We are in a very interesting time with residential power storage. It's like early pioneering times. These companies build all these great hybrid inverters and BMSs etc without really knowing what they are doing, or really understanding all the challenges. Remind of the the PC days back in the late 80s and early 90s, just when things started to heat up. We have exciting times ahead with all this solar related tech.
I bet they won't. None of them is making the software themselves. They all reliant on 3rd party companies which may have no knowledge about battery charging and BMS design.
@@OffGridGarageAustralia my seplos 10e bms reset soc 100% with high voltage alarm setting (not sure cell or pack, but they reach 100% soc without triggering the cell or pack high voltage protection ) which is separate and lower than the high voltage protection setting. maybe you have the old revision/firmware.?
Fantastic video Andy! Thank you. You are doing a great service to the community of people that need help with learning everything solar! This BMS seems very impressive for sure! One thing that I worry about with BMS' is ONE setting... Delta cell voltage shutdown should be implemented. If you have a runaway cell it can take down the whole pack before shutdown happens at a certain pack voltage. I think I have said that I use a "naked" BMS monitoring system. Instead using Home assistant, I can set a shutdown if there is a DCV of a certain voltage. Pretty cool eh? Anyway, this is just the way I do it. Thanks again for your content. Together as a community we learn! 🐸🐸🐸
Thanks, Dan. Using HA for that is pretty cool. Do you still have any hardware protection as well measuring the cell voltages and turn off MOSFETs in case of a runner? Only using HA might be a bit risky? I have seen DVC settings in other BMS but is this not a double up because if one cells runs away, the BMS will catch it anyway?
@@OffGridGarageAustralia Yes my inverters actually have a setting where you can specify a voltage that will disconnect the loads if the setting is reached. Other than that I'm running naked lol. But I don't really see a need for a physical BMS the way mine is operating..
Hey Andy, already left you couple off comments on some other video about this, I never use total voltage protection on any of my BMSes when single cell voltage protection is available, just disable the switch in the Seplos BMS set the overvoltage to 55V or 54.4V or 54.8, now when the overvoltage is triggered, it doesn't disconnect charging & battery is reset to 100%, if cells are somewhat balanced this means its actually 100%, the only issue is there is no separate floating voltage option with time limit. & since a battery with well balanced cells will drop current gradually allowing it to absorb for a longer time before reaching the actual "overvoltage protection value" I compensate for the floating absence by lowering the "charge over voltage protection recovery" to something like 53.6V like this since my system is well sized panels/batteries/load my batteries will reach almost 99% and float for about 30-60 min before reaching the "overvoltage protection" voltage, thus resetting to 100% and stopping the charge current.
Yes, I replied to you comments already. I will do some more Seplos testing soon. That is a great suggestion and 'trick'. Thank you for sharing this. The missing floating option is not great but also not a big problem. Usually people turn on loads when the battery is full anyway, so it won't stay there for too long. Using the recovery voltage... so when the battery is full, you just use the solar power for your loads and the battery voltage will slowly go down due to the fact that the charge MOSFETs are turned off. Ist that how you use it?
@Off-Grid Garage - thats true the Gobal BMS is one off the Best BMS on the Market right now when you what to parallel more Battery Packs. On an 3 Phase Victron System the Solar Input to Grid is possible over the Solar Charge Controller when the Battery is 100% full - so the BMS is not blocking the Feed in to Grid. The 100% Reset and turn OFF MOS is not the best solution because on an 3Phase Victron System with a Fronius MICO GRID Inverter when the Gird is Cut off and you have full Solar Power the Victron 1:0 Rule makes for a few miliseconds a higher Voltage on the DC side until the Inverter are throttled town by the rising Hz from the Multiplus. When the MOS are closed the can not drop the incoming Power to the Batteries. Also the 100% Turn OFF MOS is the natural way of Float Voltage - when the MOS Stays ON we need a extra Permameter for the Float Voltage to set for the the VE.Bus. I'm also FAN of your theory to keep the MOS ON at 100% Maybe Gobel Power can implement a Set Float Voltage for the Victron VE Bus when Batterie is at 100% and a NO Close Function from the Charge MOS at 100% SoC!
Hello Andy My 2 Daly 150 A BMS are ok also they do the same job . With 2 * 16 LFP 280Ah from gobelpower. Connection UART USB to the rp3 . Bmv 700 and muliplus .working fine.... Also to charge my model S off grid.
Hi Andy, you say the Gobel Bms is the best, because of the software, because the SOC reset can be determined. On the other hand, you mention but, the Gobel BMS switches off the MPPT and then uses only the battery, unlike Seplos, for example. For me, however, remains in total the Seplos the better BMS, because not only more user-friendly through Bluetooth, but what is much more important, the Battery is full in the summer at noon and it follows another 8 hours SUN, which I can immediately consume or feed and sell, the Gobel BMS can not, the whole afternoon is then lived on the battery, although the sun is shining and I could use the extra yield. For me, the Seplos remains the price performance winner. Thanks for your work and thumbs up from the crazy Germany to Australia
Thanks for sharing. Yes, that is correct. I've just summarised all pros and cons as far as I could test them. It is very difficult the compare BMS as the conditions often change when charging and it is never the same. I'll get back to the Seplos soon... more too come.
I agree with everything you say here. It's ridiculous that Daly & Jk use the same silly HVC to activate 100% SOC. I've had to fine tune my Daly BMS to do a cell HVC of 3.47v to trigger 100% SOC, but then the charge MOSFET shuts off. Please Daly & JK - LISTEN to us!! 😡
I still feel the pain i brought one of the first models of the daly smart bms at a cost of $600 aud. Its now a paper weight lol. I now have 3x jk bms. If jk could do communication it would be the best peter bms. I wont be buying a new one until there's a bms with communication and active balancer and can monitor multiple packs in series and in parrel. How good would it be to have one bms for your 3 16s modules. Ps thanks andy your the battery king of knowledge
No bluetooth is a big no go for me! My solar shed is outside. The BMS I use from Battery Hookup has bluetooth, so I can keep an eye on the cells from my office in the house.
At some point I didn't care about the BMS' SOC anymore. Instead I build myself a Smart-Shunt that tracks the SOC and reports it to the GX device. I check the BMS from time to time for the cell voltages, but that's it.
I went with the seplos a few weeks ago, tested the CAN connection to Victron and diceded to forget that and go with an extra smartshunt. The only thing i miss are the values for the highest and lowest cell but the SOC is calculated perfectly by the shunt. And this setup will work with any batteries in parallel - like your battery shelf. I am not 100% clear about the usefulness of a data connection to the battery.
You can still get all the data from the Seplos BMS but use the Smart Shunt as your battery monitor. That works great. Just leave the Seplos connected via CAN but turn off the control part in the Victron System, so it does not control charging etc...
@@OffGridGarageAustralia thanks Andy. You mean not to choose the bms as the battery monitor, right? If I remember correctly, that does not work, because the system priorized the bms in my ESS configuration. But it could be, that the turned on DVCC was responsible for that. But I need/want to turn it on, because otherwise my RS450/100 is not under "external control" what it should be (the 150/45 IS under ext. contr.) plus I cannot control the overall charging current.
@@OffGridGarageAustralia today I set up my BMS connection again and all the sh*** beginns again :-D I figuerd out, that I have to set the option "Controlling BMS: No BMS control" to ignore the values coming from the BMS (CVL, CCL, DCL). But you must know that you only find this after rebooting the cerbo. And after that, you have to reboot again but not only the cerbo, but also the MP and reset all MPPTs to ignore the BMS. What a mess. And now the best: I cannot find the min- and max-cell-voltages anymore :-D
@@mario-q6k9w Hmmm, strange, I never had to restart the Cerbo or MP when changing the battery monitor or other related settings. Yeah, the controlling BMS function should be off, so you can still use the DVCC with custom settings. As soon as you do that, the chargers won't be under external control any more. Either the MBS in in control or not.
@@OffGridGarageAustralia yes, that works now. If you have BMS controlled MPPT and either set BMS control off or disconnect the CAN BMS, then the MPPT have to do a rescan, otherwise they lack of information and report an error (bms missing or something). And maybe the MP needs a reboot because of the ESS assistant. To see the the min/max cell voltages reported by the BMS in remote console I can't find the "details" under the BMS anymore and in VRM is no data available. Did you know why that is? I use the 10E BMS
Hello Andy, I have 2 batteries ( 5KWh and 10KWh) the 5KW is lifepo4 while the 10 KW is lithium ion. They are connected in parallel using 2 jk bms's and they work 24/7 . I have not noticed any problem with them. BTW.... half of my house is totally off grid.
Yeah, you can parallel them if they have a matching voltage. However you will never get the best out of your batteries, it's always a compromise. The charge and discharge curves are very different and while the LFP battery is charging to 95%, the Li-ion battery will not charge at all until the voltage rises. Then the LFP will stop and all charging power goes into the ion battery. That can lead to high amps for either battery at various states of a cycle. But interesting project, I may look into this...
@@orkidorkid 55V on top is OK for LFP and Li-ion but 42V at the bottom is kinda low for LFP (it is OK for Li-ion thou). I would not go lower than 48V. I have 20kWh LFP battery and I am planing to connect to it Li-ion battery in parallel (I got it dirt cheap). When LFP battery works between 55 and 48V and it's capacity is almost fully used, the Li-ion battery will only be utilized in circa. 60%. For me it is OK because Li-ion battery is than very safe.
