@@OffGridGarageAustralia I been building off grid power systems for over ten years, I haven't ever kept up with lithium, I know lead acid, NiFe and nicad indepth, but I have a chance to win a 30kwh battery, and if I add 6 cells a 40kwh lithium battery for my self! This has helped a ton to get up to date, and an idea of what I need to design the build. Going to be painful testing 24 individual cells! but worth it!, thanks to your videos I am selecting the JK BMS, I'll fabricate my own copper buss bars though!
Small correction: LiFePO4 is also a lithium-ion battery. They all are. It would be better to distinguish by chemistry: Lithium Iron vs Lithium Nickel/Cobalt batteries. LiFePO4 really doesn't mind being charged to 100%. Keeping them at higher voltage (3.65V per cell) is more problematic since you can actually overcharge. Similarly, keeping them at 0% (2.5V) is generally not good, but anything over 2.5V (say, 3V which is pretty much still 0% state of charge) is fine. In other words, stay away from the very extremes and you're fine.
my observation of multiple laptop battery packs and brands is that their 100% charge requires cells to be above 4.2V. This has lead to many packs failing with just one or two bad cells; suggesting to me that these cells were over stressed. I'm imagining that something similar might happen with LiFePO4 cells driven to 3.65 or higher.
@@zaneenaz4962 "In the old days" we also pushed LiFePO4 to 4V as well. I agree that even 3.65V is pushing it, especially since there isn't much energy between 3.5V and 3.6V in the first place - at least for solar. If you do fast charging at high C rates, the voltage will rise faster and you might need to go to 3.6V or something to get the most out of the cell.If you can absorb for longer periods of time, even 3.45V like what Andy does is enough.
@@upnorthandpersonal What's the real difference between absorption @ 3.45V for longer and float @ 3.375V until the sun goes down? For example, I absorb @ 3.5V for an hour and float @ 3.375V until the sun goes down (~6 hours this time of year). Would the batteries benefit from changing to a 4 hour absorption @ 3.45V and float for @ 3.375V for 3 hours? Or are we just splitting hairs at this point?
@@bbrown_sc The key thing to know is that LiFePO4 wants to settle at a certain voltage, even after longer absorb stages. The important thing is to let it do so, which is why floating at high voltages is not a good idea, and 3.375 is often chosen below this settling value (in principle, you don't need float at all). You can absorb for longer at 3.45 compared to 3.55 for example, but that's only valid if you have the time to let the battery absorb. If you want to charge faster because you don't have this luxury, you have to go to higher charge voltage. In the end, for solar applications where we're talking low C rates and in principle can afford to charge slower, it doesn't matter. There is a lot of fine tuning one can do based on weather/seasons/etc. but one can easily go too far with this and spend way too much time tweaking parameters...
The danger with keeping a near100% full LiFePO4 on a float voltage is that the (poor quality) charger can have an AC ripple with short spikes above 4.2volt even though the DC voltmeter says it is at 3.6volt constant float. Multiply values depending on your battery; 4s, 8s or 16s. Those voltage spikes can cause metalic lithium plating on the electrode, that is permanent damage. I charge my battery untill the first cell reaches 3.65v and then stop for the first cycle of the day. BMS calibrates this event as 100%, after calibration charging is resumed at 92% and stopped at 95%. At midnight the flag for "calibration done" is reset so that it will charge again to 3.65v to recalibrate the next day. The article "Practical Characteristics of Lithium Iron Phosphate Battery Cells" at Nordkyn design discusses the memory effect of not charging the cells to full (3.65v) on a regular basis. It starts building up a hump on the curve which gradually shifts to lower SOC, this means charging to only to 3.4v may reduce the usable capacity over time bacause it reaches 3.4v at a lower SOC. In a truly off grid setup the typical daily cycle would maybe be 70% to 100% (defined as 3.65v) . Where I live there are infrequent events of consecutive rainy days where SOC drops below 10%.
I found the article, was very interesting. Regarding memory effect: it seems to take a long time for this to occur, and is generally reversible. The author states in the comments: "Memory effects don’t become permanent, they just become more difficult to overcome. [...] Performing a full recharge once or twice a year would appear to be quite enough in most cases."
@@ed-ey1yb This is very interesting to know. I made a note beside my HWT - which I drain and flush twice a year, as a self reminder to ensure the same for the battery bank.
Awesome analysis and comparison of the 2 chemistries Andy. Thanks for spending the time to bring a better understanding of the characteristic differences.
I've seen li-ion based cells subdevided into 6 families. LiFePO4 is one of them, there are 3 types with mixes of Nickel, Manganese, Cobalt and Aluminium oxide s. These became popular for cars, power tools and 18650 cells. Most people think of these 3 types when mentioning li-ion. Would be better to refer to them as li-NMC , li-NCA to avoid confusion with LiFePO4.
@@camielkotte this is my exact setup. I have 2 pcs of 435w panels and 8pcs of 3.2v 280Ah Eve. I have a Victron MPPT 100/30 scc. Pls let me know how you go. I ask Andy whether the panels are enough to fill the battery for the day. I am planning to run fridge, freezer, washing machine and dryer for this setup.
@@noelbondad7423 hahaha. Funny. My pannels are 395. I have a 100-30 and a 75-15 SC. Panel lsc is 11. Thinking about a east-west setup and maybe more panels. But since we are only weekends and high summer at the cabin 2 panels should be enough. Only problem is a massive oak tree full of various kinds of birds right at the south ... So i oversized the battery. What consumers i use will proof yhe choise of wh insyalled are enough. We want a microwave and a little oven for bread. Cant decide if it must be 24volt(thick cabling) or a 230volt (massive inverter). The later Ovens use up to 3000watts during pre heat i believe where 24volts dont. Curious. What kind is your sine inverter?
@@noelbondad7423 last weekend my brother helped install both pannels and i connected one. Today i replace the SC with the second panel inparalel. Must do some woodcrafting before placing the battery and tow the dead lead away. Succes to you. Saw you channel, will look at it tonight.
@@camielkotte I have a 24v sine wave inverter of 2500 watts continuous and 5000 watts surge. At the moment my old setup is 24v system. 1 panel of 435w, Victron 100/30 Scc, 2pcs of 100Ah LIFEPO4 which I connected in series to make it to 24v. Now I want to UPGRADE my system with Eve 280Ah I mentioned above. I bought another 435w panel. These panels are Voc = 49.9volts. I am NOT SURE whether my Victron Scc can support it since its max is 100volt which is at the boundary of my 2 panels. Actually, my other issue is whether the 2pc- 435w panels can fill the battery for the day.
Thanks. I really need this and the type of info that you plan on posting. I had thought that I would not use a smart shunt just to save money, but now, I see the need for one to monitor the state of charge. It is interesting that counting electrons in the form of coulombs is really the most important measure in relation to SOC. Will be looking forward to your next few videos.
My car audio batteries get half-ish discharged and recharged several times a day. From 3.2 average to 3.6. The high amperage demand is what I think will kill them before anything else. When I enter in competitions the batteries get abused. My amp-clamp (dmm) has shown over 300 amps on one out of the four 1/0 cables feeding the amps. My home system goes from 2.85-3.48 and NEVER see over 120 amps. Itll outlast the car audio batteries BY FAAR. Excellent video Andy
Richtig gutes Zeug. Reasons for the coulomb counting being off: - It does not count balancing losses (minor) - It does not count self discharge (minor) - Most importantly: It integrates (adds up) systematic measurement errors (offset, gain, linearity) over long or even extremely long timespans. Even an offset of just 100mA can accumulate to big deviations when no recalibration is done regularly. SoC estimators for NMC, NCA and LTO usually can use the OCV curve to recalibrate to when the cell is known to be outside of the nonlinear ranges at top and bottom. For LFP it is a different story: Your best point for calibration is when the cell starts to become nonlinear at the top end. This is easily recognisable for a BMS observing the rate of voltage change. This is also the reason why e.g. Tesla recommends charging to 100% when you have a MIC LFP model. With respect to ageing: You said it at the end, high voltages themselves will also accelerate degeneration. This is due to high voltages providing activation energy for unwanted side reactions of active material and electrolyte. The c-rate sensitivity also depends on the cell construction. For high power types, high c-rates are not that critical (except for the temperature raise they naturally entail). For high energy types, this congestion and cracking issue is much more relevant. The reaction speed (or for Li chemistries, the intercalation speed) is limited and pushing beyond it increases the energy available to deteriorating processes again (mech., chem.).
Also all of the shunts I've used don't detect less than 0.5a flowing, so small current draw such as your charge controller overnight, phone charging, or LED lights may not be counted.
@@BradCagle Yes, thats true. There are chips that do some fancy chopping to be offset-free, but if a more simple variant is used, manufacturers usually just suppress the noise and offset around 0 because that is so visible to the customer when he has nothing attached but there is still something bobbing around ;).
Wow, l learned so darn much from this video. This was one of the most informative videos I have watched about charging LifePO4 batteries. And I watch a lot. THanks so much for your great videos. You have a fan and will continue watching.
Hi Andy, more great info as usual. As soon as the temperature in Melbourne rises over zero in Melbourne! I am going out to my van to reset my 200 amp LiFePO4 to use the full capacity (not just 80% to 20% ), THANK YOU! Ok, secondly I am looking forward to the upcoming videos as mentioned, but one thing I would REALLY like you to explore/discuss is "Separate Port" BMS's. There is a lot of poo pooing separate port BMS on the Net, this seems to stem from the lower charging rate available versus the output. Now I agree that this can be a factor in a stationery system with lots of panels, which is why those systems should be a totally different discussion to RV systems. Essentially you really do need a separate port BMS in an RV for obvious reasons, secondly the lower input is never likely to be an issue on an RV simply because you would be hard pushed to get enough solar on the roof to exceed the input valuation. Personally I have a Daly 200 amp output by 50 amp input (dumb) BMS. I seems to be working fine for about two years, but I would really like to upgrade to a BMS that has Bluetooth and is fully configurable, but where to find such as beast? I am aware that "hybrid" BMS's are apparently starting to appear, but I think it is still early has for them. What I really want is something like the JK BMS with the internal balancing (nice), but with a separate charging port, any suggestions? BTW, Daly has very bad communication skills, and don't make it clear you can order a BMS from them to order, that's how I got my 50 amp input model. Also people should note, some distributors of Daly products (like Deligreen) are hopeless, they supplied me with an instruction set for connections for the BMS that was totally wrong and would have blow it up on connection. P.S. They also lie through their teeth. PPS,. currently everything is controlled by my Victron BMV712 networked to the Victron MPPT.
Exactly the information I was looking for - that's kind a scary ;) Unravel all these weird hardware settings would be a real interesting project. Thank you Andy!
13:05 Excuse me, what do you mean it's impossible? You said yourself that you have a 280K battery. Charging it to 80%, to 224 Ah, which shows 3.384V. So what's the problem to charge to 3.384V and consider it 80%? Then in the discharging chart you show that 80% discharge(224 Ah again, of course) is at 3.210V. So this means that we can setup our invertor to stop using the battery at 3.21V and stop charging at 3.384V and use it that way. I think I understood what you meant for charging and discharging and it's about not cracking the anode too much with too big voltage and not going too low, but your graphs clearly show that you can reliably determine the charge percentage using the voltage alone.
I never thought I’d tell a block I’m excited but hey there’s a first time for everything ! I’m excited to see your results and findings Andy Mate. Mate 👍
I made a 100ah 12v lifepo4. I added cheap voltmeter-usb combo to the pack. Digital voltmeter reads 14.9v. I recently bought a smart sensor which reads 13.65v at rest. Been running battery for two seasons no snd.they are amazing.
Thank you so much for your contributions, the timing could not have been better for me because I’m putting my 280ah system together now so I’m looking forward to your future videos .
12v LiFePo4 im using settings in my SCC LVD at 12.6v and charging voltage at 14.2v i like that setting compared to default (Li) 11.5v LVD and 14.4v charging using a SRNE 40A SCC.. your channel is very informative thank you
I am off-grid and charging at 13.60/floating at 13.50 (PV) based on your findings. Once I am confident my test system is functioning the way I expect, I will consider a little more investment. Thank you for your contribution to the DIY community...
@@freddyswanepoel5640 of course you can (wired in series with appropriate sized conductor). Tie the positive terminal from one battery to the negative terminal on the other. Then tie the unconnected positive to a circuit breaker or disconnect switch (for safety) and from the breaker to a bus bar or fuse block to distribute power to your loads and the charger. Tie the unconnected negative to a shunt (for monitoring) or a bus bar for distribution to connected loads and the charger...
@@freddyswanepoel5640 you will want to verify, with the manufacturer, that the two batteries in question are identical as far as the type of cells, BMS, ...etc. which, if you bought them at the same time (or within a month apart), they should be fine. The reason I say this is because, what could happen is you may have a problem with balancing, where the charger will stop charging when it thinks the battery bank is fully charged (when in fact one is and one isn't). You'll want to fully charge each battery individually (then let them rest, and check voltage) prior to wiring them in series, in an attempt to avoid this potential problem...
You don't need to float LFP batteries, that was for preventing sulfation of lead acid batteries. Suggest you float 13.4v or lower if you have to use PV as a day time source of power.
@@s.mendez7160 thank you for your suggestion and yes, I am using PV to power things throughout the day, slowly working my way off-grid in the valley of the sun...
This video was very emotional for me, so many tears Andy. Do you think you'll make a video reviewing your settings in your Charge Controller and Inverter? Would love to see that.
Extremely valid information Andy. So we can extend the life span of our, kind of expensive batteries! While utilizing as much of our investment as possible!
