scenario
I only build off grid systems with no software!
I can understand why.
I think maybe you'd need to balance at the "bottom" to avoid that ?
( all your batteries are/were, probably not carefully matched at the factory one by one, an exclusive and expensive endeavour; winston does that, on special request, i have seen once, very expensive, i was given 20 small 60 Ah batteries, Winston,yellow case, each one with a special factory engineering sticker, hand written, the person had quit a prior project, had them laying around for some 2 years and then put an ad to sell them for 400 euros,ended up giving them to me for free);
but then your system would shut off too early when charging; i do not balance at top nor bottom, but more or less at the middle, seems to work better;
i have a 48 volt system with 40 mismatched lifepo4s of all sorts and sizes, all but 4 small ones, bought used, cheap, some just given to me,like i said, had not been used in years;
i arranged them in some weird combinations for the 48 volts; used to be 24, now 48 volts;
the BMS is set at 2.65 and 3.55
i have a chinese open frame, low frequency (50 Hz), 64 euro, delivered ,inverter; this is my third, the first 2, of high frequency, they both burned up within 6 to 8 months, they did not use the famous EGS002, had sanded, erased, ICs, could not be fixed( i eventually did fix them, rewired and converted them to the SG3525 (2 or 3 euros the whole board!!), 1st stage at 23000Hz, and to the EGS002 (only 4 euros each dlivered!!) the 2nd stage, from the 380VDC to the 220VAC, 50Hz; a lot of work! i have them on the side, they are only 24 volts;
amazing to me is the Lithium chemistry, even mismatched as can be, the 40 cells all behave very similarly; voltages are king, 54.x is FULL, 49.x is empty, the BMS disconnects; at top and at bottom the cells go apart, and in a big way; it matters not, the system is sort of oversized, that is key; i have 2 strings in parallel; #1 is 8S+2P8S and #2 is 8S+2P4S;
it has worked flawlessly for a year now, this inverter CANNOT BE KILLED, 24 Mosfets, 3 proper Aluminum heat sinks, its own fan; tried ALL ways to kill it, cannot b killed; they MUST have the extra wire with the 10Kohm resistor on the EGS002!!! if not they die at any start up or any disconnect, due to the bad design with the LM393; mods are therefore needed to remedy that, the net has that,also on R7 and C21;
of course, it uses a big toroidal Si steel trafo, 19 Kgs, rated for 2000 watts, routinely up to 4100 watts for 5 to 10 minutes, i have it fan cooled with a good computer fan,12 volts,auto start, via one of those small 50 cent 40ºC thermostats, TO220 case; and a mandatory 50microHenry 200 Amp inductor on 3 stacked toroidal sendust cores that i wound with 8 turns of 3 or 4 AWG wire.
i thought i'd write this, it's an el cheapo system, well tried, completely off grid, and works amazingly well; the high frequency inverters with the small yellow trafos cannot be as reliable, they work with 380 to 400 VDC and air humidity, spiders, mosquitoes etc etc ends up killing them, some day, but the low freq one has been bullet proof; just that the trafo is VERY expensive, i paid 180 euro delivered.new, from a factry in barcelona, from Spain to Portugal but i see them on the net from 300 to 600 euros.
Thanks for sharing. Glad you could build a system from scratch.
Itll never be fixed.
Degrigation starts at manufaturing completion.
BMS is just trying to patch the holes in a sinking ship.
Thats why lithium cells have a dated life cycle
Using voltage to determine SOC in a LiFePo4 battery is basically useless.
Yep
alternate your batteries ever 2 month calibrate and parallel for24 hours
I would be ready to burn 🔥 signature wearhouse down 😂😂😂 that would drive ME INSANE...I have 4 x 14.3 chinese battery 🔋 bank NO FRIGGIN BMS ....all connected to a single bus bar ...my 64 - 280 ah cells all read 3.316 to 3.317 NO BMS ...and I got my Chinese cells 80% cheaper vs all those STUPID SERVER RACK batteries , you can now get 14.3 kwh battery for under $1,000 and I got 4 more on order ...Go tesla go sell as many powerwall 3 for $14,000 as there are suckers born ...
PS I saw a u tube video where the guy has 17 server rack batteries ALL JUST LIKE YOURS 😅😅 30 to 60% SOC