Hallo Andy, I dont know how it works, but my JKBMS always reach 100% SoC when the charge controller of the Deye is in the absorbtion phase. The Values for BMS and charge controller are set like you show on 5:52
@@OffGridGarageAustralia ...but still seems to be pretty good if the budget matters :-) If you are interested, i can send you curves of a charge cycle including the voltages of every single cell.
18:44 hey Andy, so, if it’s not one of the “high-end” manufacturers, i think it’s some new China bms. It must be the new EVE BMS, because as a Cell manufacturer they should now what functionality the Software must have. And I assume they probably want Andy to promote it. I’m excited to see if I’m right.
@@JOHNDANIEL1 Don't agree with that. BMS is a safety feature and comparable with an airbag in a car. If everything is working fine, indeed you don't need the BMS. However if a cell, for whatever reason goes out of line, the BMS kicks in and avoids much larger problems.
Hi Andy. Great video. I've ran quite a few BMSs systems. JK, ANT, seplos and a Batrium You must review a batrium BMS one day. Everything you look for in a BMS is in a batrium. 😊
You are the second TH-camr today that said something like that. I really don’t get why-particularly in a safety critical system (boat power and/or critical loads (nav, comm, surveillance and lighting)). If contactors are the problem, there are FET based devices which can do the same function, albeit at lower power.
I bought you a beer! Because it is every time nice to watch your videos! My deviation problem is under control: it is the epever charge controller, witch can't be controlled by the bms! I think about to throw it out an install a second growatt spf in series to the existing one instead! What do you think about this? Best greetings from Germany, Andy! 😎
Hello Andy! Thank you for this nice summary! You have already been able to show us a lot of secrets of various BMS's. Small question: in your videos you regularly show the Victron Remote Console on a touchscreen. I couldn't find it on your website. Which touchscreen do you use for this? Is that the official Victron screen? Thanks in advance for more info. Sunny greetings.
Thanks a lot for your feedback. The touchscreen is just my old Android tablet connecting to the VenusOS IP over the network showing the remote console. The tablet is an old Nexus 7, almost 10 years old. It still works perfectly fine
The missing DC load record is a Victron problem. By the way, for JBD there is a driver 'dbus-btbattery' that connects the BMS via Bluetooth with the Cerbo / Venus.
Wiedermals ein super Video von dir, alles auf dem Punkt gebracht was wichtig und richtig ist. Schade das es kein ähnliches Gehäusedesign von Gobel Power gibt wie das von Seplos. Wäre das BMS passend für das Gehäuse hätte ich theoretisch zugeschlagen, bis zum heutigen Video. Ich bin gespannt was du da für BMS noch gefunden hast. Bitte mehr davon!!!! 👍
@@OffGridGarageAustralia Hello Andy, Thank you very much for your effort, I'm curious to see if Gobel Power will also bring something out. Otherwise a new front shouldn't be that hard to make either. But maybe you've found a better BMS, I'm really looking forward to the next videos.
Hi, thanks for all the great videos. What would be the best bms for my boat. I was planning to use a Jk as you use to love them but then I see “jk bms go home..” I want to use it in a boat for 16 cells from eve. What do you suggest. Many thanks.
I had issue with the Overkill (JBD) and 280A Eve cells. I got a JK active balancer BMS and a Heltec active balancer. I think active balancers are almost mandatory for larger cells.
you reading my mind? I was just about to start investigating BMS's and thought of your channel.... then.. voila. This video appears. Do you have a spreadsheet of pros/cons or a feature matrix?
Thanks Stephen. I haven't got a spreadsheet with all these information. I'm sharing my own experience and what I thinks is a con maybe not one for others. It all depends a bit what battery design you have.
@@OffGridGarageAustralia For me no. On the average day I am fully charged 3/4 through the day (rest of Sun mostly waisted). I don't balance on low sun days. The software starts balancing when average voltage goes over 3.39 volts and stops at full charge (3.45).
Yah, I agree it definitely should not turn off charging based on any internal SOC tracking. I can guess where it came from though... these BMS firmwares are typically not written from scratch by a battery expert, it was probably copied or stolen from somewhere and then modified, or they commissioned it from a boilerplate. So the likely reason is that the design is left-over from a drop-in-replacement function for lead-acid. And in that situation the charging source might never disconnect so the BMS has a disconnect at 100% SOC. You've really never liked the low balancing current BMS's have, but I think it might be a bit of over-optimization. If normal charging specifications are followed, meaning we charge to 3.55V and hold there for an hour or two each cycle, and assuming the BMS is sane and starts balancing above 3.50V without any silly restrictions like only doing it while charging current is present, then the whole system works perfectly well and no active balancer is needed.
But you would need to keep the battery at 3.55V without anything connected but a power supply to keep it stable on 3.55V per cell. As soon as it is still charging or discharging a tiny bit, the balancer will not work. These undersized balancers come from Li-ion where you can start balancing at any voltage. We will see the first of these kind of BMS with an active balancer soon...
@@OffGridGarageAustralia Right, but generally speaking if the charge gets to that point, holding it there for a while isn't really a problem. Once the charge current tails off (30 minutes or less), then holding at 3.55V doesn't take any more energy than holding it at 3.375V would. So its just a matter of at what point in the day the solar gets it there in terms of holding for 2 hours to give the BMS's balancers sufficient time. And if there is extra time after that the charger simply drops to the float voltage of 3.375V or 3.4V and continues holding there. You have the same load-support capability either way.
Oh boy.- I hope this BMS programers check this video! THX Andy for that great video! I wish the JK BMS had WIFI or an ETH connector + easy (independent) PC software which connects parallel AKKUS.
Yeah, a BMS with FW upgrades is already a winner. The software should not be a problem. I don't understand why none of these manufacturers or sellers get it right though.
I'm confused I guess also needing advise. I have (us model in usa)growatt 5000es with 6 signature solar lifepower 4 batteries. Batteries are all comm linked thru the rs485 port (one designated master)hooked to the 5000es rs485 comm port. Works flawlessly so far. My question is with the communication does the master battery talk with the other batteries(bms to bms I'm assuming)and decide which battery is to charge and which is not in order to keep them balanced as a bank and the bms within each battery to keep its 16 cells in balance?I cannot set any charge parameters or cut offs this way it is all set thru the inverter. Secondly if I wanted to build my own 16s 280ah battery using catl cells. Which bms would you use and would any you know of communicate with the growatt 5000es.I do not want to parallel the inverters. Just use as separate stand alone systems but each having communication with their battery packs.I plan on running all my 240 loads on one setup while using the other for my 120v loads. I don't like(trust)having all the inverters paralleled together and risk having total failure. I have 3)5000es and plan on getting a spare.Use 2 for 240v and a single for 120v then have the spare. Any thoughts. Sorry so long winded. Howdy from Texas!
Thanks for your question, Chris. The communication does not go that far. It is more to pass error or alarm states on to the inverter, show an overall capacity and also the overall deviation across all cells combined. The master BMS will not regulate the charging of single batteries though or controls the balancers. That is still part of each individual BMS in your batteries. Any BMS supporting the Growatt protocol via CAN or RS485 will work. For example the Gobelpower BMS as well as the Seplos BMS.
@@OffGridGarageAustralia so other than the soc that it relays to the inverter as well as it has preset cutoffs and charge limitations from the inverter your really not gaining anything you cant do yourself? Also how exactly can I determine what protocol the growatts are using for comms? They use rs485 ports but I'm assuming by protocol you mean programming/language they are speaking lol.
@@chrislewis9345 In the battery, you select the brand of the inverter, or the battery protocol used by the inverter. I don't know all inverter selections, so maybe someone else can give solid fact, but you'd probably select Growatt, or maybe the Growatts employ the Pylon battery protocol. It's a protocol not a programming language or API. It how the settings are conveyed from the battery and understood by the charger or inverter.
A protocol defines the exchange process as far as the order that each device will communicate (request -> answer, field order [ie status might state the voltage followed by current draw etc in one BMS while another may send current draw followed by voltage etc]), the format used to express that exchange (is a number an integer, a floating point number, a string of characters, how many digits of a number represent the data before and after a decimal point etc) and other metadata like delay time required between each conversation, delay time expected between request and answer and so on.
Hi Andy, i have watched just about all your videos and really like your way of showing the fantastic world of solar power, i have a question if i may ask, '' do closed loop communication BMS's connected to the inverter charge multiple battery packs based on the SOC in order to charge the packs equaly ?''
I think you have sent this email too, right? I just answer it here again for everyone. No, they only charge per voltage. Parallel batteries balance themselves without problem. One banks has a bit higher voltage and will take less charge until the others have caught up. There is no communication between the BMS and inverter necessary for any of that to happen.