Thanks Andy. Looking forward to the research study and testing. Have you or anyone found any real world research testing on the actual effects of over and under charging with LFP cells? I'll include references to some research papers, which suggest that it's thermal effects that largely contribute to substrate damage, and more a function of charging rate than absorption charge voltage, and that overcharge damage typically occurred at voltages above 3.6v. I could not find one study that suggested did several hundred charge cycles to 3.2v vs 3.6V that I seem to recall found no appreciable degradation until charged to over 4.2V to 4.8V. but if I do I'll add it to this post. Here's some articles that are bit of a dive but give some insights into what is going on at the chemical level: LFP overdischarge conflicting voltage limits - Thread starter Hans Kroeger Start date Jan 21, 2021 DIY Solar Power Forum - Nature Article number: 30248 (2016) 'The Degradation Behavior of LiFePO4/C Batteries during Long-Term Calendar Aging March 2021' - Energies 2021, 14, 1732 "Because the 20% capacity fade and/or 100% internal resistance increase are considered to be the EOL criterion which are widely accepted, almost all of the tests performed in the aforementioned literature are carried out within this battery life span. There are few studies on aging behavior outside this range. Thus, this paper conducted an accelerated calendar aging test on a commercial lithium iron phosphate (LiFePO4) battery over at least 27 months. By considering the storage temperature and SOC level as the stress factors, an aging test matrix was designed such that the tested batteries were subjected to the aging test under five different conditions. Based on the aging results, the impact of storage temperature and SOC level on the long-term performance degradation behavior of fifteen LiFePO4 battery cells was analyzed." 'Failure Investigation of LiFePO4Cells under Overcharge Conditions - 159(5):A678 - researchgate' "The correlation of the changes of cell skin temperature and voltage with cycle number suggest that an internal shorting of the cell was developing with overcharging, which eventually led to the failure. Taking a close look at the temperature curve of each cycle (in the 10th cycle shown in Fig. 2c), the cell-skin temperature slowly increased with time but began a sharp increase at the end of the charge, indicating that the overcharge caused detri- mental damage to the cell components. The same pattern repeated with cycle number for each cycle, and the temperature continued to increase until failure. The temperature increased with charging, then reached the highest temperature at the end of the overcharge process (33◦C at 5.28 V, peak 1), then decreased with discharging; however, after the discharge plateau, the temperature increased again to another peak at the end of the discharge (26◦C at 2.8 V peak 2). As for the tem- perature at peak 1, the temperature increased with the cycle number at a rate of 1.36◦C/cycle, reaching 45.8◦C when the battery failed. It was also noted that in the last five cycles, the temperature increased with a much higher rate, 2.54◦C/cycle than that of the first five cycles (only 0.02◦C/cycle). Compared with the temperature change at peak
In their product specifications, EVE themselves recommends the SOC window to be 10% - 90% for the LF304 and the LF280K. I always wondered how to do that with such a flat curve. Victron's charge defaults for LiFePo4 are 56.8v (3.55v) for absorption for an hour and 54v (3.375v) for float. Based on your recommendations I dropped the absorption voltage to 56v (3.5v) and left the float at 54v for my two LF304 based batteries. My batteries never drop below 50% SOC so I've never worried about the low end. I charge to 100% (56v) every day.
I found the Victron standard settings a bit too high. That may work for batteries which are many years old but not for new ones. If you can charge to 100% anyway every day, I probably would only go to 3.4V/cell as absorption and then set a time for that to like 4h or so... That is even softer and still fully charges them.
Finally gonna be some videos on how to set up the bms parameters. I've contacted overkill about doing some videos on custom bms settings. They said it was a good idea but haven't seen anything yet. I have 6 280AH strings of batteries with 6 over kill 48volt bms and haven't used my batteries at all because I'm trying to figure out how to navigate the xiaxiang app. I want to do the max amount safely and with longevity. Also what is the charge rate? What about the float? Obsorb? Like to see how to do the bms settings. I have trouble naming all 6. I got 1 named but the app refused to let me name the other 5.
I know, there are a lot of parameters and settings. And they are more and les connected to each other, so it's really time to kick of this series! I was so confused myself and could not find answers either. Every forum, every website a different answer... So I did all the testing myself.
ANDY, everything you just said makes perfect sense. LIfPO's were designed to be fully charged and fully discharged for a certain number of times before degradation sets in. But we all hope in the long run a little less heat when charging will lengthen the serviceable life of the pack. So for me, ill keep it down to 3.55 volts max charge voltage per cell, with a 4 hour float at 3.45 volts. Beer is on me!
A 4 hour float could overcharge them. These are not car batteries. having your BMS get every cell to 3.45 and balance would be better. he showed us where the curves are for these cells. 3.5 could even work but would stress the battery a bit more.
Only when you need speed charging (boat, RV, less sun hours) it is necessary to go above 3.45. Ones mileage will vary indeed. For stationary and fair sun hours there is less need for higher voltages when charging... So Andy told...
Thanks Andy - hilarious beginning as usual.😁 Very good topic- BTW I’ve received both my JK BMSs from China, one for my van (12v) and one for home garage (48v) already. I’m impressed with the speed of delivery, obviously air freight! My 400Ah Winston LiFePO4 cells for our caravan are now 6.5 years old, and have been charged usually to 3.45 volts during daily use, and then maintenance charged during storage (we don’t usually use the van in Summer) with a small solar panel (just around 50w, and only late afternoon sun) to a lower state, typically only 3.3v just to avoid any chance of going too low. However I know from experience that : 1. The actual state of charge can be quite low, as the Victron shunt (Bmv 700) gets waaay out of calibration over time, ie: exactly as you said. The ONLY way to recalibrate it is to fully charge the battery. I’ve seen it say the battery is at 100% during storage periods when it’s actually at probably closer to about 35- 40%. 2. My degradation is still absolutely minimal. I have capacity tested them last Summer (Feb 2022) by running an aircon and found the capacity at something significantly over 380Ah- I wasn’t game to go any lower as I didn’t have a BMS on the van, (but will have one soon!) and this means I would have had to watch individual cell voltages personally, just not worth the hassle. Cheers and thanks again.
Thanks for sharing, Dave. Just be careful with charging the 12V batteries with such small voltage, it can overcharge them and you're reaching a high SOC at 3.3V already and the charge controller keeps charging. It might be better to charge them to ~60% and turn of all loads and the charger. With 3% self discharging there should be now need for a trickle charge for quite a few months. The Winston cells seem to be far higher quality and also better matched. I've got your email as wee and will reply soon. Cheers.
@@OffGridGarageAustralia Thanks again Andy. I certainly hear and understand what you’re saying about the longish periods in storage with 3.3v, however should have added that I feel I’m very safe as I very regularly use light loads such as charging my garden tools from the van’s inverter, and often even other items as well such as lights and even sometimes aircon if I am doing things in the van. (I have a 2.5 kW Panasonic split system in the van.) The maintenance solar panel only receives mid to late afternoon sun- net result being that my actual state of charge usually goes down hill over time even though the Victron usually says 90- 100%. (As you know, that % SOC figure is useless when the battery hasn’t been at top of charge for a sync for a long while.) For example, the last time we pulled the van out of it’s carport into full sun for a charge before using, I needed 250Ah going in to the battery to hit the top of charge, so the real SOC was probably something like 35-40%. Weather permitting, we hope to use the van perhaps from the end of this week. I’m guessing that I’ll need to pump something like another 200Ah or so in to fill the battery, wild (educated ??) guess. Cheers
I appreciate the graphs showing the difference between Lion and LiPO4. Unfortunately, my Bluetti solar generator was not charging its LiPO4 batteries properly. They overheated at 100%, because they bulk charged to almost 99%. Its 30C here, but no excuse to not being able to function properly when I have a BIG fan blowing on it! I told customer service, but no one understood my technical question to update the firmware. I now know where their charge rate was improper. Returned my EB3A, since it could never achieve proper life expectancy!
Correct me if I’m wrong Andy but most BMS do not start passive balancing until 3.5 V so charging to 3.4 V Will mean the BMS won’t balance the cells. The way that I look at batteries is they are a consumable item that is going to wear out over time i.e. lose capacity slowly no matter what you do you may prolong the life by only months maybe a year , it really comes down to how hard you cycle the batteries from my experience if the batteries are cycled hard you get an average of about three years out of the batteries where you have now only 80% of your original capacity regardless of how you try to look after the batteries, Like you said it depends on the C rating of your discharge so the more capacity you have the longer your batteries are going to last. The most problems I’ve seen with lithium batteries in the caravan and camping style of things is generally cell in balance over time reducing the overall capacity. I really like the effort you’re going to and I realise how much time you put into these videos thank you for busting some of the myths out there.
Thanks a lot Phil, great comment. With LiFePO4 chemistry, it makes no sense to balance below 3.4V. But stopping charging at 3.45V and balancing at 3.5V makes all the sense. The balancer will only start cutting off the voltage of each cell which dares to go over 3.5V! If the balancer does not start balancing, that's great because there is nothing to balance. Yes, if you cycle them hard it will burn them out far quicker, absolutely. Because many of us have invested a lot of money in these batteries, we try to understand how to prolong battery life and keep the usable as long as possible. The videos are taking a long time to make, edit, upload, describe, thumbnail and reading the comments afterwards. Thanks for appreciating that!
Andy, I have always agreed with you regarding not charging to 100% max. Something to also consider is how the battery is being used. I have 2P4S 560AH(LF280 cells) battery on my sailboat. The battery can be charged 3 ways: [1] shore powered charger(5A to 65A), [1] solar(350W), and via alternator(120A). Since I have been mainly day or over-night sail, I rarely charge via solar(keep flexible panels safely stored) or alternator because of the large capacity compared to my needs. When I am not sailing, the vessel is connected to shore power. I have the charger set for 45A boost charge to 14V(3.5V average cell). It stops charging until the battery hits a "boost return voltage" of 13V. This take about 3 days with loads mainly being for my refrigerator, computer, and networking. I could set the return voltage lower but I want to be able to randomly show up and go sailing with still plenty of capacity. I have a JDB SP04S034 200A BMS which has both bluetooth and RS485 communications. I readout the RS485 with a raspberry pi via a simple python program which creates plots for 2hours, 48hrs, and 7days windows. I can connect to the sailboat's raspi remotely via vnc to view the monitoring plots. It is interesting to note the cell deltas increase and then decrease during discharge cycle at what I call a "mini-nee", which is the at the 3.3V to 3.2V cell voltage drop. I wish I could send you some plots. The delta raises up only about 20 mV but then drops back down is a rather symmetric fashion. My guess is some change in a phase transitions.
Why do you keep bothering yourself with shore power? while you could just fix a 2pcs of 500w panel and forget? Alternator charging should also work very well on bad days
@@innocentusangira789 Why? Because: I have a big boat and when it is in its slip, [1] shore power is free and my automatic transfer switch will automatically switch the AC loads from the inverter with no interruptions, [2] I can run my boat's 16,500 BTU air conditioner(A/C) without concern(through the battery inverter it would draw 100A), [3] I have flexible solar panels which I can easily utilize for cruising beyond day trips and are safely stored greatly reducing degradation which is greater for flexible panels, [4] reduced battery cycling and different operational demands( I want solar to keep the battery near full so that at sunset to sunrise, I have maximum power availability(and possible use my A/C for several hours allowing for over night cruising during the hot & humid Florida summer season), [5] ...
@@dreupen how much power do you consume in a day ? The storage you have can only be filled with the few panels and yet... You will keep recharging even if you are in deep sea parked
Yep, so on a boat it's a bit of a different situation and you want to charge your battery with anything you have as fast as possible. Different to our solar stationary setup. I agree, in this case it makes totally sense to charge the 3.55V or even 3.6V as fast as possible so you're ready for your next trip. You certainly don't want to leave a port with only a half charged battery.
@@innocentusangira789 The issue with boats is space, and going from a 300 watt to 500watt when the panel is mounted horizontal is not worth the investment or the time to install.
Verry clear explenation Andy, nice job as allways Looking forward to see the results, i kind off winged it when i made my packs. To make full use of the chargers capabilities i even went as far as to make 18S packs and charge them to 64V (max Victron voltage) this would be 3.55V per cell. Dangerously close to 3.65 i guess but in a perfect world... i can never overcharge them. In my installation you can clearly see that at 98%SOC the pack voltage is still only 60V (3.33V per cell) and after that it goes realy quickly, but to prevent cell runaway the batrium lowers the charging rate for the "top balalnce". there are so many cool tools out there to manage your batteries, gonna be nice to see you play with them. Batrium should send you a sample, they are "local" for you i guess... Keep it up Andy, 40K subs soon!
The only good thing about a lithium cell is they are lightweight. All forms of battery cells are garbage. Our bodies are the best energy devices ever created due to the small amount of resources needed to do a large amount of work. An electric generator is great , the only problem is we need something portable that is good enough to power it. Fuel combustion is still our best bet. I can make my own fuel and get a lot of energy out of it.
Well my new off-grid hunt camp power system goes active tomorrow. 520 watts of solar panels powering a 40 amp SCC to a 12v 200ah DIY Lifepo4 battery. I've been watching all of your videos along with Will's DIY solar forum and I'm still confused on what parameter settings I should use. You say charge to 3.4 (13.6) but what voltage should I use to start top balancing my battery? Guess I should rewatch some videos to see what I probably missed the first time. Really enjoy your videos and the time you spend finding and suggesting new or better possibilities for this battery chemistry. 🇺🇸☀️🍺
Hey, great setup and design. This should work well. For a rough start: - 3.45V Absorption, Bulk or Boost voltage in your solar charge controller (different names for the same thing) - 3.35V for Float (when the battery gets full) - 3.45V Balancing start voltage (depends a bit on your BMS) More details coming soon.
Andy, If I may ask.... What is the calendar shelf life of lithium cells? I am currently using grade B CALB cells which are 10 years old.... Still going strong at 90% capacity, 100% off-grid with daily capacity discharge from 100% -30% level. I just keep wondering does it mean there's a day I may just wake up to find the storage bank is 40% or less ? Then why should I not just consume all my capacity while still alive ? I can imagine .... I might take another 15yrs to get to 80% capacity.
Calendar life only means that batteries degrade from age if they are used or not. It does not mean they will be dead at a certain time or date, but still lose a tiny bit of capacity every single day regardless how and if you sue them. If yours are 10 years old and still have 90% capacity, that is great
@@OffGridGarageAustralia I am in my 40s, I guess they will be still alive when am gone.....I appreciate your educative videos which will help us to extend lives of these cells for quite so many years to come. I always remember the argument of Bulk Voltage being same as Float Voltage. My cells get full by mid-day, I keep using the PV power without cycling my cells for the reminder of the day. This makes me to cycle the battery bank for less than 1 cycle per day.....This is very helpful insight I see many of solar users do not know... I really appreciate your work..
@@innocentusangira789 thank you. It depends on the exact settings in your charge controller if bulk should be the same as absorption and also the devices you're using. Watch the upcoming videos about the settings to learn more. It will be interesting.
I bought a Renogy 100W solar "kit" for my boat. It came with everything you need..panel, Z brackets, connectors, wiring, and a PWM 30A charge controller all for 129. Am charging a 100A Lifepo4 battery. (yes I immediately see I need more panels.) The charging problem I perceive is that the "Wanderer" PWM charge controller,set to Li. charges to over 14.2 volts! There are no user settings. Question I have is: What about using a Buck/Boost converter to regulate the current to 13.6 or somewhere in that range?
Hey Andy - A test idea for you: Variable discharge rate curves. Many times we have seen you show us the flat voltage curve of LiFePO4 when discharging (or charging) between 90% and 10% state of charge and with the knees at each end of the SOC. Most of these show the curve when tested with a constant discharge rate. I was wondering however what the discharge voltage curve looked like if you programmed a variable rate of discharge, say cycling between 0.1C and 1C for 5-minutes at a time, or some other variable load profile. How much "wobble" would the voltage display when a battery is being discharge with variable rates? You have a battery tester which can be programmed to run such a test I believe. It seems to me that variable rate of discharge would be a more realistic test of the demands placed on many batteries. Thanks and keep up the good work.
Thanks Alex. I have tried that in some experiment already and the result was... there was no difference. The think is the Peukert factor for lithium cells is almost 1 so it does no matter with which current we charge or discharge the cells. The curves will look almost identical. I can do the test with my tester and show you in a future video, no problem. Thanks for your suggestion.
Andy, have plotted any charge/discharge curves for some of the common 100Ah 12.8V LFP's by Li-Time or Power Qween? Their chargers are CC/CV 14.6V which is not necessary except for cell balancing. Seeing a charge curve would be helpful for setting up a Solar Charge Controller specific to those batteries. Thank you for your awesome productions.
Seems to me if cells are not perfectly ballanced, Smart Shunt is not enough. It won't protect single cells to go in "unhealth" voltage teritory. Each cell should be measured and BMS should send commands to charger to lower the charging current. All three of them. Maybe use some kind of BMS system with master and slave shunts capability. Or just rely on three batteryes. If one BMS disconects its battery, the other two are still running, and the system stays on. But if you have just one battery, BMS must comunicate with charger. Only the charger should finish charging, not the BMS (by disconnecting the battery). Can't wait, what is further going to develop in this project. Top work Andy. Respect!
Another thing to consider might be the swelling that occurs while charging and discharging. I don't know how much this physical movement of the cells contributes to degradation, but the swelling on my cells tends to occur more at above 3.35V/cell (26.8/8 cells). I also consider how long my battery will play my stereo at full volume and whether or not it's long enough to make me happy. Will my hearing be degraded before I need to replace my battery? Time will tell the result of that experiment. You might want to consider/research adding closed captioning to your videos in the long term😁 Thanks for your input. I have enjoyed your videos 👍
The swelling is minimal and often cannot be measured. I'm using my cells uncompressed with some minimal space in between and had some swelling after 1.5 years of using them like this, so the cells actually touched in certain areas. So you either compress with recommended 300kgf, leave some space between the cells or use flexible busbars, so cells can move without putting force on the terminals. I like you stereo experiment. That is really outstanding😂
the bench is clean and free - looks like new project on the horizon but that was just at the beginning of the video before you unpacked that lot of stuff.
As usual a very interesting video, Andy. I am looking forward to all the settings. Still strubbling with my absorption and float specs from the Victron RS450(wich I set close together). Hope to see/hear you soon.....
Andy, my interest is in a system for my boat and since I don't sail during winter I would like to know if I have to change settings when minimal battery drain ? Come for a sail when it gets warmer. André in Sydney
I bought a server rack 48V battery, and its BMS talks to the inverter. I am still experimenting with the charge parameter settings. With this communication (battery & inverter both EG4), do I really need a shunt?
Appreciate the explanation and frankly was quite surprised. My Lifepo4 battery manufacturer wants me to charge it with 14,6V only. I guess they want to sell me new batteries soon? :)
It depends. If it is a 12V battery with built-in BMS and a very small balance current for example, this would be the only chance for them to balance this pack at a higher voltage. I would not bother and just stop at 13.8V and let it absorb there. It will still be 100% full
@@OffGridGarageAustralia Yeah. Now I have noticed that my Ultimatron 200Ah batteries work nicely with 13.6V (Epever 10420), but Wulills 200Ah batteries need 14.4V (with Victron charger) to fill them up.
The problem I have is that while the solar controllers are great for regulating the battery there seems to be no way to automatically divert any extra available current which for me is such a waste and I find that I can only go by voltage and have a diversion system that turns on by voltage but this does not tell me how much power is actually available so I can either take too much or too little power to my diverted load; any ideas you have how not to waste this potential power would be very helpful. Thanks
Thanks Jo. Yeah, that is a very difficult topic and problem. You can use some automation for that but as you said, it won't be perfect as the solar power may change and then you will use battery power. Basically you can only set a certain threshold to turn on a water heater when you hit absorption and then turn it off again if the battery goes under 95%SOC. This is something we will explore as part of the automation I have planned for my system.
This is my first experience with you, having only just found the solar garage. I’m extremely interested as I plan to be building an off grid system in the near future & things have changed drastically since I was last fully off the grid. Looking forward to more 👍
Hi. I found your video very informative. I run my Escooter 48v 20aH (960wH) Voltage is basically my 'fuel gauge'. :) Calculating in my head (and after a few test runs), I worked out that 47v (resting voltage) is my 50% mark (or turn around time!) About 50Km round trip.... I'm thinking about installing a 48v 100aH battery for a long distance endurance run (project w solar) Cheers von Australien 👍 L.E > Ideally i would require a 48v small petrol generator, say 850w which I can recharge on the fly, so to speak. I've not yet looked into availability.... However, with 100Ah i sink I would get around 250 Km out of it. (5Kw x .30 cents = $1.50 to 'fill er up' on AC charger..) 😅
I could literally write a book about what I know about flooded deep cycles. But my switch to lifepo4 has been frustrating because there are too many opinions and so called "experts." What I really like about this channel is that you don't write your information onto stone tablets. I've only been researching lifepo4 charge/discharge and haven't seen the 20 -80% rule. I can tell you, however, that my laptop only charges to 95% to prevent shortening the life of the battery.
C-rate and temperature are probably the biggest factors. In a solar application, where the C-rate is almost universally very low, cycle life should be very high. Basically it is the expansion and contraction stress which creates the problem, not so much the voltage. The chemical structure of the medium has a different 'natural' size with vs without the ions, so the matrix actually changes dimensions with each charge and discharge cycle. Sometimes very significantly. SOC matters some but I don't recall the exact reference... basically the degradation curve doesn't go exponential for LFP within the 0-100% range, but it does 'curl up' a bit and go non-linear close to 0% and close to 100%. But the biggest factor is C rate and the second biggest factor is temperature. -Matt
Danke für das sehr informative Video. Du inspirierst mich meine "Balkonanlage" umzubauen. Victron Laderegler ist schon hier, Hankzor JK ist bestellt. Wenn die Akkupreise wieder gegen Normal gehen werden auch neue (280ah) Akkus bestellt.
WENN die Preise wieder runtergehen, kaufe ich gleich so viele, das die gleich wieder steigen. Darauf warten wohl viele. Die Nachfrage ist einfach viel zu gross...
Andy - you're correct theoretically, but in a practical sense you can get pretty close to a % by selecting a voltage AND a time. I want to have about 65% in my battery each morning, so I set it to charge between 2-5am to a voltage of 53.7v. Every single day this gets me 65-67% full as per the JK BMS readout. It quickly hits the voltage & holds, then the capacity slowly creeps up. So you need to "learn" what voltage level & how long & then you can practically get a specific % pretty reliably. If it starts at 0% maybe i'll get 61%, if it starts at 40% i'll see my 65% within that timeframe of charging. Hope that helps!
That's almost impossible, Nate. The SOC the JK shows is not accurate and it will drift with every cycle if it does not get calibrated frequently. At 3.356V you still in the flat area of the curve and will have inconsistent results. It may work in your case because you have a grid charger with a constant current and hence you may see a more constant result with this approach. Most people charge their batteries through solar and at this voltage it can be anything. Even 3.4V per cell is not enough to be consistent unless you absorb for a quite a while.
Hi, I have an 18v electric saw with a push button that I can see Soc it shows less when the saw is running so I assume that is the more correct measurement.
Hi Andy. I have a question for you. I build my power wall/solar battery out of a Tesla model 3. So, what SOC do you charge your model 3 car to? My battery is a "Standard" battery with 32 cells per group and 14 groups for a 57.6 volt battery. Sooooo, should I charge to 57.6 volts or 58.0 volts? Do I consider the 100% SOC at 4.2 volts or something else? What is the100% SOC voltage per cell in your car? Lou Little
Hey, hey, hey, I wasn't aware you can harvest a model 3 battery modules, I thought they were compounded and glued together in on big block. Need to rewatch Sandy Munro's videos again. Edit: just did it. So you 32 cells in parallel and 14 of these bricks in series? That would make 14*3.7V=51.8V nominal voltage. Anyway, I only charge mine to 60% for daily usage but 4.2V is 100%SOC. I have done this twice since I have the car and watched it on the Scan my Tesla app. So in your case I would charge them to 3.8V max (53.2V pack voltage) and let them absorb. Because they are NCA cells you need to have the right BMS for them.
@@OffGridGarageAustralia I'll send you a pix of what I have done. I am using an active balancer model JK-B1A24S. It is a scailed down version of your JK that you have. Only one amp charge/discharge. Running all the time down to 0.002 volts. The groups (32 cells) don't wonder like the lithium iron phosphate cells that you have. But then again, you don't have 32 cells in parallel to keep the others in check. [:~) More to come.
Very valuable information Andy. I am glued to your presentations. I have a question. If you have 3 150A batteries (in parallel) charging at 0.2C, would you set the Charge Controller at 30A or 90A ?
Three 150 amp-HOUR batteries, not three 150 amp ones. Think of it like this (simplified); In parallel, you have three lanes. Whatever the charger outputs, it gets split up over those lanes. To send 30 amps through each battery, you thus need 90 amps. In a series connection, you only have one lane going through all the batteries which requires only 30 amps.
Hi Andy. Great Idea! Please include in to the scope also solar charge controller setings. E.g. where to set a voltage of 3,4? Is this in SCC or BMS etc. Another topic is why not to read state of charge from BMS using CANBUS vs using SmartShunt? Maybe this is clear when you are at battery version 2 or 3, but for those who are at version 1.2 there are a more questions than answers
Yeah, we will go through all the settings soon... I'm not a fan of communication between BMS and solar charger of inverter. It brings no benefits in my experience and only ads further complexity. It's not necessary.
I keep it simple and set my BMS to 3.0 and 3.5 for now. I will adjust as your videos indicate otherwise. Should I need more current overall I will add more cells as I go. It spreads out the cost and the prices seem to drop a bit each year. EZ-PZ. Thanks for the video.
but they BMS just want to be your last backup to protect the cells. you should better make the voltage settings with the charge and discharge controller.
The BMS should *never* be used to control charging a battery, neither charging nor discharging, whether with or without comms. That's a total NO NO! The BMS turns off your battery if something goes wrong or exceeds otherwise set specifications, nothing else. Always use a specific dedicated charger to charge your batteries which controls the high and low of charging. The BMS will be there if this charger fails. Don't ever charge you battery with a BMS only!
@Off-Grid Garage: You mixed up the anode and the cathode. It is the anode that contains the graphite and optionally a bit of silicon. Look up Lithium-Silicon battery on Wikipedia: In a "lithiated" (charged) state, the anode swells up.
Another great vid, I have often had similar thoughts with all ur testing video experiments, (hours and hours of them), why push the curves? Look forward to ur new coulomb counting methods and trickery. Cheers
I could use some help. Running 200watt solar on a camper for the 12v fridge. have it connected to 30amp pwm controller then to the lifepo4 battery. from there I go to a 25a breaker for what breakers do and to isolate the lifepo4 from the lead system when in use (the truck alternator/shore power). Then I put another pwm controller set for 13volts going to the lead battery. So the solar charges the lifepo4 then the lifepo4 charges the lead. The lifepo4 does most of the lifting charging and discharging. Am I off my rocker or is this viable. only on the second test right now. First time I was pumping to much to the lead so backed it down a bit.
That won't work. You can have only one controller connected to your panel. So, solar panel - PWM controller -LiFePO4 battery. If you want to charge the 12V battery as well from this system, you would need a DC-DC charger. The takes energy from the LiFePO4 and charges the 12V lead acid battery. Renogy makes them for example.
@@OffGridGarageAustralia Thanks I'll look into DC-DC charger. I was thinking the second pwm would work as a dc-dc charger. This last test looks promising but did not want to go to far and find what the breaking point is and damage the battery.