@@OffGridGarageAustralia Hi Andy, im so sorry for doubling my question, yes i did see your email and thank you very much for replying, it was important for me to know this, i just installed 20 kw of solar panels, 24 kw (4 x 6000W) hybrid inverters and 40 kw of Lifepo4 battery's, im learning every day new things thanks to people like you 🙂
I love your videos, but I must correct you since i have been seeing 10 videoes where you say seplos stop charging and reset to 100% only at 3,65v. I have a seplos case and it reset to 100% at 3,50v, because i have set the total battery voltage to 56v. When it hits that it resets to 100%, also when I have set the cell ov protection to 3,6v. The seplos reset to 100 percent when it hit an overvoldtage protection for the cells or total voltage you have set yourself. I use it with the 10A limiter but my deye 12kv 3 phaee inveter wants to take things easy and only charge with 5A until the battery does not absorb anymore. After that, the battery goes down to around 3,4v. Best inverter ever, 17A on each phase if all phases is loaded 100% or 27A on one phase if the total load is under 12000w forever. And it can do 200% for a small amount of time on both the 17A and 27A. And it can do all sorts of smart things, like use the generator input as smart load that turns on when the battery reaches a preset % and off again at a preset %, I use for charging a tesla. It works with alot of batteries like seplos. It have the shunt build in so it can run with no connection to the bms in % mode. It can take 18kw solarcells and run another inverter like micro inverters on the generator input, even in off grid mode I run the seplos with the neey 4.0. I got some very good eve280 cells there dont drift much. I could run it without the neey and I did for one month because I turned it off by mistake, and the call was only 25mv from each other when the battery was charged to 100%. The settings in the seplos bms its set to that the balancing starts at 30mv and stop again at 20mv, so that is exactly what it was set to do. I have the neey set to 8mv because I like the the quick balancing the neey offer, then the seplus work as a backup balancing if the neey stop working or more then one cell runs off. I have not updated the bms or anything, and it does just work. The system runs as an off grid system now, until the permit to export to the grid comes.
The only gotcha with the 4S JKBMS is it seems to have a significant current draw. Did a test of 100AHr 12V 4S LiFePO4 with full charge and left it for 3 months and it lost 40% of capacity - fortunately there is an on/off switch that you use to easily switch on/off BMS. I have a suspicion that the current draw is due to the inefficient DC DC converter that produces >21V as this was the voltage required to drive the BMS in 16S versions.
JBD bms tells you it's 100% according to what you program into it. Maybe you need to revisit it Andy? The last one you looked at was really small and soo old. My batteries reset at 3.6 volts, where I need to program them because of the Canadian climate, instead of the 3.45 you use. We get much colder here then Australia and 3.45 is too low to charge completely at anything below 8 degrees C. Program the new JBD bms, at 2.8 to 3.45 (as an example). Test capacity by going from full (3.45 ) to empty (2.8). The JBD will now read and communicate the AH stored within those parameters as 0 to 100% . There is no need for the BMS to tell the Victron what to do if the Victron system is programed correctly. The inverter should always turn off before the battery, this is easily achieved with voltage parameters because of the LiFePO4 discharge curve. The inverter throttles down any loads at shut down, preventing spikes to the battery. Charging MOSFETS turning off is essential. Relying on an outside device to stop charging is added complication that invites disaster if something goes wrong with the communication.
I'm in contact with JBD and see what we can organise... That is the whole purpose of a smart BMS to tell your inverter what to do without having to program your inverter (or Victron) in this regards. I don't think it is essential to turn off MOSFETs at all. If the charger follows the BMS request it will just stop charging at the set numbers. If not, the BMS is there to catch the battery and disconnect. It's a 2-step system.
@@OffGridGarageAustralia I see how in your situation this will work fine, and might be the goal. In other situations, say for instance if life safety equipment is involved (like we have on ships), this setup could cause major problems. Lets say your battery shelf with 3 batteries is your main system in this example, and your backup system is the Frankenstein battery. Your main system has 3 BMS's that communicate to turn off the inverter. One of your three batteries, has a failure and turns off the inverter. You switch to the backup Frankenstein battery, because someone's life depends on having the inverter running for their medical equipment. The inverter doesn't turn on. You quickly isolate the defective battery with it's circuit breaker and switch back from your back up Frankenstein battery to your main battery. Your inverter still doesn't work, it's being told not to by the BMS of the problem battery still. Somebody dies before you can reprogram things. The same problem happens if you run the main battery bank to empty then try to switch to a backup battery. The empty batteries will tell the inverter to stay off even though a backup power source is available. By using the voltage profiles of the LiFePO4 as the triggers for the inverter, you ensure that a backup system will turn the inverter back on in an emergency. JBD makes a BMS capable of 750 amps peak, so you're not going to blow any MOSFETs like you have with other BMSs. If we are talking about massive storage installations, Victron isn't involved and neither are any of the BMS's you and I are using. A much more complicated computer control system is used for these.
@@boatelectricaldiy Thanks for your comment. OK, so one single BMS will not turn off the inverter, it will turn off the faulty battery bank and the inverter keeps running on the remaining batteries (we have tested this with the Seplos batteries). Also turning off the breaker or battery switch for one battery banks should also isolate the BMS. If not the order they are wired is wrong (or bad design). Switching from the main battery to a backup battery should always start the inverter, otherwise it is useless as a backup as you explained. Life saved, maybe?!
@@OffGridGarageAustralia it seems my lack of interest in other BMS's has caught up to me here. The price and size of these BMS's really put me off, and I I haven't had lots of time on my hands so it seems I've missed a vital bit of information that you have presented in your videos. I guess I should go back and pay better attention while watching. I still don't see my customers being interested in this, but you've peaked my interest yet again. Thank you, keep up the good work!
@@OffGridGarageAustralia I have been able to revisit this video. My company has expanded from just marine electrical installations to remote offgrid home installations now too. I was able to locate the manufacturer of the BMS that these company's use, and have ordered a few for testing. Thank you for the great videos, it's rare that I find something on TH-cam that makes me reconsider my business model.
Hey Andy! Given your current knowledge, would you still recommend JKBMS or not? Do you know which hardware and software version of the JKBMS you are using in your system?
The JK-BMS is a very different device in comparison to these larger BMS. As a rule of thumb: use teh JK-BMS for 12V/24V batteries or if you have only one battery. If you have several larger batteries in parallel and want more safety and also communication, Seplos, Gobelpower and co. are better for that purpose. Mine is a very old JK-BMS, one without a power switch or display option.
@@OffGridGarageAustralia Thank you very much for this information. I already have several JK BMS versions in use and I am of the opinion that the current hardware/software version 11.x finally calculates the SOC value more sensibly. Apparently the problems of the older models have improved significantly here. I would also like to test the Gobel Power BMS, but since it doesn't have Bluetooth, I'm still considering it. I am currently in the planning phase for a 43kwh storage system that will consist of 3 individual batteries. I'm really undecided on which BMS to use ;-)
@@cyborgxxl Thanks for sharing about the improvements with the JK. With three banks you could still use the JKs but won't have any comms or elevated safety features like with the other ones. The Pace, Seplos and Gobel are all very close together and at the moment I would go with the Seplos just due to the fact, that it works with my Victron system and the connected DC load. This is a major drawback for all the other BMS. And we also have found a solution to overcome some of the problems with the Seplos BMS now. Well, one of my viewers has found a pretty good workaround for the Seplos...
@@OffGridGarageAustralia Ok, thanks for the detailed answer. I think maybe I'll go for the Gobel or Seplos BMS after all. I only use Victron components. Greetings from Austria!
@@Der_Hannes Yep, I got my OVA to 3.55, OVP to 3.6, and fullpack voltage to 55.2 and current to 1A, and the view is to slowly increase the FPV to 56 over the next week so the built in balancer does its job.
@@sarahjrandomnumbers my FPV sits on 55.2V and FPC at 1A cause i have added a NEEEEEEEEY balancer wich of course is doing a great job. Anyway, the SOC resets to 100% just if both parameters are reached. The only downside i noticed is that the charging mosfets are turned off after reaching that "Fully-Voltage". That should only happen at a protective situation
@@Der_Hannes Just fully charged, and it didn't turn off the charge fets, but the remaining current and presumably SoC reset once it hit the FPV and FPC levels. Recorded it just in case, i would post the link here but youtube is blocking it, it's on my channel anyway. That also includes the view of the BMS settings I have at the moment for those that want to compare. No cell/pack alarms or protect, and the balancer stayed on until it went under 500ma, as expected. Currently at a deviation of 13mv, compared to 270mv when I first got it, although I did open the pack up and charge every cell individually to 3.6v, then let the internal balancer do its job. Although, thinking about it, it's not really a balancer, all it's doing is adding 43ohm to the cell, which makes it look like the internal cell resistance is much higher, meaning it'll charge slower, which makes a whole lot more sense than it draining power from the cell to balance it. But that's a different conversation :).
Very interesting video I have 48v byd lithium batteries 24v each which are connected in series to make up to 48v I misplaced the bic card which means I can't use the bms in the Battery pack. I want to know if there is a way to build my own bms in both packs. Thank you
@OffGridGarageAustralia I don't have access to the cells the batteries arr still new and never been used before.i have this connector hanging from the cells to the bic card then to the bms am not sure if that can be used instead.if not I have to take the Battery apart
i swaer your videos are like the movie "Inception" videos inside of another videos inside of a short editing video 😂😂don't blink or you will mis the whole video Love it.
Im acting as a human BMS by discharging and charging each cell when they become unbalanced, using a car light and a power supply. I can balance the cells in no time and perfectly. The cells will most likely become unbalanced quicker then the time it took to balance them when i connect it to my Solar system. I have a Hellcat 4s Balancer but thats only for top balancing. Im going to have to buy something i can program. Maybe the BatGo ?
What would you think about making the JK-BMS to be able to do the same. Like reducing charge current when it comes close to overvoltage or reduce charge and discharge current when it reaches a critical temperature. Let me know if you are interested.
@@OffGridGarageAustralia But i think i could make it to communicate with the inverter and tell it to reduce the current in these cases, by letting it speak the Pylontech protokoll. I guess your Victron system can understand Pylontech as well.