@@swamprat9018 nah, the voltage difference is not there to charge one 12V battery from another 12V battery. The DC-DC charger has a boost converter to raise the voltage to 16-18V. PWM solar chargers cannot do that.
@@OffGridGarageAustralia I only have a 200 watt solar panel rated for 17.?? volts going to the first battery. I've only seen it get up to 11v. in daylight so far. So I don't even have enough solar to charge the lithium battery with the provided PWM set for lithium settings. This is making me into the un happy camper group. LOL Thank You for sharing your knowledge and time.
@@swamprat9018 disconnect the solar panel from the battery and measure the voltage without any load. It should show the 17V. It's normal once it is connected to the battery, that the voltage will go down.
Hi Andy, I am looking forward to the new series, it is very timely since I am just starting to configure my components. I have a good understanding of how each component (chargers, inverters and BMS) work individually, but when looking at the entire system as one, it is confusing. for example, If I set the BMS to stop charging when any cell reaches 3.45v and similarly set the charge controller to stop at 13.8v (3.45 X 4) the BMS is always going to stop the charging before the charge controller, so why bother with those settings? Will you be addressing these settings? thanks!
Hello Malcolm. It may not be the best settings, so I don’t suggest anyone to do the same, but my settings, if I translate it to a 12V system, are 13.8V for the charge controler, but instead of limiting the bms at 3,45 per cell, I alow 3.58V, and limit the total to 14.4V, wich the battery will never reach of course, but one high cell could reach 3.58V, and one low cell could stay at 3,32V while the 2 remaining cells would be spot on 3,45V, and the charge controler would stop charging, but my fckng daly bms would not disconect the battery from the charger and load, it would, if set to 3.45V and 13.8v limits. My external balancer is set to start at 13.6V, so those extreem limits never occure. Single cell drift happen, and not always the same cell. 😂 I hope it helps. Cheers.
@@hommerdalor6301 Hi Hommer, thanks for your input. I am using the electrodakus BMS, which will remotely turn on/off the chargers/inverters. After giving it more thought, I'm going to try the more conservative route: have the BMS stop charging when the first cell reaches 3.45v and leave the charger setting at 13.8v. that way if for some reason the BMS does not stop charging, the charger will stop at 13.8v. thanks!
excellent as always ... smiles .. beerwah swamp is appearing again above the water line .. the grass has grown deep and the sun is once again shining upon us ... lol..
I'm sure I read somewhere about how not having about 50% charge - with equal charge on both sides had long-term degradation potentially - but not so much vs calendar life wear anyway, but I thought I'd also read somewhere that you want to keep your charge cycles short which favours higher voltage. But my memory's vague and imprecise and I couldn't comment on reasons. What I've struggled with is standby SOC. All I want are UPS systems which seem simple to make now using ideal diodes for DC standby power ... but I'm not sure it's good to hold LifePO4 cells at 100% standby all the time? If it's accepted that holding the cells for long durations at lower SOC leads to longer calendar life, and if the cells are capable of 3000 to 6000 even cycle life, are there not charge systems that will simply charge your cells to 100% every day say, or maybe once a week / whatever makes sense, and then let them power the load 'til they drop to 50% to 70% SOC ... whatever the optimum is ... and then keep the load powered by your mains supply etc. Would that not optimise your calendar life for a standby use? They could have "intelligent" modifications to their profiles for brown-out/black-out events based on their intervals etc. I've never been able to find items to purchase that performs this kind of charging/status-maintenance profile. Am I mad in some way for thinking it would be a good thing to want?
Edit ... 50% charge - I meant "unequal" charge on the opposite electrodes leading to degradation. I tried chatting to Chat GPT about it ... extract: CHAT GPT: Indeed, the nominal voltage of LFP (Lithium Iron Phosphate) cells hovers around 3.2V or 3.3V, and you're right that the full charge voltage peaks at about 3.6V before settling back down when the cell is at rest or in an open-circuit state. This difference in voltage might seem like a short hop in the cosmos, but in the world of battery chemistry, it's a significant leap. The crux of the degradation question lies not so much in the specific voltage value itself (whether it's 3.3V or 3.6V), but in the battery's state of charge (SOC) and how long it remains at or near full capacity. High SOC (Close to Full Charge) Degradation: When we talk about degradation at high SOC, we're considering the battery's condition when it's kept fully charged (or near that state) for extended periods. At this state, even if the voltage settles to around 3.3V after charging to 3.6V, the battery chemistry is in a state that is more prone to degradation. This is due to several factors: Stress on Electrode Materials: Being at a high SOC means the lithium ions are densely packed into the cathode, putting stress on the material. Potential for Electrolyte Decomposition: While less of an issue at 3.3V, the battery having been at 3.6V and full SOC means there's a potential for more reactive conditions that could have initiated decomposition processes, which might not fully reverse even if the voltage drops. Voltage and SOC Relationship: The flat discharge curve of LFP batteries means that they operate over a relatively narrow voltage range for much of their discharge cycle. At any given voltage point on this curve, the battery's SOC can significantly impact its health: High SOC (near 3.3V): The cell has been recently charged and could be experiencing the after-effects of high SOC stress even if the voltage has normalized. Low SOC (also near 3.3V): The cell is in a more relaxed state with less stress on the materials, less densely packed lithium in the cathode, and generally more favorable conditions for longevity. So, to navigate through the asteroid field of confusion: The concern with degradation at high SOC is less about the voltage being momentarily higher during charging (the journey to 3.6V) and more about the state of the battery's chemistry when it remains fully charged. It's the difference between sprinting to a finish line and then staying tense and ready to sprint again, versus relaxing after the race. The latter is akin to a battery resting at a comfortable SOC, where its chemical components are under less stress and thus more durable in the long haul of their operational life. Think of it as ensuring your starship's dilithium crystals are not under constant pressure, even if they can handle the power flow; giving them a breather extends their useful life in the vast expanse of space exploration. --- I was wondering - seeing how prices for battery's have come down so much with a 280Ah Eco Worthy 12.8V offering on Amazon in the UK for £699 - £150 voucher = £550 ish ... and looking at Victron's MPTT solar chargers not because I have solar but because of the VE Direct thingy and higher charge rates than their chargers ... where I could just use a 480W 48V supply I imagine as input to a 100V/50A MPTT charger, whether I could use those components to implement the charging profile for UPS standby use I want. I was looking at the Renology 3000W inverter (I'm unlikely to exceed 1000W and would live mostly under 400W). It says it switches over in less than 50ms which is a bit slow, but I use Seasonic ATX supplies in my computers which have a long hold-up time. And that inverter supports a remote switch ... I'm not quite sure if that can just toggle the inverting function but allow mains AC if present to be available at the output sockets but if not I'm sure I could rig something up. Yeah - I was thinking - I could let the charger do it's thing, but monitor it - maybe with an ESP32 or something and with MQTT to Home Assistant, so that when the battery's fully charged I could activate the inverter and switch to it's output until I've counted (oh - I'd probably want a smart shunt too - I forgot that - with ve direct etc. - if I've got that terminology right) roughly 20% to 30% as a compromise battery SOC used up, and then switch the invertor to mains again so that the UPS is in standby mode. I'm imagining, maybe, the ve-direct thing would allow not just profile changes, but active disabling of charging ... so I could deliberately turn-off charging in the MPTT controller until there's a brown-out / black-out event in which case I'd turn it back on again until logic based on event intervals and with buffer times tailing etc. gives the all clear (after 100% SOC is again achieved and the deliberate discharge occurs again). With home experiment stuff I was messing about with lots of smaller batteries so I could over-engineer wring and have less current and play things super-safe with lots of protection but that introduced lots of management complexity and I never had the time and space to see things through. But I wasn't aware of Victron stuff. I always thought of the things that I occasionally noticed on Amazon as being for solar panels etc. It didn't occur to me I might be able to use them for my UPS ideas without solar needed. (I'd love to have solar but I think we'd need a new roof first in our tiny Victorian terrace with very little space to accommodate green tech). But now costs have come down and with the revelation on what's available ... well ... now my debts got out of hand and it'd be risky putting more on credit cards lol ... but still ... it's tempting. Again though - am I mad for wanting the set-up I suggest? To me it seems like "common-sense" for a UPS. I want something that might last 10-years plus with having to think about it or worry about it or be bothered if there's complex brown-out/black-out events or whatever. And on a budget. I just couldn't seem to find stuff ready-made. That wasn't huge or really expensive or that did stuff I didn't need.
Oh oops - I keep conflating numerical values lol ... for a DIY 12V system I have on the wall I have an LED panel with the voltage showing ... and because that can show 13.1 to 13.3 before dropping to 12.8 etc. and sit at 13.6V for a bit during charging, I keep confusing that with "3.3" and "3.6" cell voltages when I'm tired. Anyway. I think I communicated the essence of what I was trying to say. Though I forgot to say about scheduled recharge/discharge and occasional discharge/recharge/discharge cycles on top of awaiting brown-out/black-out events in my little imagined scenario.
It seems to me that keeping a low charge voltage risks not being able to harness all available power before sundown especially on bad days. So what works on sunny days is not optimal on rainy days and those days are the most important
Andy, well done, since it is years ago, call me a Late Bloomer.,, in the LifePo4 world, I hope as I look thru your 300+ videos, we can find your Take on the MFG's warranty using the 80% reference and 6000 cycles.. as my understanding in Solar design is trying to keep your Discharge ideally down to only a remaining available of 75-80%, using only 20-25%, with a max stop down to 50% balance remaining.. from what I am taking from this video for the LifePo4, is 20-80% operating range is not best, as in summary, discharging to such lower rates and then charging back with higher voltage to get the batteries back up your creating degradation to the health of the battery over time... Please correct my take on this.. Thank You for all sharing and testing.. sent a cava from Buyflyer
Doesn't make sense to charge to only 80 percent. If you use your lifepo4 battery for survival you will need the maximum power output you can get. Also charging to only 80 percent, your voltage is much lower and your run times are shorter and less voltage throughout load. More importantly you need to only charge to 100 percent when you only need to, as long as you use battery and drain only down to 20 percent. Use your battery immediately after charged. Don't leave your battery at full charge for days. Lifepo4 average lifespan is roughly 10 years anyways, so why worry. Get as much use out of it that you need, the lifepo4 battery works for you, not you slaving over the battery state of charge. Simple and dont fear all the hype about 80 percent charge. Your lifepo4 battery only balances at 100 percent charge anyways, hello. Also make sure to use only a Lifepo4 battery charger ( or a Lifepo4 charge controller for solar) so it doesn't under charge or over charge. Absolutely there are more things to worry about in life. Like getting maximum voltage use out of your batteries.
Because of your presentation style, my brain is absorbing the information without undue fracturing!
Thanks.
This is one of the most informative videos I've ever watched in regard to SOC on Li-Ion and LiFePO4 cells. Thank you!
Thanks a lot, glad it made sense and helped.
@@OffGridGarageAustralia look up carpen battery..70 yrs,,no charge..
@@OffGridGarageAustralia I been building off grid power systems for over ten years, I haven't ever kept up with lithium, I know lead acid, NiFe and nicad indepth, but I have a chance to win a 30kwh battery, and if I add 6 cells a 40kwh lithium battery for my self! This has helped a ton to get up to date, and an idea of what I need to design the build.
Going to be painful testing 24 individual cells! but worth it!, thanks to your videos I am selecting the JK BMS, I'll fabricate my own copper buss bars though!
@@Refertech101 Thanks for your feedback. All the best with your project. It will be great and very rewarding!
@@harrywalker968 da daaaa, vor sa distruga muzeul tehnic unde se afla...
Small correction: LiFePO4 is also a lithium-ion battery. They all are. It would be better to distinguish by chemistry: Lithium Iron vs Lithium Nickel/Cobalt batteries. LiFePO4 really doesn't mind being charged to 100%. Keeping them at higher voltage (3.65V per cell) is more problematic since you can actually overcharge. Similarly, keeping them at 0% (2.5V) is generally not good, but anything over 2.5V (say, 3V which is pretty much still 0% state of charge) is fine. In other words, stay away from the very extremes and you're fine.
Thank you!
my observation of multiple laptop battery packs and brands is that their 100% charge requires cells to be above 4.2V. This has lead to many packs failing with just one or two bad cells; suggesting to me that these cells were over stressed. I'm imagining that something similar might happen with LiFePO4 cells driven to 3.65 or higher.
@@zaneenaz4962 "In the old days" we also pushed LiFePO4 to 4V as well. I agree that even 3.65V is pushing it, especially since there isn't much energy between 3.5V and 3.6V in the first place - at least for solar. If you do fast charging at high C rates, the voltage will rise faster and you might need to go to 3.6V or something to get the most out of the cell.If you can absorb for longer periods of time, even 3.45V like what Andy does is enough.
@@upnorthandpersonal What's the real difference between absorption @ 3.45V for longer and float @ 3.375V until the sun goes down? For example, I absorb @ 3.5V for an hour and float @ 3.375V until the sun goes down (~6 hours this time of year). Would the batteries benefit from changing to a 4 hour absorption @ 3.45V and float for @ 3.375V for 3 hours? Or are we just splitting hairs at this point?
@@bbrown_sc The key thing to know is that LiFePO4 wants to settle at a certain voltage, even after longer absorb stages. The important thing is to let it do so, which is why floating at high voltages is not a good idea, and 3.375 is often chosen below this settling value (in principle, you don't need float at all).
You can absorb for longer at 3.45 compared to 3.55 for example, but that's only valid if you have the time to let the battery absorb. If you want to charge faster because you don't have this luxury, you have to go to higher charge voltage.
In the end, for solar applications where we're talking low C rates and in principle can afford to charge slower, it doesn't matter. There is a lot of fine tuning one can do based on weather/seasons/etc. but one can easily go too far with this and spend way too much time tweaking parameters...
Exactly the way I understood it after playing with LiFePo4 batteries for a few years. Most knowledgeable videos about these batteries.
The danger with keeping a near100% full LiFePO4 on a float voltage is that the (poor quality) charger can have an AC ripple with short spikes above 4.2volt even though the DC voltmeter says it is at 3.6volt constant float. Multiply values depending on your battery; 4s, 8s or 16s.
Those voltage spikes can cause metalic lithium plating on the electrode, that is permanent damage.
I charge my battery untill the first cell reaches 3.65v and then stop for the first cycle of the day. BMS calibrates this event as 100%, after calibration charging is resumed at 92% and stopped at 95%.
At midnight the flag for "calibration done" is reset so that it will charge again to 3.65v to recalibrate the next day.
The article "Practical Characteristics of Lithium Iron Phosphate Battery Cells" at Nordkyn design discusses the memory effect of not charging the cells to full (3.65v) on a regular basis. It starts building up a hump on the curve which gradually shifts to lower SOC, this means charging to only to 3.4v may reduce the usable capacity over time bacause it reaches 3.4v at a lower SOC.
In a truly off grid setup the typical daily cycle would maybe be 70% to 100% (defined as 3.65v) . Where I live there are infrequent events of consecutive rainy days where SOC drops below 10%.
I found the article, was very interesting. Regarding memory effect: it seems to take a long time for this to occur, and is generally reversible. The author states in the comments: "Memory effects don’t become permanent, they just become more difficult to overcome. [...] Performing a full recharge once or twice a year would appear to be quite enough in most cases."
@@ed-ey1yb This is very interesting to know. I made a note beside my HWT - which I drain and flush twice a year, as a self reminder to ensure the same for the battery bank.
WOW cheap charlie chargers with ripple and spikes - may explain the quite common sort of fire from these e scooters left on over night charge
Awesome analysis and comparison of the 2 chemistries Andy. Thanks for spending the time to bring a better understanding of the characteristic differences.
I've seen li-ion based cells subdevided into 6 families. LiFePO4 is one of them, there are 3 types with mixes of Nickel, Manganese, Cobalt and Aluminium oxide s. These became popular for cars, power tools and 18650 cells. Most people think of these 3 types when mentioning li-ion. Would be better to refer to them as li-NMC , li-NCA to avoid confusion with LiFePO4.
Perfect timing Andy, I'm finally setting up my own battery 2.0 and the Smart Shunt is part of that. So I'm keen for a deep dive into the settings.
Me too just in time. Perfect!
Next weekend i will install my first ☺️ 8x 280 with jk BMS and victron charger(800 watts panels)
@@camielkotte this is my exact setup. I have 2 pcs of 435w panels and 8pcs of 3.2v 280Ah Eve. I have a Victron MPPT 100/30 scc. Pls let me know how you go. I ask Andy whether the panels are enough to fill the battery for the day. I am planning to run fridge, freezer, washing machine and dryer for this setup.
@@noelbondad7423 hahaha. Funny. My pannels are 395. I have a 100-30 and a 75-15 SC. Panel lsc is 11. Thinking about a east-west setup and maybe more panels. But since we are only weekends and high summer at the cabin 2 panels should be enough. Only problem is a massive oak tree full of various kinds of birds right at the south ... So i oversized the battery.
What consumers i use will proof yhe choise of wh insyalled are enough. We want a microwave and a little oven for bread. Cant decide if it must be 24volt(thick cabling) or a 230volt (massive inverter).
The later Ovens use up to 3000watts during pre heat i believe where 24volts dont. Curious.
What kind is your sine inverter?
@@noelbondad7423 last weekend my brother helped install both pannels and i connected one. Today i replace the SC with the second panel inparalel. Must do some woodcrafting before placing the battery and tow the dead lead away.
Succes to you. Saw you channel, will look at it tonight.
@@camielkotte I have a 24v sine wave inverter of 2500 watts continuous and 5000 watts surge. At the moment my old setup is 24v system. 1 panel of 435w, Victron 100/30 Scc, 2pcs of 100Ah LIFEPO4 which I connected in series to make it to 24v. Now I want to UPGRADE my system with Eve 280Ah I mentioned above. I bought another 435w panel. These panels are Voc = 49.9volts. I am NOT SURE whether my Victron Scc can support it since its max is 100volt which is at the boundary of my 2 panels. Actually, my other issue is whether the 2pc- 435w panels can fill the battery for the day.
Thanks. I really need this and the type of info that you plan on posting. I had thought that I would not use a smart shunt just to save money, but now, I see the need for one to monitor the state of charge. It is interesting that counting electrons in the form of coulombs is really the most important measure in relation to SOC. Will be looking forward to your next few videos.
Thanks Bill. You can us the 'smart shunt' of the BMS as well but, but, but... it's not as reliable as the real smart shunt.
My car audio batteries get half-ish discharged and recharged several times a day. From 3.2 average to 3.6. The high amperage demand is what I think will kill them before anything else. When I enter in competitions the batteries get abused. My amp-clamp (dmm) has shown over 300 amps on one out of the four 1/0 cables feeding the amps. My home system goes from 2.85-3.48 and NEVER see over 120 amps. Itll outlast the car audio batteries BY FAAR. Excellent video Andy
Thanks for sharing. I always wondered why you guys don't use supercaps for this sort of usage. Or at least LTO cells...?
Richtig gutes Zeug.
Reasons for the coulomb counting being off:
- It does not count balancing losses (minor)
- It does not count self discharge (minor)
- Most importantly: It integrates (adds up) systematic measurement errors (offset, gain, linearity) over long or even extremely long timespans. Even an offset of just 100mA can accumulate to big deviations when no recalibration is done regularly.
SoC estimators for NMC, NCA and LTO usually can use the OCV curve to recalibrate to when the cell is known to be outside of the nonlinear ranges at top and bottom. For LFP it is a different story: Your best point for calibration is when the cell starts to become nonlinear at the top end. This is easily recognisable for a BMS observing the rate of voltage change. This is also the reason why e.g. Tesla recommends charging to 100% when you have a MIC LFP model.
With respect to ageing: You said it at the end, high voltages themselves will also accelerate degeneration. This is due to high voltages providing activation energy for unwanted side reactions of active material and electrolyte.
The c-rate sensitivity also depends on the cell construction. For high power types, high c-rates are not that critical (except for the temperature raise they naturally entail). For high energy types, this congestion and cracking issue is much more relevant. The reaction speed (or for Li chemistries, the intercalation speed) is limited and pushing beyond it increases the energy available to deteriorating processes again (mech., chem.).
Also all of the shunts I've used don't detect less than 0.5a flowing, so small current draw such as your charge controller overnight, phone charging, or LED lights may not be counted.
@@BradCagle Yes, thats true. There are chips that do some fancy chopping to be offset-free, but if a more simple variant is used, manufacturers usually just suppress the noise and offset around 0 because that is so visible to the customer when he has nothing attached but there is still something bobbing around ;).
Wow, l learned so darn much from this video. This was one of the most informative videos I have watched about charging LifePO4 batteries. And I watch a lot. THanks so much for your great videos. You have a fan and will continue watching.
Hi Andy, more great info as usual. As soon as the temperature in Melbourne rises over zero in Melbourne! I am going out to my van to reset my 200 amp LiFePO4 to use the full capacity (not just 80% to 20% ), THANK YOU!
Ok, secondly I am looking forward to the upcoming videos as mentioned, but one thing I would REALLY like you to explore/discuss is "Separate Port" BMS's.
There is a lot of poo pooing separate port BMS on the Net, this seems to stem from the lower charging rate available versus the output.
Now I agree that this can be a factor in a stationery system with lots of panels, which is why those systems should be a totally different discussion to RV systems.
Essentially you really do need a separate port BMS in an RV for obvious reasons, secondly the lower input is never likely to be an issue on an RV simply because you would be hard pushed to get enough solar on the roof to exceed the input valuation.
Personally I have a Daly 200 amp output by 50 amp input (dumb) BMS. I seems to be working fine for about two years, but I would really like to upgrade to a BMS that has Bluetooth and is fully configurable, but where to find such as beast?
I am aware that "hybrid" BMS's are apparently starting to appear, but I think it is still early has for them.
What I really want is something like the JK BMS with the internal balancing (nice), but with a separate charging port, any suggestions?
BTW, Daly has very bad communication skills, and don't make it clear you can order a BMS from them to order, that's how I got my 50 amp input model.
Also people should note, some distributors of Daly products (like Deligreen) are hopeless, they supplied me with an instruction set for connections for the BMS that was totally wrong and would have blow it up on connection.
P.S. They also lie through their teeth.
PPS,. currently everything is controlled by my Victron BMV712 networked to the Victron MPPT.
Exactly the information I was looking for - that's kind a scary ;) Unravel all these weird hardware settings would be a real interesting project. Thank you Andy!
Great video, and looking forward to your settings series.
Coming soon!
13:05 Excuse me, what do you mean it's impossible? You said yourself that you have a 280K battery. Charging it to 80%, to 224 Ah, which shows 3.384V. So what's the problem to charge to 3.384V and consider it 80%? Then in the discharging chart you show that 80% discharge(224 Ah again, of course) is at 3.210V. So this means that we can setup our invertor to stop using the battery at 3.21V and stop charging at 3.384V and use it that way.
I think I understood what you meant for charging and discharging and it's about not cracking the anode too much with too big voltage and not going too low, but your graphs clearly show that you can reliably determine the charge percentage using the voltage alone.
The flat charging/discharging curve of the LFP is addictive. Voltage sag was a mayor pain in the arsch with AGM, especially in freezing temps.
Voltagesag with AGMs suck but charging LFPs in freezing temps suck even more...
@@windsolarupnorth7084 Yep, built LFP batterybox with heating.
I never thought I’d tell a block I’m excited but hey there’s a first time for everything ! I’m excited to see your results and findings Andy Mate. Mate 👍
Great to got you excited, Paul Hardie 😃
There will be more for you soon here. And thank you for being here!
Looking forward to dive in to those settings. Your expertise make a lot of sense. Learning so much as always. Cheers my friend
I made a 100ah 12v lifepo4. I added cheap voltmeter-usb combo to the pack. Digital voltmeter reads 14.9v. I recently bought a smart sensor which reads 13.65v at rest. Been running battery for two seasons no snd.they are amazing.
Thank you so much for your contributions, the timing could not have been better for me because I’m putting my 280ah system together now so I’m looking forward to your future videos .
12v LiFePo4 im using settings in my SCC LVD at 12.6v and charging voltage at 14.2v i like that setting compared to default (Li) 11.5v LVD and 14.4v charging using a SRNE 40A SCC.. your channel is very informative thank you
I am off-grid and charging at 13.60/floating at 13.50 (PV) based on your findings. Once I am confident my test system is functioning the way I expect, I will consider a little more investment. Thank you for your contribution to the DIY community...
Can i connect two of the same lifePO4 24v to get 48v ?
@@freddyswanepoel5640 of course you can (wired in series with appropriate sized conductor). Tie the positive terminal from one battery to the negative terminal on the other. Then tie the unconnected positive to a circuit breaker or disconnect switch (for safety) and from the breaker to a bus bar or fuse block to distribute power to your loads and the charger. Tie the unconnected negative to a shunt (for monitoring) or a bus bar for distribution to connected loads and the charger...
@@freddyswanepoel5640 you will want to verify, with the manufacturer, that the two batteries in question are identical as far as the type of cells, BMS, ...etc. which, if you bought them at the same time (or within a month apart), they should be fine. The reason I say this is because, what could happen is you may have a problem with balancing, where the charger will stop charging when it thinks the battery bank is fully charged (when in fact one is and one isn't). You'll want to fully charge each battery individually (then let them rest, and check voltage) prior to wiring them in series, in an attempt to avoid this potential problem...
You don't need to float LFP batteries, that was for preventing sulfation of lead acid batteries. Suggest you float 13.4v or lower if you have to use PV as a day time source of power.
@@s.mendez7160 thank you for your suggestion and yes, I am using PV to power things throughout the day, slowly working my way off-grid in the valley of the sun...
This video was very emotional for me, so many tears Andy. Do you think you'll make a video reviewing your settings in your Charge Controller and Inverter? Would love to see that.
The video is coming soon, stay tuned... wipe of your tears first!
Extremely valid information Andy. So we can extend the life span of our, kind of expensive batteries! While utilizing as much of our investment as possible!
Thanks!
Thanks a lot for your support, Ray🙏🍺
Thanks Andy. Looking forward to the research study and testing. Have you or anyone found any real world research testing on the actual effects of over and under charging with LFP cells? I'll include references to some research papers, which suggest that it's thermal effects that largely contribute to substrate damage, and more a function of charging rate than absorption charge voltage, and that overcharge damage typically occurred at voltages above 3.6v. I could not find one study that suggested did several hundred charge cycles to 3.2v vs 3.6V that I seem to recall found no appreciable degradation until charged to over 4.2V to 4.8V. but if I do I'll add it to this post.
Here's some articles that are bit of a dive but give some insights into what is going on at the chemical level:
LFP overdischarge conflicting voltage limits - Thread starter Hans Kroeger Start date Jan 21, 2021
DIY Solar Power Forum - Nature Article number: 30248 (2016)
'The Degradation Behavior of LiFePO4/C Batteries during Long-Term Calendar Aging March 2021' -
Energies 2021, 14, 1732
"Because the 20% capacity fade and/or 100% internal resistance increase are considered
to be the EOL criterion which are widely accepted, almost all of the tests performed in
the aforementioned literature are carried out within this battery life span. There are few
studies on aging behavior outside this range. Thus, this paper conducted an accelerated
calendar aging test on a commercial lithium iron phosphate (LiFePO4) battery over at
least 27 months. By considering the storage temperature and SOC level as the stress
factors, an aging test matrix was designed such that the tested batteries were subjected to
the aging test under five different conditions. Based on the aging results, the impact of
storage temperature and SOC level on the long-term performance degradation behavior of
fifteen LiFePO4 battery cells was analyzed."