@@ackmaniac5101 Yeah, you can let it communicate to the Victron system and other inverters. I made a video about it a while back. The question is though, how do you handle several battery banks in parallel and how do they report combined back to your system?
Im still a huge fan of the electrodacus sbms0, its tiny but powerful and leaves all the heavy lifting to the devices that are meant to do it.. I really dont get why this is the only bms that works like that.
Not true my JK BMS shows 100% Charge state for 16S at 54.97V and 3.439V on avg. The JK BMS can also nicely connected via Eib Kabel and USB Port to a RPI4.
Oooh! Another BMS SOC problem? Where have we last had it? Oh, yes! Outlander :D Ps. Anything coming on the Teslander soon? Interesting updates? Degradation of the battery, perhaps? 😎
@@OffGridGarageAustralia Great news Andy. It’s NOT something I lose sleep over like some, and nor do I do deep dives with software programs and OBD readers etc- I suspect we’ve lost something more like 5 to 6% but it was ALL in the first year. Since then the “full range” hasn’t altered more than about 1km, and always goes up again for a while after a long trip. (Our car has literally never in it’s life been above 95% as I like to maintain some decent regen, and only below 20% maybe three times.) This is in line with many overseas users with very high kilometres- initial degradation over the first year (ie: reduced projected range) and then it levels off very dramatically. I do NOT miss the Outlander’s battery issues and bloody useless BMS - though it was otherwise a pretty good car.
my reply got deleted?! I am using both xiaoxiang and overkillsolar apps. Setting you shoul check is single full voltage or cell full voltage or something similar depending of version of app used
6:10 ich widersprech zwar ungern aber das stimmt nicht was du sagst , bei meinen 3Seplos hab ich derzeit jeden tag 100% mit 56V, es dauert ~1,5h von 95Soc auf 100%Soc, alles gut
Hi Andy, Great vid, I went to look at your GobelPower BMS link to buy one, but they're all for 16S, is that corrent? I didn't see a 4S12V system, so I guess my option is the JK one you reviewed about a year ago, this one: "The JK-BMS 4S-8S Edition - World Premiere. The best BMS now for your 12V LiFePO4 battery system" So I'll order a B2A8S20P unit, with your link. For that unit, do you know if that 4.3" screen that they sell is any better for the amount of information that it shows versus the smaller 2.3" one with the on/off button you showed in the video?
These sort of BMS are all 16S for 48V battery system and communication between them and an inverter. Yeah, for a 12V or 24V battery, use the JK-BMS. that is a great device and will work perfectly with its active balancer. The screens are really optional as you can always connect via the app. There is just basic information on either screen. They are great if you install your battery in an RV and also other persons can read the information of SOC. And the on/off button... it again depends how you use your battery but the BMS should be staying connected to it all the time anyway. And thanks a lot for your support and using my links
I'm quite surprised that in this day & age, BMS manufacturers are still making those "common" mistakes. And they often are all making the same mistakes, which I find very worrying. The fact that their firmware on the BMS often can not be upgraded is a very concerning one as any bugs discovered, and you have found many, can not be fixed for their existing customers. I think the best thing we all can do is cease buying such BMSses and focus as much as possible, for the Gobelpower solutions. (and JBD haha)
Like we had with the NEEEEY balancer last year, where it took them 4 (hardware) attempts to get it right. How good would a FW upgrade function bee in these devices. I don't understand either why none of these develop a 'perfect' BMS and listen to the community, be on the ball with new FW upgrades and support. They would sell like fresh baked, hot cookies.
@@OffGridGarageAustralia I totally agree with you, a good BMS would be a true bomberfest, like the GobelPower seems to be. Thanks to you & your channel when I was about to purchase a Daly BMS, it was one of the first videos I found you discussing it in informative details. That was a very cold shower when I watched that video and opened my eyes to dig deeper & further and I got bitten by the LiFePO4 & BMS "bug" due to you dear Andy & your wonderful channel (& informative website!).
After discussing the translation issues with several companies in China, I received an email for the "Central department of export control" of the Chinese Government to "Cease and Desist" suggesting that individual manufacturers to use independent translators. "All translations are completed correctly by the central translation offices of the export control offices by appropriately trained professional translaters" I stopped my correspondence immediately as I feared for the safety of my correspondents. You can draw you own conclusions.....
Keep your cells balanced and rely on a real battery monitor like Victron to do SoC measurements. BMS's with a bunch of cheap non-zero temp compensated paralleled chip resistors for current shunt and non-auto offset zeroing sense op amplifiers for shunt resistor voltage sensing is not a good way to make an accurate SoC indicator.
@@OffGridGarageAustralia consumer grade bms just do not cut the mustard. Ive got 50 kwh batteries that unfortunately are at 24 volts due to everything thats powered by them being 24 volt, replacing inverters and everything else for 48 volt would cost far to much
![[#2024MAMA] BIGBANG (빅뱅) - 뱅뱅뱅 (BANG BANG BANG) + FANTASTIC BABY | Mnet 241123 방송](http://i.ytimg.com/vi/EQsYfiPR9uM/mqdefault.jpg)
We are in a very interesting time with residential power storage. It's like early pioneering times. These companies build all these great hybrid inverters and BMSs etc without really knowing what they are doing, or really understanding all the challenges. Remind of the the PC days back in the late 80s and early 90s, just when things started to heat up. We have exciting times ahead with all this solar related tech.
It is definitely an exciting time and our pioneer work will hopefully help to speed up this process and get as many people on board as possible.
@@OffGridGarageAustraliathanks, Andy!
Thanks for all the testing and reporting back your opinion, very helpful!
Let’s hope all these bms manufacturers take note, it’s in their best interest, thanks for all your testing Andy🇦🇺
I bet they won't. None of them is making the software themselves. They all reliant on 3rd party companies which may have no knowledge about battery charging and BMS design.
@@OffGridGarageAustralia my seplos 10e bms reset soc 100% with high voltage alarm setting (not sure cell or pack, but they reach 100% soc without triggering the cell or pack high voltage protection ) which is separate and lower than the high voltage protection setting. maybe you have the old revision/firmware.?
@@Pey5531 I'm not aware that there is a FW update for the 10E.
Fantastic video Andy! Thank you. You are doing a great service to the community of people that need help with learning everything solar!
This BMS seems very impressive for sure!
One thing that I worry about with BMS' is ONE setting... Delta cell voltage shutdown should be implemented. If you have a runaway cell it can take down the whole pack before shutdown happens at a certain pack voltage.
I think I have said that I use a "naked" BMS monitoring system. Instead using Home assistant, I can set a shutdown if there is a DCV of a certain voltage. Pretty cool eh?
Anyway, this is just the way I do it.
Thanks again for your content. Together as a community we learn! 🐸🐸🐸
Thanks, Dan. Using HA for that is pretty cool. Do you still have any hardware protection as well measuring the cell voltages and turn off MOSFETs in case of a runner? Only using HA might be a bit risky?
I have seen DVC settings in other BMS but is this not a double up because if one cells runs away, the BMS will catch it anyway?
@@OffGridGarageAustralia Yes my inverters actually have a setting where you can specify a voltage that will disconnect the loads if the setting is reached.
Other than that I'm running naked lol. But I don't really see a need for a physical BMS the way mine is operating..
Hey Andy,
already left you couple off comments on some other video about this,
I never use total voltage protection on any of my BMSes when single cell voltage protection is available,
just disable the switch in the Seplos BMS set the overvoltage to 55V or 54.4V or 54.8,
now when the overvoltage is triggered, it doesn't disconnect charging & battery is reset to 100%, if cells are somewhat balanced this means its actually 100%, the only issue is there is no separate floating voltage option with time limit.
& since a battery with well balanced cells will drop current gradually allowing it to absorb for a longer time before reaching the actual "overvoltage protection value"
I compensate for the floating absence by lowering the "charge over voltage protection recovery" to something like 53.6V
like this since my system is well sized panels/batteries/load my batteries will reach almost 99% and float for about 30-60 min before reaching the "overvoltage protection" voltage, thus resetting to 100% and stopping the charge current.
Yes, I replied to you comments already. I will do some more Seplos testing soon. That is a great suggestion and 'trick'. Thank you for sharing this.
The missing floating option is not great but also not a big problem. Usually people turn on loads when the battery is full anyway, so it won't stay there for too long.
Using the recovery voltage... so when the battery is full, you just use the solar power for your loads and the battery voltage will slowly go down due to the fact that the charge MOSFETs are turned off. Ist that how you use it?
As Always, thank you for your beautiful and worth knowing broadcast👍
I'm looking for a BMS for two banks of 16p 280 AH cells and I came across this channel. Fantastic information, great reviews, great channel.
@Off-Grid Garage - thats true the Gobal BMS is one off the Best BMS on the Market right now when you what to parallel more Battery Packs. On an 3 Phase Victron System the Solar Input to Grid is possible over the Solar Charge Controller when the Battery is 100% full - so the BMS is not blocking the Feed in to Grid.
The 100% Reset and turn OFF MOS is not the best solution because on an 3Phase Victron System with a Fronius MICO GRID Inverter when the Gird is Cut off and you have full Solar Power the Victron 1:0 Rule makes for a few miliseconds a higher Voltage on the DC side until the Inverter are throttled town by the rising Hz from the Multiplus. When the MOS are closed the can not drop the incoming Power to the Batteries.
Also the 100% Turn OFF MOS is the natural way of Float Voltage - when the MOS Stays ON we need a extra Permameter for the Float Voltage to set for the the VE.Bus. I'm also FAN of your theory to keep the MOS ON at 100%
Maybe Gobel Power can implement a Set Float Voltage for the Victron VE Bus when Batterie is at 100% and a NO Close Function from the Charge MOS at 100% SoC!
Hello Andy
My 2 Daly 150 A BMS are ok also they do the same job .
With 2 * 16 LFP 280Ah from gobelpower.
Connection UART USB to the rp3 .
Bmv 700 and muliplus .working fine....
Also to charge my model S off grid.
Nice, that's sounds like an interesting setup you have!
Thanks Andy- a really great explanation ❤️ as usual. BIG 👍👍👍
Heya, it looks like for parralleling batteries you found the only bms that does olmost everything we need for or systems thanks for testing for us all
Well done. Great summary. Thanks for sharing
Hi Andy, you say the Gobel Bms is the best, because of the software, because the SOC reset can be determined. On the other hand, you mention but, the Gobel BMS switches off the MPPT and then uses only the battery, unlike Seplos, for example. For me, however, remains in total the Seplos the better BMS, because not only more user-friendly through Bluetooth, but what is much more important, the Battery is full in the summer at noon and it follows another 8 hours SUN, which I can immediately consume or feed and sell, the Gobel BMS can not, the whole afternoon is then lived on the battery, although the sun is shining and I could use the extra yield. For me, the Seplos remains the price performance winner. Thanks for your work and thumbs up from the crazy Germany to Australia
Thanks for sharing. Yes, that is correct. I've just summarised all pros and cons as far as I could test them.
It is very difficult the compare BMS as the conditions often change when charging and it is never the same.
I'll get back to the Seplos soon... more too come.
Andy, give me a shout to discuss DIYBMS!
this comment needs more attention!!
Who are you again?
Thank you for this update an information very interesting. Have a great week, M
I agree with everything you say here. It's ridiculous that Daly & Jk use the same silly HVC to activate 100% SOC. I've had to fine tune my Daly BMS to do a cell HVC of 3.47v to trigger 100% SOC, but then the charge MOSFET shuts off.
Please Daly & JK - LISTEN to us!! 😡
I still feel the pain i brought one of the first models of the daly smart bms at a cost of $600 aud. Its now a paper weight lol.
I now have 3x jk bms. If jk could do communication it would be the best peter bms.
I wont be buying a new one until there's a bms with communication and active balancer and can monitor multiple packs in series and in parrel. How good would it be to have one bms for your 3 16s modules.
Ps thanks andy your the battery king of knowledge
You get what you pay for as the saying goes. Been enjoying all the BMS testing.
No bluetooth is a big no go for me! My solar shed is outside. The BMS I use from Battery Hookup has bluetooth, so I can keep an eye on the cells from my office in the house.
Yeah, a bit annoying but not critical for me. Once set, I don't need to login that often any more...
At some point I didn't care about the BMS' SOC anymore. Instead I build myself a Smart-Shunt that tracks the SOC and reports it to the GX device.
I check the BMS from time to time for the cell voltages, but that's it.
That's what I do with the Victron shunt and my battery shelf. The single SOC of the BMSes are off by a good margin.
This video was 7 months ago - Have there been any recent developments or is there any other BMS that you would recommend over the Gobel at this point?
I went with the seplos a few weeks ago, tested the CAN connection to Victron and diceded to forget that and go with an extra smartshunt. The only thing i miss are the values for the highest and lowest cell but the SOC is calculated perfectly by the shunt. And this setup will work with any batteries in parallel - like your battery shelf. I am not 100% clear about the usefulness of a data connection to the battery.
You can still get all the data from the Seplos BMS but use the Smart Shunt as your battery monitor. That works great. Just leave the Seplos connected via CAN but turn off the control part in the Victron System, so it does not control charging etc...
@@OffGridGarageAustralia thanks Andy. You mean not to choose the bms as the battery monitor, right? If I remember correctly, that does not work, because the system priorized the bms in my ESS configuration. But it could be, that the turned on DVCC was responsible for that. But I need/want to turn it on, because otherwise my RS450/100 is not under "external control" what it should be (the 150/45 IS under ext. contr.) plus I cannot control the overall charging current.
@@OffGridGarageAustralia today I set up my BMS connection again and all the sh*** beginns again :-D
I figuerd out, that I have to set the option "Controlling BMS: No BMS control" to ignore the values coming from the BMS (CVL, CCL, DCL). But you must know that you only find this after rebooting the cerbo. And after that, you have to reboot again but not only the cerbo, but also the MP and reset all MPPTs to ignore the BMS. What a mess.
And now the best: I cannot find the min- and max-cell-voltages anymore :-D
@@mario-q6k9w Hmmm, strange, I never had to restart the Cerbo or MP when changing the battery monitor or other related settings.
Yeah, the controlling BMS function should be off, so you can still use the DVCC with custom settings. As soon as you do that, the chargers won't be under external control any more. Either the MBS in in control or not.
@@OffGridGarageAustralia yes, that works now.
If you have BMS controlled MPPT and either set BMS control off or disconnect the CAN BMS, then the MPPT have to do a rescan, otherwise they lack of information and report an error (bms missing or something).
And maybe the MP needs a reboot because of the ESS assistant.
To see the the min/max cell voltages reported by the BMS in remote console I can't find the "details" under the BMS anymore and in VRM is no data available. Did you know why that is? I use the 10E BMS
Hello Andy, I have 2 batteries ( 5KWh and 10KWh) the 5KW is lifepo4 while the 10 KW is lithium ion. They are connected in parallel using 2 jk bms's and they work 24/7 . I have not noticed any problem with them. BTW.... half of my house is totally off grid.
Is your LFP battery at 16s and Li-ion at 14s configuration? What voltage range you have in your parallel connected LFP and Li-ion batteries?
@@marek8928 yes you are right. Max voltage in both is 55v and minimum is 42v
Yeah, you can parallel them if they have a matching voltage. However you will never get the best out of your batteries, it's always a compromise. The charge and discharge curves are very different and while the LFP battery is charging to 95%, the Li-ion battery will not charge at all until the voltage rises. Then the LFP will stop and all charging power goes into the ion battery. That can lead to high amps for either battery at various states of a cycle.
But interesting project, I may look into this...
@@orkidorkid 55V on top is OK for LFP and Li-ion but 42V at the bottom is kinda low for LFP (it is OK for Li-ion thou). I would not go lower than 48V.
I have 20kWh LFP battery and I am planing to connect to it Li-ion battery in parallel (I got it dirt cheap). When LFP battery works between 55 and 48V and it's capacity is almost fully used, the Li-ion battery will only be utilized in circa. 60%. For me it is OK because Li-ion battery is than very safe.
Hi Andy, ich habe es mir geahnt, dass du ursprünglich aus Deutschland kommst. ❤
Was ein Quatsch! Link in the video description of aaaaaaall my videos. 😉
My daly BMS's are still going well no problems at all
Great, what Daly's are they?
Hallo Andy,
I dont know how it works, but my JKBMS always reach 100% SoC when the charge controller of the Deye is in the absorbtion phase. The Values for BMS and charge controller are set like you show on 5:52
JK is very different...
@@OffGridGarageAustralia
...but still seems to be pretty good if the budget matters :-)
If you are interested, i can send you curves of a charge cycle including the voltages of every single cell.
@@Juergen_Miessmer I like the JK as well as I said. It's just another BMS for a different battery design.
Idem, my JK BMS reach the 100% every day and I never get an OVP alarm. I'm charging at 3,5V and JK BMS OVP setting is set to 3,6V.
@@M-Clem Do you know what triggers these 100%SOC reset in the JK? Lot's of people are asking for that.
Did you see JBD new BMS similar to this one . they call it : "Energy storage BMS" , I have been waiting your review on that
Yeah, there are a few on Ali now, I have seen. They also contacted me to review it. I may do it in the future...
18:44 hey Andy, so, if it’s not one of the “high-end” manufacturers, i think it’s some new China bms. It must be the new EVE BMS, because as a Cell manufacturer they should now what functionality the Software must have. And I assume they probably want Andy to promote it.
I’m excited to see if I’m right.
Good guess...
@@OffGridGarageAustralia it seems to be of Chinese origin, due to some of the writing on the big box.
@@davidpenfold Are they not all coming from the motherland? 😊
That mystery BMS is driving me crazy :-) Can't wait to see whats in the box.
@@JOHNDANIEL1 Don't agree with that. BMS is a safety feature and comparable with an airbag in a car. If everything is working fine, indeed you don't need the BMS. However if a cell, for whatever reason goes out of line, the BMS kicks in and avoids much larger problems.
@@JOHNDANIEL1 the airbags in my car never deployed yet. Happy for that. Still have no intention to remove them, or get a car without airbags.
You always need a BMS.
If any questions for BMS, welcome to contact me.
Hi Andy. Great video. I've ran quite a few BMSs systems. JK, ANT, seplos and a Batrium You must review a batrium BMS one day.
Everything you look for in a BMS is in a batrium. 😊
I'm not a fan of the external shunt and contactor situation of their design, tbh...
You are the second TH-camr today that said something like that. I really don’t get why-particularly in a safety critical system (boat power and/or critical loads (nav, comm, surveillance and lighting)).
If contactors are the problem, there are FET based devices which can do the same function, albeit at lower power.