'Failure Investigation of LiFePO4Cells under Overcharge Conditions - 159(5):A678 - researchgate'
"The correlation of the changes of
cell skin temperature and voltage with cycle number suggest that an
internal shorting of the cell was developing with overcharging, which
eventually led to the failure. Taking a close look at the temperature
curve of each cycle (in the 10th cycle shown in Fig. 2c), the cell-skin
temperature slowly increased with time but began a sharp increase
at the end of the charge, indicating that the overcharge caused detri-
mental damage to the cell components. The same pattern repeated
with cycle number for each cycle, and the temperature continued to
increase until failure. The temperature increased with charging, then
reached the highest temperature at the end of the overcharge process
(33◦C at 5.28 V, peak 1), then decreased with discharging; however,
after the discharge plateau, the temperature increased again to another
peak at the end of the discharge (26◦C at 2.8 V peak 2). As for the tem-
perature at peak 1, the temperature increased with the cycle number at
a rate of 1.36◦C/cycle, reaching 45.8◦C when the battery failed. It was
also noted that in the last five cycles, the temperature increased with a
much higher rate, 2.54◦C/cycle than that of the first five cycles (only
0.02◦C/cycle). Compared with the temperature change at peak
Jack, did your text get cut off. Seems to end adruptly.
You are an excellent scientist, and I very much appreciate your explanation of charge voltage at the atomic level!
In their product specifications, EVE themselves recommends the SOC window to be 10% - 90% for the LF304 and the LF280K. I always wondered how to do that with such a flat curve. Victron's charge defaults for LiFePo4 are 56.8v (3.55v) for absorption for an hour and 54v (3.375v) for float. Based on your recommendations I dropped the absorption voltage to 56v (3.5v) and left the float at 54v for my two LF304 based batteries. My batteries never drop below 50% SOC so I've never worried about the low end. I charge to 100% (56v) every day.
I found the Victron standard settings a bit too high. That may work for batteries which are many years old but not for new ones. If you can charge to 100% anyway every day, I probably would only go to 3.4V/cell as absorption and then set a time for that to like 4h or so... That is even softer and still fully charges them.
@@OffGridGarageAustralia if you have the capacity, go for de less degradation posible thanks Andy and Up North and Personal exelent information
Excellent and really knowledgeable video. Better than most others that really do not know much about lithium batteries.
This will be interesting. Looking forward to all the different opinions here and the facts as you present this.
Same, just started reading...
Very good video. Great detail on this subject. I have a 512wh Lifepo4 powerbank. I have set the upper charge limit to 90%.
Finally gonna be some videos on how to set up the bms parameters. I've contacted overkill about doing some videos on custom bms settings. They said it was a good idea but haven't seen anything yet.
I have 6 280AH strings of batteries with 6 over kill 48volt bms and haven't used my batteries at all because I'm trying to figure out how to navigate the xiaxiang app. I want to do the max amount safely and with longevity. Also what is the charge rate? What about the float? Obsorb? Like to see how to do the bms settings. I have trouble naming all 6. I got 1 named but the app refused to let me name the other 5.
I know, there are a lot of parameters and settings. And they are more and les connected to each other, so it's really time to kick of this series! I was so confused myself and could not find answers either. Every forum, every website a different answer... So I did all the testing myself.
Thank you Andy, just in time my first 16 280Ah cells arrived this week. 🙂
Hey great!
ANDY, everything you just said makes perfect sense. LIfPO's were designed to be fully charged and fully discharged for a certain number of times before degradation sets in. But we all hope in the long run a little less heat when charging will lengthen the serviceable life of the pack. So for me, ill keep it down to 3.55 volts max charge voltage per cell, with a 4 hour float at 3.45 volts. Beer is on me!
A 4 hour float could overcharge them. These are not car batteries. having your BMS get every cell to 3.45 and balance would be better. he showed us where the curves are for these cells. 3.5 could even work but would stress the battery a bit more.
Only when you need speed charging (boat, RV, less sun hours) it is necessary to go above 3.45. Ones mileage will vary indeed.
For stationary and fair sun hours there is less need for higher voltages when charging...
So Andy told...
Thanks a lot for the SPAT 🍺, Bruce. Lower your float a bit and I'm happy with your settings. Thanks again for your support!
2nd time watched it . skipped all easy informations those years you gave and i still find new ones.... . thanks.
Thanks Andy - hilarious beginning as usual.😁
Very good topic- BTW I’ve received both my JK BMSs from China, one for my van (12v) and one for home garage (48v) already.
I’m impressed with the speed of delivery, obviously air freight!
My 400Ah Winston LiFePO4 cells for our caravan are now 6.5 years old, and have been charged usually to 3.45 volts during daily use, and then maintenance charged during storage (we don’t usually use the van in Summer) with a small solar panel (just around 50w, and only late afternoon sun) to a lower state, typically only 3.3v just to avoid any chance of going too low. However I know from experience that :
1. The actual state of charge can be quite low, as the Victron shunt (Bmv 700) gets waaay out of calibration over time, ie: exactly as you said. The ONLY way to recalibrate it is to fully charge the battery. I’ve seen it say the battery is at 100% during storage periods when it’s actually at probably closer to about 35- 40%.
2. My degradation is still absolutely minimal. I have capacity tested them last Summer (Feb 2022) by running an aircon and found the capacity at something significantly over 380Ah- I wasn’t game to go any lower as I didn’t have a BMS on the van, (but will have one soon!) and this means I would have had to watch individual cell voltages personally, just not worth the hassle.
Cheers and thanks again.
Thanks for sharing, Dave.
Just be careful with charging the 12V batteries with such small voltage, it can overcharge them and you're reaching a high SOC at 3.3V already and the charge controller keeps charging. It might be better to charge them to ~60% and turn of all loads and the charger. With 3% self discharging there should be now need for a trickle charge for quite a few months.
The Winston cells seem to be far higher quality and also better matched.
I've got your email as wee and will reply soon. Cheers.
@@OffGridGarageAustralia Thanks again Andy.
I certainly hear and understand what you’re saying about the longish periods in storage with 3.3v, however should have added that I feel I’m very safe as I very regularly use light loads such as charging my garden tools from the van’s inverter, and often even other items as well such as lights and even sometimes aircon if I am doing things in the van. (I have a 2.5 kW Panasonic split system in the van.)
The maintenance solar panel only receives mid to late afternoon sun- net result being that my actual state of charge usually goes down hill over time even though the Victron usually says 90- 100%. (As you know, that % SOC figure is useless when the battery hasn’t been at top of charge for a sync for a long while.)
For example, the last time we pulled the van out of it’s carport into full sun for a charge before using, I needed 250Ah going in to the battery to hit the top of charge, so the real SOC was probably something like 35-40%.
Weather permitting, we hope to use the van perhaps from the end of this week. I’m guessing that I’ll need to pump something like another 200Ah or so in to fill the battery, wild (educated ??) guess. Cheers
I appreciate the graphs showing the difference between Lion and LiPO4.
Unfortunately, my Bluetti solar generator was not charging its LiPO4 batteries properly. They overheated at 100%, because they bulk charged to almost 99%. Its 30C here, but no excuse to not being able to function properly when I have a BIG fan blowing on it!
I told customer service, but no one understood my technical question to update the firmware. I now know where their charge rate was improper. Returned my EB3A, since it could never achieve proper life expectancy!
Bulk charging to 99% is not a bad thing with LiFePO4 and usually works well. Once the voltage rises to over 3.45V, the cells are 99%+ charged anyway.
Very interesting. I am looking for the next videos about this topic
Correct me if I’m wrong Andy but most BMS do not start passive balancing until 3.5 V so charging to 3.4 V Will mean the BMS won’t balance the cells.
The way that I look at batteries is they are a consumable item that is going to wear out over time i.e. lose capacity slowly no matter what you do you may prolong the life by only months maybe a year , it really comes down to how hard you cycle the batteries from my experience if the batteries are cycled hard you get an average of about three years out of the batteries where you have now only 80% of your original capacity regardless of how you try to look after the batteries, Like you said it depends on the C rating of your discharge so the more capacity you have the longer your batteries are going to last.
The most problems I’ve seen with lithium batteries in the caravan and camping style of things is generally cell in balance over time reducing the overall capacity. I really like the effort you’re going to and I realise how much time you put into these videos thank you for busting some of the myths out there.
Thanks a lot Phil, great comment.
With LiFePO4 chemistry, it makes no sense to balance below 3.4V. But stopping charging at 3.45V and balancing at 3.5V makes all the sense. The balancer will only start cutting off the voltage of each cell which dares to go over 3.5V! If the balancer does not start balancing, that's great because there is nothing to balance.
Yes, if you cycle them hard it will burn them out far quicker, absolutely. Because many of us have invested a lot of money in these batteries, we try to understand how to prolong battery life and keep the usable as long as possible.
The videos are taking a long time to make, edit, upload, describe, thumbnail and reading the comments afterwards. Thanks for appreciating that!
Andy, I have always agreed with you regarding not charging to 100% max. Something to also consider is how the battery is being used. I have 2P4S 560AH(LF280 cells) battery on my sailboat. The battery can be charged 3 ways: [1] shore powered charger(5A to 65A), [1] solar(350W), and via alternator(120A). Since I have been mainly day or over-night sail, I rarely charge via solar(keep flexible panels safely stored) or alternator because of the large capacity compared to my needs. When I am not sailing, the vessel is connected to shore power. I have the charger set for 45A boost charge to 14V(3.5V average cell). It stops charging until the battery hits a "boost return voltage" of 13V. This take about 3 days with loads mainly being for my refrigerator, computer, and networking. I could set the return voltage lower but I want to be able to randomly show up and go sailing with still plenty of capacity. I have a JDB SP04S034 200A BMS which has both bluetooth and RS485 communications. I readout the RS485 with a raspberry pi via a simple python program which creates plots for 2hours, 48hrs, and 7days windows. I can connect to the sailboat's raspi remotely via vnc to view the monitoring plots. It is interesting to note the cell deltas increase and then decrease during discharge cycle at what I call a "mini-nee", which is the at the 3.3V to 3.2V cell voltage drop. I wish I could send you some plots. The delta raises up only about 20 mV but then drops back down is a rather symmetric fashion. My guess is some change in a phase transitions.
Why do you keep bothering yourself with shore power? while you could just fix a 2pcs of 500w panel and forget?
Alternator charging should also work very well on bad days
@@innocentusangira789 Why? Because: I have a big boat and when it is in its slip, [1] shore power is free and my automatic transfer switch will automatically switch the AC loads from the inverter with no interruptions, [2] I can run my boat's 16,500 BTU air conditioner(A/C) without concern(through the battery inverter it would draw 100A), [3] I have flexible solar panels which I can easily utilize for cruising beyond day trips and are safely stored greatly reducing degradation which is greater for flexible panels, [4] reduced battery cycling and different operational demands( I want solar to keep the battery near full so that at sunset to sunrise, I have maximum power availability(and possible use my A/C for several hours allowing for over night cruising during the hot & humid Florida summer season), [5] ...
@@dreupen how much power do you consume in a day ?
The storage you have can only be filled with the few panels and yet... You will keep recharging even if you are in deep sea parked
Yep, so on a boat it's a bit of a different situation and you want to charge your battery with anything you have as fast as possible. Different to our solar stationary setup. I agree, in this case it makes totally sense to charge the 3.55V or even 3.6V as fast as possible so you're ready for your next trip. You certainly don't want to leave a port with only a half charged battery.
@@innocentusangira789 The issue with boats is space, and going from a 300 watt to 500watt when the panel is mounted horizontal is not worth the investment or the time to install.
Actually Andy, you are so spot on that I’ve got nothing else to say but…..thanks!
And... thank you, Karl!
Andy, please explain calendar aging? Thumbs up and still subscribed!
Verry clear explenation Andy, nice job as allways
Looking forward to see the results, i kind off winged it when i made my packs.
To make full use of the chargers capabilities i even went as far as to make 18S packs and charge them to 64V (max Victron voltage) this would be 3.55V per cell.
Dangerously close to 3.65 i guess but in a perfect world... i can never overcharge them.
In my installation you can clearly see that at 98%SOC the pack voltage is still only 60V (3.33V per cell) and after that it goes realy quickly, but to prevent cell runaway the batrium lowers the charging rate for the "top balalnce".
there are so many cool tools out there to manage your batteries, gonna be nice to see you play with them.
Batrium should send you a sample, they are "local" for you i guess...
Keep it up Andy, 40K subs soon!
To add even more uncertainty to the SOC, is charge rate. When charging to 80%, the voltage will vary depending on how many Amps you are charging at.
The only good thing about a lithium cell is they are lightweight. All forms of battery cells are garbage. Our bodies are the best energy devices ever created due to the small amount of resources needed to do a large amount of work. An electric generator is great , the only problem is we need something portable that is good enough to power it. Fuel combustion is still our best bet. I can make my own fuel and get a lot of energy out of it.
I saw your eyes sparkle when you looked at those beer plaques! 👍 🍻
It's fun to learn together.
We love you to Andy!
Be safe all
Thanks 👍
Well my new off-grid hunt camp power system goes active tomorrow. 520 watts of solar panels powering a 40 amp SCC to a 12v 200ah DIY Lifepo4 battery. I've been watching all of your videos along with Will's DIY solar forum and I'm still confused on what parameter settings I should use. You say charge to 3.4 (13.6) but what voltage should I use to start top balancing my battery? Guess I should rewatch some videos to see what I probably missed the first time. Really enjoy your videos and the time you spend finding and suggesting new or better possibilities for this battery chemistry. 🇺🇸☀️🍺
Hey, great setup and design. This should work well. For a rough start:
- 3.45V Absorption, Bulk or Boost voltage in your solar charge controller (different names for the same thing)
- 3.35V for Float (when the battery gets full)
- 3.45V Balancing start voltage (depends a bit on your BMS)
More details coming soon.