I bought you a beer! Because it is every time nice to watch your videos! My deviation problem is under control: it is the epever charge controller, witch can't be controlled by the bms! I think about to throw it out an install a second growatt spf in series to the existing one instead! What do you think about this?
Best greetings from Germany, Andy! 😎
Hello Andy!
Thank you for this nice summary! You have already been able to show us a lot of secrets of various BMS's.
Small question: in your videos you regularly show the Victron Remote Console on a touchscreen.
I couldn't find it on your website. Which touchscreen do you use for this? Is that the official Victron screen? Thanks in advance for more info.
Sunny greetings.
Thanks a lot for your feedback.
The touchscreen is just my old Android tablet connecting to the VenusOS IP over the network showing the remote console. The tablet is an old Nexus 7, almost 10 years old. It still works perfectly fine
The missing DC load record is a Victron problem. By the way, for JBD there is a driver 'dbus-btbattery' that connects the BMS via Bluetooth with the Cerbo / Venus.
Why does the DC consumption work with other BMS though?
Yeah, I have shown the connection of Daly, JK, JBD,... in a video a while back.
Wiedermals ein super Video von dir, alles auf dem Punkt gebracht was wichtig und richtig ist. Schade das es kein ähnliches Gehäusedesign von Gobel Power gibt wie das von Seplos. Wäre das BMS passend für das Gehäuse hätte ich theoretisch zugeschlagen, bis zum heutigen Video. Ich bin gespannt was du da für BMS noch gefunden hast. Bitte mehr davon!!!! 👍
Thanks Matthias. I've asked Gobel for a similar DIY case. I think they look into it so watch this space...
@@OffGridGarageAustralia Hello Andy,
Thank you very much for your effort, I'm curious to see if Gobel Power will also bring something out. Otherwise a new front shouldn't be that hard to make either. But maybe you've found a better BMS, I'm really looking forward to the next videos.
Very nice! When I finally go lifePo4 I think this’ll be my bms 😊
Hello Any, any chance you will test the 10A QUCC Active balancer? Maybe a cheaper replacement for the NEEEEEY
I did test the 10A Heltec active balancer a while back. Garbage though...
th-cam.com/video/-X1xh8CUuK8/w-d-xo.html
Hi, thanks for all the great videos. What would be the best bms for my boat. I was planning to use a Jk as you use to love them but then I see “jk bms go home..” I want to use it in a boat for 16 cells from eve. What do you suggest. Many thanks.
I had issue with the Overkill (JBD) and 280A Eve cells.
I got a JK active balancer BMS and a Heltec active balancer.
I think active balancers are almost mandatory for larger cells.
Yes, I think so too. It may work with very well matched cells in a 4s installation, but 16s... there are for to many cells which are drifting.
Hey Andy, what about BMSs on Deye and Pytes batteries?
you reading my mind? I was just about to start investigating BMS's and thought of your channel.... then.. voila. This video appears.
Do you have a spreadsheet of pros/cons or a feature matrix?
Thanks Stephen. I haven't got a spreadsheet with all these information. I'm sharing my own experience and what I thinks is a con maybe not one for others. It all depends a bit what battery design you have.
I am working on a heavy duty diy bms that uses 1 ohm power resistors on each cell (which gives about 3 amp balancing current).
But passive balancing? Is that not a thing from the past? A bit wasteful...
@@OffGridGarageAustralia For me no. On the average day I am fully charged 3/4 through the day (rest of Sun mostly waisted). I don't balance on low sun days. The software starts balancing when average voltage goes over 3.39 volts and stops at full charge (3.45).
Yah, I agree it definitely should not turn off charging based on any internal SOC tracking. I can guess where it came from though... these BMS firmwares are typically not written from scratch by a battery expert, it was probably copied or stolen from somewhere and then modified, or they commissioned it from a boilerplate. So the likely reason is that the design is left-over from a drop-in-replacement function for lead-acid. And in that situation the charging source might never disconnect so the BMS has a disconnect at 100% SOC.
You've really never liked the low balancing current BMS's have, but I think it might be a bit of over-optimization. If normal charging specifications are followed, meaning we charge to 3.55V and hold there for an hour or two each cycle, and assuming the BMS is sane and starts balancing above 3.50V without any silly restrictions like only doing it while charging current is present, then the whole system works perfectly well and no active balancer is needed.
But you would need to keep the battery at 3.55V without anything connected but a power supply to keep it stable on 3.55V per cell. As soon as it is still charging or discharging a tiny bit, the balancer will not work. These undersized balancers come from Li-ion where you can start balancing at any voltage.
We will see the first of these kind of BMS with an active balancer soon...
@@OffGridGarageAustralia Right, but generally speaking if the charge gets to that point, holding it there for a while isn't really a problem. Once the charge current tails off (30 minutes or less), then holding at 3.55V doesn't take any more energy than holding it at 3.375V would.
So its just a matter of at what point in the day the solar gets it there in terms of holding for 2 hours to give the BMS's balancers sufficient time. And if there is extra time after that the charger simply drops to the float voltage of 3.375V or 3.4V and continues holding there. You have the same load-support capability either way.
Oh boy.- I hope this BMS programers check this video! THX Andy for that great video! I wish the JK BMS had WIFI or an ETH connector + easy (independent) PC software which connects parallel AKKUS.
Yeah, a BMS with FW upgrades is already a winner. The software should not be a problem. I don't understand why none of these manufacturers or sellers get it right though.
thanks again for your explanation and evaluation.
Thanks Joe...
Hey Andy, von Seplos gibt es wohl eine Version 3.0 ihres BMS und in der Werbung steht was von aktiv Balancer! Würdest du das mal testen? 😂
I'm confused I guess also needing advise. I have (us model in usa)growatt 5000es with 6 signature solar lifepower 4 batteries. Batteries are all comm linked thru the rs485 port (one designated master)hooked to the 5000es rs485 comm port. Works flawlessly so far. My question is with the communication does the master battery talk with the other batteries(bms to bms I'm assuming)and decide which battery is to charge and which is not in order to keep them balanced as a bank and the bms within each battery to keep its 16 cells in balance?I cannot set any charge parameters or cut offs this way it is all set thru the inverter. Secondly if I wanted to build my own 16s 280ah battery using catl cells. Which bms would you use and would any you know of communicate with the growatt 5000es.I do not want to parallel the inverters. Just use as separate stand alone systems but each having communication with their battery packs.I plan on running all my 240 loads on one setup while using the other for my 120v loads. I don't like(trust)having all the inverters paralleled together and risk having total failure. I have 3)5000es and plan on getting a spare.Use 2 for 240v and a single for 120v then have the spare. Any thoughts. Sorry so long winded. Howdy from Texas!
Thanks for your question, Chris.
The communication does not go that far. It is more to pass error or alarm states on to the inverter, show an overall capacity and also the overall deviation across all cells combined. The master BMS will not regulate the charging of single batteries though or controls the balancers. That is still part of each individual BMS in your batteries.
Any BMS supporting the Growatt protocol via CAN or RS485 will work. For example the Gobelpower BMS as well as the Seplos BMS.
@@OffGridGarageAustralia so other than the soc that it relays to the inverter as well as it has preset cutoffs and charge limitations from the inverter your really not gaining anything you cant do yourself? Also how exactly can I determine what protocol the growatts are using for comms? They use rs485 ports but I'm assuming by protocol you mean programming/language they are speaking lol.
@@chrislewis9345 In the battery, you select the brand of the inverter, or the battery protocol used by the inverter. I don't know all inverter selections, so maybe someone else can give solid fact, but you'd probably select Growatt, or maybe the Growatts employ the Pylon battery protocol.
It's a protocol not a programming language or API. It how the settings are conveyed from the battery and understood by the charger or inverter.
A protocol defines the exchange process as far as the order that each device will communicate (request -> answer, field order [ie status might state the voltage followed by current draw etc in one BMS while another may send current draw followed by voltage etc]), the format used to express that exchange (is a number an integer, a floating point number, a string of characters, how many digits of a number represent the data before and after a decimal point etc) and other metadata like delay time required between each conversation, delay time expected between request and answer and so on.
Thank you sir for the info
Next project for Andy - Frankenstein BMS to combine the good features from all BMS’s & an active balancer.
If I just could...
Hi Andy, i have watched just about all your videos and really like your way of showing the fantastic world of solar power, i have a question if i may ask, '' do closed loop communication BMS's connected to the inverter charge multiple battery packs based on the SOC in order to charge the packs equaly ?''
I think you have sent this email too, right? I just answer it here again for everyone.
No, they only charge per voltage. Parallel batteries balance themselves without problem. One banks has a bit higher voltage and will take less charge until the others have caught up.
There is no communication between the BMS and inverter necessary for any of that to happen.
@@OffGridGarageAustralia Hi Andy, im so sorry for doubling my question, yes i did see your email and thank you very much for replying, it was important for me to know this, i just installed 20 kw of solar panels, 24 kw (4 x 6000W) hybrid inverters and 40 kw of Lifepo4 battery's, im learning every day new things thanks to people like you 🙂
@@creamshop Not a problem at all 😉
have u tried batrium bms the do the 100% reset too. with 2 amps balancing per cell
Hello Andy, nice review! Thank you very much. Do you know if the 200A Gobel BMS is the same? Greetings from germany
Yes, exactly the same, just with 200A.
I love your videos, but I must correct you since i have been seeing 10 videoes where you say seplos stop charging and reset to 100% only at 3,65v. I have a seplos case and it reset to 100% at 3,50v, because i have set the total battery voltage to 56v. When it hits that it resets to 100%, also when I have set the cell ov protection to 3,6v.
The seplos reset to 100 percent when it hit an overvoldtage protection for the cells or total voltage you have set yourself.
I use it with the 10A limiter but my deye 12kv 3 phaee inveter wants to take things easy and only charge with 5A until the battery does not absorb anymore. After that, the battery goes down to around 3,4v. Best inverter ever, 17A on each phase if all phases is loaded 100% or 27A on one phase if the total load is under 12000w forever. And it can do 200% for a small amount of time on both the 17A and 27A. And it can do all sorts of smart things, like use the generator input as smart load that turns on when the battery reaches a preset % and off again at a preset %, I use for charging a tesla. It works with alot of batteries like seplos. It have the shunt build in so it can run with no connection to the bms in % mode. It can take 18kw solarcells and run another inverter like micro inverters on the generator input, even in off grid mode
I run the seplos with the neey 4.0. I got some very good eve280 cells there dont drift much. I could run it without the neey and I did for one month because I turned it off by mistake, and the call was only 25mv from each other when the battery was charged to 100%. The settings in the seplos bms its set to that the balancing starts at 30mv and stop again at 20mv, so that is exactly what it was set to do. I have the neey set to 8mv because I like the the quick balancing the neey offer, then the seplus work as a backup balancing if the neey stop working or more then one cell runs off.
I have not updated the bms or anything, and it does just work. The system runs as an off grid system now, until the permit to export to the grid comes.
Hey Andy, my jk bms shows 100% without the need to hit the upper limit...
Yes, mine too. I still don't know the trigger but I rarely look at its SOC.
I like when the white board comes out
Says more than 1000 words, right? 😁
What BMS would you recommend for a simple 4s battery pack then?
The JK-BMS all the way! Link is here: off-grid-garage.com/battery-management-systems-bms/
The only gotcha with the 4S JKBMS is it seems to have a significant current draw. Did a test of 100AHr 12V 4S LiFePO4 with full charge and left it for 3 months and it lost 40% of capacity - fortunately there is an on/off switch that you use to easily switch on/off BMS. I have a suspicion that the current draw is due to the inefficient DC DC converter that produces >21V as this was the voltage required to drive the BMS in 16S versions.
JBD bms tells you it's 100% according to what you program into it. Maybe you need to revisit it Andy? The last one you looked at was really small and soo old. My batteries reset at 3.6 volts, where I need to program them because of the Canadian climate, instead of the 3.45 you use. We get much colder here then Australia and 3.45 is too low to charge completely at anything below 8 degrees C. Program the new JBD bms, at 2.8 to 3.45 (as an example). Test capacity by going from full (3.45 ) to empty (2.8). The JBD will now read and communicate the AH stored within those parameters as 0 to 100% .
There is no need for the BMS to tell the Victron what to do if the Victron system is programed correctly. The inverter should always turn off before the battery, this is easily achieved with voltage parameters because of the LiFePO4 discharge curve. The inverter throttles down any loads at shut down, preventing spikes to the battery.
Charging MOSFETS turning off is essential. Relying on an outside device to stop charging is added complication that invites disaster if something goes wrong with the communication.
I'm in contact with JBD and see what we can organise...
That is the whole purpose of a smart BMS to tell your inverter what to do without having to program your inverter (or Victron) in this regards.
I don't think it is essential to turn off MOSFETs at all. If the charger follows the BMS request it will just stop charging at the set numbers. If not, the BMS is there to catch the battery and disconnect. It's a 2-step system.
@@OffGridGarageAustralia I see how in your situation this will work fine, and might be the goal. In other situations, say for instance if life safety equipment is involved (like we have on ships), this setup could cause major problems.
Lets say your battery shelf with 3 batteries is your main system in this example, and your backup system is the Frankenstein battery. Your main system has 3 BMS's that communicate to turn off the inverter. One of your three batteries, has a failure and turns off the inverter. You switch to the backup Frankenstein battery, because someone's life depends on having the inverter running for their medical equipment. The inverter doesn't turn on. You quickly isolate the defective battery with it's circuit breaker and switch back from your back up Frankenstein battery to your main battery. Your inverter still doesn't work, it's being told not to by the BMS of the problem battery still. Somebody dies before you can reprogram things.
The same problem happens if you run the main battery bank to empty then try to switch to a backup battery. The empty batteries will tell the inverter to stay off even though a backup power source is available.
By using the voltage profiles of the LiFePO4 as the triggers for the inverter, you ensure that a backup system will turn the inverter back on in an emergency.
JBD makes a BMS capable of 750 amps peak, so you're not going to blow any MOSFETs like you have with other BMSs.
If we are talking about massive storage installations, Victron isn't involved and neither are any of the BMS's you and I are using. A much more complicated computer control system is used for these.
@@boatelectricaldiy Thanks for your comment.
OK, so one single BMS will not turn off the inverter, it will turn off the faulty battery bank and the inverter keeps running on the remaining batteries (we have tested this with the Seplos batteries).
Also turning off the breaker or battery switch for one battery banks should also isolate the BMS. If not the order they are wired is wrong (or bad design).
Switching from the main battery to a backup battery should always start the inverter, otherwise it is useless as a backup as you explained. Life saved, maybe?!
@@OffGridGarageAustralia it seems my lack of interest in other BMS's has caught up to me here. The price and size of these BMS's really put me off, and I I haven't had lots of time on my hands so it seems I've missed a vital bit of information that you have presented in your videos. I guess I should go back and pay better attention while watching. I still don't see my customers being interested in this, but you've peaked my interest yet again. Thank you, keep up the good work!
@@OffGridGarageAustralia I have been able to revisit this video. My company has expanded from just marine electrical installations to remote offgrid home installations now too. I was able to locate the manufacturer of the BMS that these company's use, and have ordered a few for testing. Thank you for the great videos, it's rare that I find something on TH-cam that makes me reconsider my business model.
Hello everyone. Can i use this BMS with Li-Ion 18650/21700 cells? Configuration 15S with max voltage of 63[V].
Hey Andy!
Given your current knowledge, would you still recommend JKBMS or not? Do you know which hardware and software version of the JKBMS you are using in your system?
The JK-BMS is a very different device in comparison to these larger BMS.
As a rule of thumb: use teh JK-BMS for 12V/24V batteries or if you have only one battery. If you have several larger batteries in parallel and want more safety and also communication, Seplos, Gobelpower and co. are better for that purpose.
Mine is a very old JK-BMS, one without a power switch or display option.
@@OffGridGarageAustralia Thank you very much for this information. I already have several JK BMS versions in use and I am of the opinion that the current hardware/software version 11.x finally calculates the SOC value more sensibly. Apparently the problems of the older models have improved significantly here. I would also like to test the Gobel Power BMS, but since it doesn't have Bluetooth, I'm still considering it. I am currently in the planning phase for a 43kwh storage system that will consist of 3 individual batteries. I'm really undecided on which BMS to use ;-)
@@cyborgxxl Thanks for sharing about the improvements with the JK.
With three banks you could still use the JKs but won't have any comms or elevated safety features like with the other ones.
The Pace, Seplos and Gobel are all very close together and at the moment I would go with the Seplos just due to the fact, that it works with my Victron system and the connected DC load. This is a major drawback for all the other BMS.
And we also have found a solution to overcome some of the problems with the Seplos BMS now. Well, one of my viewers has found a pretty good workaround for the Seplos...
@@OffGridGarageAustralia Ok, thanks for the detailed answer. I think maybe I'll go for the Gobel or Seplos BMS after all. I only use Victron components. Greetings from Austria!
great thoughts 👍
Newer versions of the PaceBMS resets the SoC at the "full pack charge" voltage, just an fyi :)
I totally agree. @off-grid Garage: you just have to set the parameters the right way 😉
@@Der_Hannes Yep, I got my OVA to 3.55, OVP to 3.6, and fullpack voltage to 55.2 and current to 1A, and the view is to slowly increase the FPV to 56 over the next week so the built in balancer does its job.
@@sarahjrandomnumbers my FPV sits on 55.2V and FPC at 1A cause i have added a NEEEEEEEEY balancer wich of course is doing a great job. Anyway, the SOC resets to 100% just if both parameters are reached.
The only downside i noticed is that the charging mosfets are turned off after reaching that "Fully-Voltage". That should only happen at a protective situation
@@Der_Hannes I haven't noticed the charge fets turn off when it hits FPV, but i'm charging up now so I'll see if I get the same behaviour. :)
@@Der_Hannes Just fully charged, and it didn't turn off the charge fets, but the remaining current and presumably SoC reset once it hit the FPV and FPC levels.
Recorded it just in case, i would post the link here but youtube is blocking it, it's on my channel anyway. That also includes the view of the BMS settings I have at the moment for those that want to compare.
No cell/pack alarms or protect, and the balancer stayed on until it went under 500ma, as expected. Currently at a deviation of 13mv, compared to 270mv when I first got it, although I did open the pack up and charge every cell individually to 3.6v, then let the internal balancer do its job.
Although, thinking about it, it's not really a balancer, all it's doing is adding 43ohm to the cell, which makes it look like the internal cell resistance is much higher, meaning it'll charge slower, which makes a whole lot more sense than it draining power from the cell to balance it. But that's a different conversation :).
It is the Seplos BMS 3.0!
Good guess...
😂
You said that it has a lot of problems, how can one want to buy it knowing that it has such a problem.