Andy,
If I may ask.... What is the calendar shelf life of lithium cells?
I am currently using grade B CALB cells which are 10 years old.... Still going strong at 90% capacity, 100% off-grid with daily capacity discharge from 100% -30% level. I just keep wondering does it mean there's a day I may just wake up to find the storage bank is 40% or less ?
Then why should I not just consume all my capacity while still alive ? I can imagine .... I might take another 15yrs to get to 80% capacity.
I don’t think anyone yet knows the calendar life of these if looked after properly. Suffice to say: it’s good.
Calendar life only means that batteries degrade from age if they are used or not. It does not mean they will be dead at a certain time or date, but still lose a tiny bit of capacity every single day regardless how and if you sue them. If yours are 10 years old and still have 90% capacity, that is great
@@OffGridGarageAustralia
I am in my 40s, I guess they will be still alive when am gone.....I appreciate your educative videos which will help us to extend lives of these cells for quite so many years to come.
I always remember the argument of Bulk Voltage being same as Float Voltage. My cells get full by mid-day, I keep using the PV power without cycling my cells for the reminder of the day. This makes me to cycle the battery bank for less than 1 cycle per day.....This is very helpful insight I see many of solar users do not know...
I really appreciate your work..
@@innocentusangira789 thank you. It depends on the exact settings in your charge controller if bulk should be the same as absorption and also the devices you're using. Watch the upcoming videos about the settings to learn more. It will be interesting.
You are a absolute star . From south africa
Finally I learned what happens when batteries charge and discharge. What anode snd cathod are. Such a great content. Love from Pakistan.
Looking forward to this series.
Great topic, exactly what I need, can't wait to watch next episodes
I bought a Renogy 100W solar "kit" for my boat. It came with everything you need..panel, Z brackets, connectors, wiring, and a PWM 30A charge controller all for 129. Am charging a 100A Lifepo4 battery. (yes I immediately see I need more panels.) The charging problem I perceive is that the "Wanderer" PWM charge controller,set to Li. charges to over 14.2 volts! There are no user settings. Question I have is: What about using a Buck/Boost converter to regulate the current to 13.6 or somewhere in that range?
Merci pour tout ces judicieux conseils ANDY. Bonne continuation.👍
Hey Andy - A test idea for you: Variable discharge rate curves.
Many times we have seen you show us the flat voltage curve of LiFePO4 when discharging (or charging) between 90% and 10% state of charge and with the knees at each end of the SOC.
Most of these show the curve when tested with a constant discharge rate.
I was wondering however what the discharge voltage curve looked like if you programmed a variable rate of discharge, say cycling between 0.1C and 1C for 5-minutes at a time, or some other variable load profile. How much "wobble" would the voltage display when a battery is being discharge with variable rates? You have a battery tester which can be programmed to run such a test I believe.
It seems to me that variable rate of discharge would be a more realistic test of the demands placed on many batteries.
Thanks and keep up the good work.
Thanks Alex. I have tried that in some experiment already and the result was... there was no difference. The think is the Peukert factor for lithium cells is almost 1 so it does no matter with which current we charge or discharge the cells. The curves will look almost identical.
I can do the test with my tester and show you in a future video, no problem.
Thanks for your suggestion.
Andy, have plotted any charge/discharge curves for some of the common 100Ah 12.8V LFP's by Li-Time or Power Qween? Their chargers are CC/CV 14.6V which is not necessary except for cell balancing. Seeing a charge curve would be helpful for setting up a Solar Charge Controller specific to those batteries. Thank you for your awesome productions.
Seems to me if cells are not perfectly ballanced, Smart Shunt is not enough. It won't protect single cells to go in "unhealth" voltage teritory.
Each cell should be measured and BMS should send commands to charger to lower the charging current. All three of them. Maybe use some kind of BMS system with master and slave shunts capability. Or just rely on three batteryes. If one BMS disconects its battery, the other two are still running, and the system stays on. But if you have just one battery, BMS must comunicate with charger. Only the charger should finish charging, not the BMS (by disconnecting the battery). Can't wait, what is further going to develop in this project. Top work Andy. Respect!
Another thing to consider might be the swelling that occurs while charging and discharging. I don't know how much this physical movement of the cells contributes to degradation, but the swelling on my cells tends to occur more at above 3.35V/cell (26.8/8 cells).
I also consider how long my battery will play my stereo at full volume and whether or not it's long enough to make me happy. Will my hearing be degraded before I need to replace my battery? Time will tell the result of that experiment. You might want to consider/research adding closed captioning to your videos in the long term😁
Thanks for your input. I have enjoyed your videos 👍
The swelling is minimal and often cannot be measured. I'm using my cells uncompressed with some minimal space in between and had some swelling after 1.5 years of using them like this, so the cells actually touched in certain areas.
So you either compress with recommended 300kgf, leave some space between the cells or use flexible busbars, so cells can move without putting force on the terminals.
I like you stereo experiment. That is really outstanding😂
the bench is clean and free - looks like new project on the horizon
but that was just at the beginning of the video before you unpacked that lot of stuff.
Great unboxing video!😀 Looking forward to the upcoming series on settings.
Coming soon!
As usual a very interesting video, Andy. I am looking forward to all the settings. Still strubbling with my absorption and float specs from the Victron RS450(wich I set close together).
Hope to see/hear you soon.....
Very soon, Victor, it is pretty much in production...
Hello can I ask you if the positive connection it’s in contact with battery case ?
Correct, the positive of the battery is the outside case.
@@OffGridGarageAustralia thank you
Andy, my interest is in a system for my boat and since I don't sail during winter I would like to know if I have to change settings when minimal battery drain ? Come for a sail when it gets warmer. André in Sydney
Another great episode. Thanks so much!!
I bought a server rack 48V battery, and its BMS talks to the inverter. I am still experimenting with the charge parameter settings. With this communication (battery & inverter both EG4), do I really need a shunt?
Appreciate the explanation and frankly was quite surprised. My Lifepo4 battery manufacturer wants me to charge it with 14,6V only. I guess they want to sell me new batteries soon? :)
It depends. If it is a 12V battery with built-in BMS and a very small balance current for example, this would be the only chance for them to balance this pack at a higher voltage. I would not bother and just stop at 13.8V and let it absorb there. It will still be 100% full
@@OffGridGarageAustralia Yeah. Now I have noticed that my Ultimatron 200Ah batteries work nicely with 13.6V (Epever 10420), but Wulills 200Ah batteries need 14.4V (with Victron charger) to fill them up.
The problem I have is that while the solar controllers are great for regulating the battery there seems to be no way to automatically divert any extra available current which for me is such a waste and I find that I can only go by voltage and have a diversion system that turns on by voltage but this does not tell me how much power is actually available so I can either take too much or too little power to my diverted load; any ideas you have how not to waste this potential power would be very helpful. Thanks
Thanks Jo. Yeah, that is a very difficult topic and problem. You can use some automation for that but as you said, it won't be perfect as the solar power may change and then you will use battery power. Basically you can only set a certain threshold to turn on a water heater when you hit absorption and then turn it off again if the battery goes under 95%SOC.
This is something we will explore as part of the automation I have planned for my system.
This is my first experience with you, having only just found the solar garage. I’m extremely interested as I plan to be building an off grid system in the near future & things have changed drastically since I was last fully off the grid. Looking forward to more 👍
Thank you and welcome to the channel. It's all here...
@@OffGridGarageAustralia Wow, you truely are a nice man. Thankyou. Can’t wait until we build my new system……Victron all the way 👍
Your informations are really good, as I just got my new LiFePo4 Bat -Pack! It helps me a lot! I will make 16S4P 280Ah Eve-cells
Ur the best for all of us newbyes in lithium battery world. Fantasik videos
Thank you very much!
Hi Andy. Where did you buy your EVE cells? I would like to buy some for a solar project here in NZ.
There are links in the description of each video I ever made and also on my website off-grid-garage.com/batteries/
Hi. I found your video very informative.
I run my Escooter 48v 20aH (960wH)
Voltage is basically my 'fuel gauge'. :)
Calculating in my head (and after a few test runs), I worked out that 47v (resting voltage) is my 50% mark (or turn around time!)
About 50Km round trip....
I'm thinking about installing a 48v 100aH battery for a long distance endurance run (project w solar)
Cheers von Australien 👍
L.E > Ideally i would require a 48v small petrol generator, say 850w which I can recharge on the fly, so to speak. I've not yet looked into availability....
However, with 100Ah i sink I would get around 250 Km out of it.
(5Kw x .30 cents = $1.50 to 'fill er up' on AC charger..) 😅
I could literally write a book about what I know about flooded deep cycles. But my switch to lifepo4 has been frustrating because there are too many opinions and so called "experts." What I really like about this channel is that you don't write your information onto stone tablets. I've only been researching lifepo4 charge/discharge and haven't seen the 20 -80% rule. I can tell you, however, that my laptop only charges to 95% to prevent shortening the life of the battery.
Ditto bro
Great stuff Andy, I look forward to the following experiments
C-rate and temperature are probably the biggest factors. In a solar application, where the C-rate is almost universally very low, cycle life should be very high. Basically it is the expansion and contraction stress which creates the problem, not so much the voltage. The chemical structure of the medium has a different 'natural' size with vs without the ions, so the matrix actually changes dimensions with each charge and discharge cycle. Sometimes very significantly. SOC matters some but I don't recall the exact reference... basically the degradation curve doesn't go exponential for LFP within the 0-100% range, but it does 'curl up' a bit and go non-linear close to 0% and close to 100%.
But the biggest factor is C rate and the second biggest factor is temperature.
-Matt
Danke für das sehr informative Video. Du inspirierst mich meine "Balkonanlage" umzubauen. Victron Laderegler ist schon hier, Hankzor JK ist bestellt.
Wenn die Akkupreise wieder gegen Normal gehen werden auch neue (280ah) Akkus bestellt.
WENN die Preise wieder runtergehen, kaufe ich gleich so viele, das die gleich wieder steigen. Darauf warten wohl viele. Die Nachfrage ist einfach viel zu gross...
another great video, very informative video with all of the key points I have been needing so far
Great, thank you!
Andy - you're correct theoretically, but in a practical sense you can get pretty close to a % by selecting a voltage AND a time. I want to have about 65% in my battery each morning, so I set it to charge between 2-5am to a voltage of 53.7v. Every single day this gets me 65-67% full as per the JK BMS readout. It quickly hits the voltage & holds, then the capacity slowly creeps up. So you need to "learn" what voltage level & how long & then you can practically get a specific % pretty reliably. If it starts at 0% maybe i'll get 61%, if it starts at 40% i'll see my 65% within that timeframe of charging.
Hope that helps!
That's almost impossible, Nate. The SOC the JK shows is not accurate and it will drift with every cycle if it does not get calibrated frequently. At 3.356V you still in the flat area of the curve and will have inconsistent results. It may work in your case because you have a grid charger with a constant current and hence you may see a more constant result with this approach. Most people charge their batteries through solar and at this voltage it can be anything. Even 3.4V per cell is not enough to be consistent unless you absorb for a quite a while.
Andy, I would love to get your views on BMS settings for charge/discharge hysteresis
Hi, I have an 18v electric saw with a push button that I can see Soc it shows less when the saw is running so I assume that is the more correct measurement.
Nissan leaf, with leaf spy pro, you get ALL the information! and it has a setting you can select for it to stop at 80%
Hi Andy. I have a question for you. I build my power wall/solar battery out of a Tesla model 3. So, what SOC do you charge your model 3 car to? My battery is a "Standard" battery with 32 cells per group and 14 groups for a 57.6 volt battery. Sooooo, should I charge to 57.6 volts or 58.0 volts? Do I consider the 100% SOC at 4.2 volts or something else? What is the100% SOC voltage per cell in your car?
Lou Little
Hey, hey, hey, I wasn't aware you can harvest a model 3 battery modules, I thought they were compounded and glued together in on big block. Need to rewatch Sandy Munro's videos again.
Edit: just did it. So you 32 cells in parallel and 14 of these bricks in series? That would make 14*3.7V=51.8V nominal voltage.
Anyway, I only charge mine to 60% for daily usage but 4.2V is 100%SOC. I have done this twice since I have the car and watched it on the Scan my Tesla app.
So in your case I would charge them to 3.8V max (53.2V pack voltage) and let them absorb. Because they are NCA cells you need to have the right BMS for them.
@@OffGridGarageAustralia I'll send you a pix of what I have done. I am using an active balancer model JK-B1A24S. It is a scailed down version of your JK that you have. Only one amp charge/discharge. Running all the time down to 0.002 volts. The groups (32 cells) don't wonder like the lithium iron phosphate cells that you have. But then again, you don't have 32 cells in parallel to keep the others in check. [:~) More to come.
Very valuable information Andy. I am glued to your presentations. I have a question. If you have 3 150A batteries (in parallel) charging at 0.2C, would you set the Charge Controller at 30A or 90A ?
Three 150 amp-HOUR batteries, not three 150 amp ones.
Think of it like this (simplified);
In parallel, you have three lanes. Whatever the charger outputs, it gets split up over those lanes. To send 30 amps through each battery, you thus need 90 amps.
In a series connection, you only have one lane going through all the batteries which requires only 30 amps.
Hi Andy. Great Idea! Please include in to the scope also solar charge controller setings. E.g. where to set a voltage of 3,4? Is this in SCC or BMS etc. Another topic is why not to read state of charge from BMS using CANBUS vs using SmartShunt? Maybe this is clear when you are at battery version 2 or 3, but for those who are at version 1.2 there are a more questions than answers
Yeah, we will go through all the settings soon...