You never tried the daisy chaining.
Why not also include OrionBMS or REC BMS? They do all that and so much more. It’s a shame to say Gobel is the only one to do things like that.
When I first mentioned testing the REC BMS, Andy said it was too expensive. Maybe now he's going up to higher priced models he might reconsider 😁
I said, it's the only one so far...
Yeah, why not test the Orion and Rec as well?
Very interesting video
I have 48v byd lithium batteries
24v each which are connected in series to make up to 48v
I misplaced the bic card which means I can't use the bms in the Battery pack. I want to know if there is a way to build my own bms in both packs. Thank you
If you have access to the battery cells, I don't see a problem replacing the existing BMS with another 16s BMS.
@OffGridGarageAustralia I don't have access to the cells the batteries arr still new and never been used before.i have this connector hanging from the cells to the bic card then to the bms am not sure if that can be used instead.if not I have to take the Battery apart
The best bms is the first bms that has a victron smartshunt build in.. still waiting for it tho to be made.
i swaer your videos are like the movie "Inception" videos inside of another videos inside of a short editing video 😂😂don't blink or you will mis the whole video Love it.
Yes, I like to add all kind of information so people have to watch the full video. Otherwise they are confused.
"this totally sucks! DONKEY BALLS!!!! CLASSIC 🤣🤣🤣🤣🤣
Im acting as a human BMS by discharging and charging each cell when they become unbalanced, using a car light and a power supply. I can balance the cells in no time and perfectly. The cells will most likely become unbalanced quicker then the time it took to balance them when i connect it to my Solar system. I have a Hellcat 4s Balancer but thats only for top balancing. Im going to have to buy something i can program. Maybe the BatGo ?
Have you tried Stuart Pitterways DIY BMS yet?
No
Can it work in parallel with different capacity like we tested the 2 seplos ?
I have only one, so don't know.
What would you think about making the JK-BMS to be able to do the same. Like reducing charge current when it comes close to overvoltage or reduce charge and discharge current when it reaches a critical temperature. Let me know if you are interested.
The JK does not have this functionality and no hardware limiter. It also lacks the ability to communicate with parallel batteries.
@@OffGridGarageAustralia But i think i could make it to communicate with the inverter and tell it to reduce the current in these cases, by letting it speak the Pylontech protokoll. I guess your Victron system can understand Pylontech as well.
@@ackmaniac5101 Yeah, you can let it communicate to the Victron system and other inverters. I made a video about it a while back. The question is though, how do you handle several battery banks in parallel and how do they report combined back to your system?
Boo! We want a test of Batrium and REC BMS 🤣
Anyway, nice video.
Hahaha, just wait for it...
Can anyone tell me how to connect a JK_BMS to a laptop and what program to use please?
7:07 Donkey Balls!! LOL
Im still a huge fan of the electrodacus sbms0, its tiny but powerful and leaves all the heavy lifting to the devices that are meant to do it.. I really dont get why this is the only bms that works like that.
Yeah, but only 12V/24V systems...
New model out with Bluetooth?
Thanks!
Thank you very much for your support, Liviu!
I'm looking forward to your Gobel battery review with the active balancer. This could suit me nicely.
It's on its way...
They're galahs ya galah lolol just ask Alf from the telly ;) Didn't realize you're in Oz too until this video!
Sulphur Crested White Cockatoos
It is the new Daly BMS with the parallel board, I am most certain....
Good guess...
@@OffGridGarageAustralia I bought two myself and now I can follow along at home 😊
@@TrevorFraserAU You are committed, my friend. They should work really well. You still need an active balancer though...
Not true my JK BMS shows 100% Charge state for 16S at 54.97V and 3.439V on avg. The JK BMS can also nicely connected via Eib Kabel and USB Port to a RPI4.
Oooh! Another BMS SOC problem? Where have we last had it? Oh, yes! Outlander :D
Ps. Anything coming on the Teslander soon? Interesting updates? Degradation of the battery, perhaps? 😎
We got our Model 3 just maybe a week before Andy. Our battery has not lost anything at all after the first year. I suspect Andy’s will be the same.
@@FutureSystem738 that's quite impossible 😉
I'm 3% down after 3.5 yrs
@@OffGridGarageAustralia Great news Andy. It’s NOT something I lose sleep over like some, and nor do I do deep dives with software programs and OBD readers etc- I suspect we’ve lost something more like 5 to 6% but it was ALL in the first year. Since then the “full range” hasn’t altered more than about 1km, and always goes up again for a while after a long trip. (Our car has literally never in it’s life been above 95% as I like to maintain some decent regen, and only below 20% maybe three times.)
This is in line with many overseas users with very high kilometres- initial degradation over the first year (ie: reduced projected range) and then it levels off very dramatically.
I do NOT miss the Outlander’s battery issues and bloody useless BMS - though it was otherwise a pretty good car.
@@OffGridGarageAustralia how much of that is true though. DBCAM is needed 😂
Do they have a 12-volt version and 300 amp?
It's not a consumer grade BMS like DALY, JK or JBD...
@@OffGridGarageAustralia what would you classify as consumer-grade.?
Mensch Andy deine Video sind genial
Bin auch deiner Meinung.
Thank you! 😊
JBD BMS can give you 100% SoC at level selected
Which JBD ist that and whip app/software do you use?
my reply got deleted?! I am using both xiaoxiang and overkillsolar apps. Setting you shoul check is single full voltage or cell full voltage or something similar depending of version of app used
mein letzter versuch zu antworten, youtube blendet scheints meine antworten aus, hab ein video eingestellt mit seplos einstellungen
gruß Klaus
You can't post any links. We would be drowning in spam...
6:10 ich widersprech zwar ungern aber das stimmt nicht was du sagst , bei meinen 3Seplos hab ich derzeit jeden tag 100% mit 56V, es dauert ~1,5h von 95Soc auf 100%Soc, alles gut
What are your settings then and what Seplos BMS with what FW are you using? It all depends on the settings...
Hi Andy,
Great vid, I went to look at your GobelPower BMS link to buy one, but they're all for 16S, is that corrent?
I didn't see a 4S12V system, so I guess my option is the JK one you reviewed about a year ago, this one:
"The JK-BMS 4S-8S Edition - World Premiere. The best BMS now for your 12V LiFePO4 battery system"
So I'll order a B2A8S20P unit, with your link. For that unit, do you know if that 4.3" screen that they sell is any better for
the amount of information that it shows versus the smaller 2.3" one with the on/off button you showed in the video?
These sort of BMS are all 16S for 48V battery system and communication between them and an inverter.
Yeah, for a 12V or 24V battery, use the JK-BMS. that is a great device and will work perfectly with its active balancer. The screens are really optional as you can always connect via the app. There is just basic information on either screen. They are great if you install your battery in an RV and also other persons can read the information of SOC. And the on/off button... it again depends how you use your battery but the BMS should be staying connected to it all the time anyway.
And thanks a lot for your support and using my links
what part number is he reviewing?
I'm quite surprised that in this day & age, BMS manufacturers are still making those "common" mistakes.
And they often are all making the same mistakes, which I find very worrying.
The fact that their firmware on the BMS often can not be upgraded is a very concerning one as any bugs discovered, and you have found many, can not be fixed for their existing customers.
I think the best thing we all can do is cease buying such BMSses and focus as much as possible, for the Gobelpower solutions. (and JBD haha)
Like we had with the NEEEEY balancer last year, where it took them 4 (hardware) attempts to get it right. How good would a FW upgrade function bee in these devices.
I don't understand either why none of these develop a 'perfect' BMS and listen to the community, be on the ball with new FW upgrades and support. They would sell like fresh baked, hot cookies.
@@OffGridGarageAustralia I totally agree with you, a good BMS would be a true bomberfest, like the GobelPower seems to be.
Thanks to you & your channel when I was about to purchase a Daly BMS, it was one of the first videos I found you discussing it in informative details.
That was a very cold shower when I watched that video and opened my eyes to dig deeper & further and I got bitten by the LiFePO4 & BMS "bug" due to you dear Andy & your wonderful channel (& informative website!).
Like# 10!
Pylontech BMS?
Good guess...
@@OffGridGarageAustralia ...though l am wrong... (awaits next installment)
Hello!! From uk
Hello from Down Under👋
👍👍
Hahahaha donkey balls had me rolling, classic
🐸🐸🐸
After discussing the translation issues with several companies in China, I received an email for the "Central department of export control" of the Chinese Government to "Cease and Desist" suggesting that individual manufacturers to use independent translators. "All translations are completed correctly by the central translation offices of the export control offices by appropriately trained professional translaters"
I stopped my correspondence immediately as I feared for the safety of my correspondents.
You can draw you own conclusions.....
Hahaha, great... I guess I'm appropriately trained in many things as well...
Keep your cells balanced and rely on a real battery monitor like Victron to do SoC measurements.
BMS's with a bunch of cheap non-zero temp compensated paralleled chip resistors for current shunt and non-auto offset zeroing sense op amplifiers for shunt resistor voltage sensing is not a good way to make an accurate SoC indicator.
If it came in 24 volts i would buy it but it doesnt 😢
No, it's not a consumer grade BMS like DALY, JK or JBD
@@OffGridGarageAustralia consumer grade bms just do not cut the mustard. Ive got 50 kwh batteries that unfortunately are at 24 volts due to everything thats powered by them being 24 volt, replacing inverters and everything else for 48 volt would cost far to much