I'm not a fan of communication between BMS and solar charger of inverter. It brings no benefits in my experience and only ads further complexity. It's not necessary.
Andy, thanks for this fantastic video. Incredibly well done.
I keep it simple and set my BMS to 3.0 and 3.5 for now. I will adjust as your videos indicate otherwise. Should I need more current overall I will add more cells as I go. It spreads out the cost and the prices seem to drop a bit each year. EZ-PZ. Thanks for the video.
but they BMS just want to be your last backup to protect the cells. you should better make the voltage settings with the charge and discharge controller.
@@alexanderp.8075 of you have comms set up between your inverter charger and the BMS then this is where you make those settings
@@edc1569 on the inverter or the bms?
When you have the comm setup.
The BMS should *never* be used to control charging a battery, neither charging nor discharging, whether with or without comms. That's a total NO NO!
The BMS turns off your battery if something goes wrong or exceeds otherwise set specifications, nothing else. Always use a specific dedicated charger to charge your batteries which controls the high and low of charging. The BMS will be there if this charger fails. Don't ever charge you battery with a BMS only!
No @Ed C, you neve set any charging parameters in the BMS. It is not a charger!
@Off-Grid Garage: You mixed up the anode and the cathode. It is the anode that contains the graphite and optionally a bit of silicon. Look up Lithium-Silicon battery on Wikipedia: In a "lithiated" (charged) state, the anode swells up.
I thought I said carbon material. I did not say the word graphite or graphene in my video😉
@@OffGridGarageAustralia carbon, graphite: same atom. Point is that you swapped them.
Another great vid, I have often had similar thoughts with all ur testing video experiments, (hours and hours of them), why push the curves? Look forward to ur new coulomb counting methods and trickery. Cheers
Yes, it is far easier to stay withing the knees and use the rising/falling voltage as a trigger to stop charging/discharging.
I could use some help. Running 200watt solar on a camper for the 12v fridge. have it connected to 30amp pwm controller then to the lifepo4 battery. from there I go to a 25a breaker for what breakers do and to isolate the lifepo4 from the lead system when in use (the truck alternator/shore power). Then I put another pwm controller set for 13volts going to the lead battery. So the solar charges the lifepo4 then the lifepo4 charges the lead. The lifepo4 does most of the lifting charging and discharging. Am I off my rocker or is this viable. only on the second test right now. First time I was pumping to much to the lead so backed it down a bit.
That won't work. You can have only one controller connected to your panel.
So, solar panel - PWM controller -LiFePO4 battery.
If you want to charge the 12V battery as well from this system, you would need a DC-DC charger. The takes energy from the LiFePO4 and charges the 12V lead acid battery. Renogy makes them for example.
@@OffGridGarageAustralia Thanks I'll look into DC-DC charger. I was thinking the second pwm would work as a dc-dc charger. This last test looks promising but did not want to go to far and find what the breaking point is and damage the battery.
@@swamprat9018 nah, the voltage difference is not there to charge one 12V battery from another 12V battery. The DC-DC charger has a boost converter to raise the voltage to 16-18V. PWM solar chargers cannot do that.
@@OffGridGarageAustralia I only have a 200 watt solar panel rated for 17.?? volts going to the first battery. I've only seen it get up to 11v. in daylight so far. So I don't even have enough solar to charge the lithium battery with the provided PWM set for lithium settings. This is making me into the un happy camper group. LOL Thank You for sharing your knowledge and time.
@@swamprat9018 disconnect the solar panel from the battery and measure the voltage without any load. It should show the 17V. It's normal once it is connected to the battery, that the voltage will go down.
Hi Andy, I am looking forward to the new series, it is very timely since I am just starting to configure my components. I have a good understanding of how each component (chargers, inverters and BMS) work individually, but when looking at the entire system as one, it is confusing. for example, If I set the BMS to stop charging when any cell reaches 3.45v and similarly set the charge controller to stop at 13.8v (3.45 X 4) the BMS is always going to stop the charging before the charge controller, so why bother with those settings? Will you be addressing these settings? thanks!
Hello Malcolm. It may not be the best settings, so I don’t suggest anyone to do the same, but my settings, if I translate it to a 12V system, are 13.8V for the charge controler, but instead of limiting the bms at 3,45 per cell, I alow 3.58V, and limit the total to 14.4V, wich the battery will never reach of course, but one high cell could reach 3.58V, and one low cell could stay at 3,32V while the 2 remaining cells would be spot on 3,45V, and the charge controler would stop charging, but my fckng daly bms would not disconect the battery from the charger and load, it would, if set to 3.45V and 13.8v limits. My external balancer is set to start at 13.6V, so those extreem limits never occure. Single cell drift happen, and not always the same cell. 😂
I hope it helps.
Cheers.
@@hommerdalor6301 Hi Hommer, thanks for your input. I am using the electrodakus BMS, which will remotely turn on/off the chargers/inverters. After giving it more thought, I'm going to try the more conservative route: have the BMS stop charging when the first cell reaches 3.45v and leave the charger setting at 13.8v. that way if for some reason the BMS does not stop charging, the charger will stop at 13.8v.
thanks!
The electrodacus is great.
Sure, try the conservative settings, experimenting is the best we can do. :-)
Fantastic explanation Andy.
excellent as always ... smiles .. beerwah swamp is appearing again above the water line .. the grass has grown deep and the sun is once again shining upon us ... lol..
I'm sure I read somewhere about how not having about 50% charge - with equal charge on both sides had long-term degradation potentially - but not so much vs calendar life wear anyway, but I thought I'd also read somewhere that you want to keep your charge cycles short which favours higher voltage. But my memory's vague and imprecise and I couldn't comment on reasons.
What I've struggled with is standby SOC. All I want are UPS systems which seem simple to make now using ideal diodes for DC standby power ... but I'm not sure it's good to hold LifePO4 cells at 100% standby all the time?
If it's accepted that holding the cells for long durations at lower SOC leads to longer calendar life, and if the cells are capable of 3000 to 6000 even cycle life, are there not charge systems that will simply charge your cells to 100% every day say, or maybe once a week / whatever makes sense, and then let them power the load 'til they drop to 50% to 70% SOC ... whatever the optimum is ... and then keep the load powered by your mains supply etc. Would that not optimise your calendar life for a standby use? They could have "intelligent" modifications to their profiles for brown-out/black-out events based on their intervals etc.
I've never been able to find items to purchase that performs this kind of charging/status-maintenance profile.
Am I mad in some way for thinking it would be a good thing to want?
Edit ... 50% charge - I meant "unequal" charge on the opposite electrodes leading to degradation.
I tried chatting to Chat GPT about it ... extract:
CHAT GPT:
Indeed, the nominal voltage of LFP (Lithium Iron Phosphate) cells hovers around 3.2V or 3.3V, and you're right that the full charge voltage peaks at about 3.6V before settling back down when the cell is at rest or in an open-circuit state. This difference in voltage might seem like a short hop in the cosmos, but in the world of battery chemistry, it's a significant leap.
The crux of the degradation question lies not so much in the specific voltage value itself (whether it's 3.3V or 3.6V), but in the battery's state of charge (SOC) and how long it remains at or near full capacity.
High SOC (Close to Full Charge) Degradation: When we talk about degradation at high SOC, we're considering the battery's condition when it's kept fully charged (or near that state) for extended periods. At this state, even if the voltage settles to around 3.3V after charging to 3.6V, the battery chemistry is in a state that is more prone to degradation. This is due to several factors:
Stress on Electrode Materials: Being at a high SOC means the lithium ions are densely packed into the cathode, putting stress on the material.
Potential for Electrolyte Decomposition: While less of an issue at 3.3V, the battery having been at 3.6V and full SOC means there's a potential for more reactive conditions that could have initiated decomposition processes, which might not fully reverse even if the voltage drops.
Voltage and SOC Relationship: The flat discharge curve of LFP batteries means that they operate over a relatively narrow voltage range for much of their discharge cycle. At any given voltage point on this curve, the battery's SOC can significantly impact its health:
High SOC (near 3.3V): The cell has been recently charged and could be experiencing the after-effects of high SOC stress even if the voltage has normalized.
Low SOC (also near 3.3V): The cell is in a more relaxed state with less stress on the materials, less densely packed lithium in the cathode, and generally more favorable conditions for longevity.
So, to navigate through the asteroid field of confusion: The concern with degradation at high SOC is less about the voltage being momentarily higher during charging (the journey to 3.6V) and more about the state of the battery's chemistry when it remains fully charged. It's the difference between sprinting to a finish line and then staying tense and ready to sprint again, versus relaxing after the race. The latter is akin to a battery resting at a comfortable SOC, where its chemical components are under less stress and thus more durable in the long haul of their operational life.
Think of it as ensuring your starship's dilithium crystals are not under constant pressure, even if they can handle the power flow; giving them a breather extends their useful life in the vast expanse of space exploration.
---
I was wondering - seeing how prices for battery's have come down so much with a 280Ah Eco Worthy 12.8V offering on Amazon in the UK for £699 - £150 voucher = £550 ish ... and looking at Victron's MPTT solar chargers not because I have solar but because of the VE Direct thingy and higher charge rates than their chargers ... where I could just use a 480W 48V supply I imagine as input to a 100V/50A MPTT charger, whether I could use those components to implement the charging profile for UPS standby use I want. I was looking at the Renology 3000W inverter (I'm unlikely to exceed 1000W and would live mostly under 400W). It says it switches over in less than 50ms which is a bit slow, but I use Seasonic ATX supplies in my computers which have a long hold-up time. And that inverter supports a remote switch ... I'm not quite sure if that can just toggle the inverting function but allow mains AC if present to be available at the output sockets but if not I'm sure I could rig something up.
Yeah - I was thinking - I could let the charger do it's thing, but monitor it - maybe with an ESP32 or something and with MQTT to Home Assistant, so that when the battery's fully charged I could activate the inverter and switch to it's output until I've counted (oh - I'd probably want a smart shunt too - I forgot that - with ve direct etc. - if I've got that terminology right) roughly 20% to 30% as a compromise battery SOC used up, and then switch the invertor to mains again so that the UPS is in standby mode. I'm imagining, maybe, the ve-direct thing would allow not just profile changes, but active disabling of charging ... so I could deliberately turn-off charging in the MPTT controller until there's a brown-out / black-out event in which case I'd turn it back on again until logic based on event intervals and with buffer times tailing etc. gives the all clear (after 100% SOC is again achieved and the deliberate discharge occurs again).
With home experiment stuff I was messing about with lots of smaller batteries so I could over-engineer wring and have less current and play things super-safe with lots of protection but that introduced lots of management complexity and I never had the time and space to see things through. But I wasn't aware of Victron stuff. I always thought of the things that I occasionally noticed on Amazon as being for solar panels etc. It didn't occur to me I might be able to use them for my UPS ideas without solar needed. (I'd love to have solar but I think we'd need a new roof first in our tiny Victorian terrace with very little space to accommodate green tech). But now costs have come down and with the revelation on what's available ... well ... now my debts got out of hand and it'd be risky putting more on credit cards lol ... but still ... it's tempting.
Again though - am I mad for wanting the set-up I suggest? To me it seems like "common-sense" for a UPS. I want something that might last 10-years plus with having to think about it or worry about it or be bothered if there's complex brown-out/black-out events or whatever. And on a budget. I just couldn't seem to find stuff ready-made. That wasn't huge or really expensive or that did stuff I didn't need.
Oh oops - I keep conflating numerical values lol ... for a DIY 12V system I have on the wall I have an LED panel with the voltage showing ... and because that can show 13.1 to 13.3 before dropping to 12.8 etc. and sit at 13.6V for a bit during charging, I keep confusing that with "3.3" and "3.6" cell voltages when I'm tired. Anyway. I think I communicated the essence of what I was trying to say. Though I forgot to say about scheduled recharge/discharge and occasional discharge/recharge/discharge cycles on top of awaiting brown-out/black-out events in my little imagined scenario.
My brain didn't draw too many Amps to understand this explanation.
It seems to me that keeping a low charge voltage risks not being able to harness all available power before sundown especially on bad days. So what works on sunny days is not optimal on rainy days and those days are the most important
Andy, well done, since it is years ago, call me a Late Bloomer.,, in the LifePo4 world, I hope as I look thru your 300+ videos, we can find your Take on the MFG's warranty using the 80% reference and 6000 cycles.. as my understanding in Solar design is trying to keep your Discharge ideally down to only a remaining available of 75-80%, using only 20-25%, with a max stop down to 50% balance remaining.. from what I am taking from this video for the LifePo4, is 20-80% operating range is not best, as in summary, discharging to such lower rates and then charging back with higher voltage to get the batteries back up your creating degradation to the health of the battery over time... Please correct my take on this.. Thank You for all sharing and testing.. sent a cava from Buyflyer
Thanks as a lot of good information is always learned from your videos. Keep them coming!
Thanks, will do!
Heya. you made it very clear the why and remember lifipo4 don't have a "memory"
Thank you for the info and knowledge sharing! Keep it up!
Doesn't make sense to charge to only 80 percent. If you use your lifepo4 battery for survival you will need the maximum power output you can get. Also charging to only 80 percent, your voltage is much lower and your run times are shorter and less voltage throughout load.
More importantly you need to only charge to 100 percent when you only need to, as long as you use battery and drain only down to 20 percent. Use your battery immediately after charged. Don't leave your battery at full charge for days. Lifepo4 average lifespan is roughly 10 years anyways, so why worry.
Get as much use out of it that you need, the lifepo4 battery works for you, not you slaving over the battery state of charge.
Simple and dont fear all the hype about 80 percent charge. Your lifepo4 battery only balances at 100 percent charge anyways, hello. Also make sure to use only a Lifepo4 battery charger ( or a Lifepo4 charge controller for solar) so it doesn't under charge or over charge.
Absolutely there are more things to worry about in life. Like getting maximum voltage use out of your batteries.