I am glad that Andy is finally seeing the value of HA. I have been running HA for several years now, it is my "PMS" (Powerplant Management System). It gathers all of the data from my Inverter, BMSs, etc. and provides me with a LOCAL human interface to monitor and control everything. I built several boards to monitor & control JK & JBD BMSs, MPP & PowMr Inverters, as well as PZEM Power monitors (Shunts and Current Transformers), and automatically populate all the data points into Home Assistant. The net result is that HA is the "Brains" of my system, not the BMS.
I hope the system will still safely work if HA does not for whatever reason? I'm using HA only for monitoring atm. Even if the BMS fails, the MPPTs will only charge to a certain voltage...
Thanks for the update video. Love the build process!! Question/Observation: I don't think the "Charging Enabled" toggle is a good option as it implies charge is ON/OFF while it is not. It is basically a toggle that reduces the CCL (removes the battery from the CCL calculation in Victron). Maybe call this CCL include or exclude. Thanks again Andy!
Yes, that's what I said and explained, the charging enable toggle has only an effect on the CCL. In Home Assistant, you can call it whatever you like, this is just the default name of the toggle. I know what it does, so it's OK for me, but thanks for your advice and explanation for others again.
It was very nice to see how you made the home assistant work. Maybe you could show how you did it on another video a good day. I have trouble to make connect to the victron system😮
Yeah, that goes a bit away from the solar and battery topic though and only a few people would be interested i such a video. I'll have it on my list...
I have been using Home Assistant for 2 years to monitor and control all aspects of my home powerplant. Having no Victron components, it was the best choice for me. I have designed a few "Dave" boards that interface my Inverter, BMSs, and Shunts to HA.
@@OffGridGarageAustralia What if the BMS Fails? What if the RasPi fails? What if a circuit breaker fails? After 2 years of daily use, I have very strong confidence in HA, and I have no dependency on the internet to update a web site in some foreign country (that I have no control over) and I have full control of the data display and control logic. I have WAY more confidence in HA's development approach than I do in JK's, based on several years of direct experience with both.
Excellent I am glad that you finaly use Home Assistant to display / control your system. So much nicer so much powerful 😊 Would also be nice if you could recap how data are exchange in your complete system ( ve bus, cerbo gx, can ...)
I want to point out that all the software that makes the peter board possible is open source and done by a lot of people in their free time. Esphome is the key word if you want to look it up. Some other guy did all the work implementing the comunication from the bms ti esphome. He took all this work from others, added some charing logic, then makes it cloesed source and sells it for an insane amount of money. Keep the source open and simply sell your designed hardware with your firmware pre installed, so people which want a plug and play version can but it from you, but taking an open source Projekt modify it and not letting the whole comunity profit from it but just yourself is just wrong and you know it.
The data come from the Peter boards from the single BMS (2x JK and 1x JBD). Some Info also comes from Victron. All data is coming together in Home Assistant.
I like the Dashboard Andy, big improvement. At 15:45, I as you have a good milliohm tester and a baseline on your you pack with the Aluminum busbars, I might suggest testing an aluminum anti-oxide compound to see if it adds any resistance and prevents corrosion and resistance build-up. Many years ago we used a lot of aluminum conductors in the oil and gas industry at a remote location and found 'Penetrox A' (DXE-P8A) an industrial Aluminum Conductor Anti-Oxidant indispensable, and that it effectively prevented aluminum to aluminum and aluminum to copper connection corrosion if used a liberally. The stuff is relatively cheap and a bit messy, but worked well on connecting aluminum to copper buss bars and even using copper compression lugs not rated for aluminum, on aluminum wire.
Thanks Jack. I have tested three different paste, greases and compounds in the past. They all seem to work well for a while but eventually have made the situation worse. I had the best experience with NO grease whatsoever.
I also have a JK BMS 60A / 0.6A Balance and its just working fine for a 16S 48v system w 9.6kWh. Waiting for my Peterboard and will also start Home Assistent later this year.
@@nodelayfordays8083 That must be 300/320 Ah cells. Looks ok to me. When all cells are almost the same voltage, balancing with 0.6A is enough I guess. DIY Battery banks I like more than ready made Storage products as they are a lot cheaper. And I like JK BMS. So I have ordered the so called Peterboard from Uksa 007 to be able to connect to my Deye Inverter over Canbus .
Awesome, I really liked it, it will be very inspiring for many people. Automatic Plug & Play Home Assistant integration should be part of all smart solar devices nowadays. I hope solar device manufacturers will wake up and start putting this in their devices. It would be great to follow up with a video showing the process of integrating the whole Victron System into Home Assistant.
I have all Victron Gear in HA for monitoring purposes. I still like the VRM better, so will only have a small dashboard in HA with some selected numbers I'm interested in. For now at least😆 HA is a can of worms!
Thanks Andy - love your work as always (though way beyond my level of expertise). Stay cool, and stay charged. I’m very happy to have 2A balancing with my JK BMS’s, and would actually like even more - I occasionally need to extend the absorption charge time to fully balance if it’s been a while since they’ve been fully charged.
Because they are grade B cells not matched what you often get from alibaba and aliexpress plus a lot of vendors. Or people think they safe money when buying garde b labeled cells from reliable sellers when the save 20 Euro on a 130Euro cell and think they just get a bit less capacity....
I agree with him its better to have more balancing current then you actually need right now as the cells aging over time and not evenly so the delta the balancer has to cope is getting bigger over time. So 0.6A might be ok now but in 5 years not anymore. And if i am building a system its designed to last the battery life.
Forgot to add the more cells you have in series the more the whole cellpack will deviate over time, means a 48V pack needs always a BMS and a capable balancer while if you have high quality cells in 4S a battery monitor with an external relay like Victron BMV712 that does LVC and HVC is more then sufficient. Know over 40 installs (half i done myself) in cars older then 7 years that have no BMS with 4x40AH till 4x100AH Winston cells as only battery, no BMS. Took some appart and deviation after 7 years were below 5mV. And they often see 5C peak use from the engine starter surge. Forget doing that with 24 or 48V as the production differences of each cell add up when you put them in series. And that added differences which grows over time need to be covered by the balancer. And if you use that in eg boats where the bank capacity is often twice or more of the solar total wattage these banks are massivly oversized so getting seldomly to 100%SOC. Means when they do the active balancer needs to be bigger to balance the bigger deviation as a result of that.
@@chrisr819 Yes, the more cells the bigger the problem, as only two cells can balance at any one time. I have a 2 x 18s system, and often go many part charge cycles before filling and getting a chance to balance. I usually need every bit of the 2A balance current if it’s been a long time since the last full charge and balance.
Andy, have you had a problem with firmware update failing and then the JK BMS locking up? I tried to update the BMS, got a failed message and now I can't see the BMS in the app or computer.
Hello Can you tell me if I can use Peltier cooling of the battery enclosure, as is the case with you? In summer, in Romania, in my garage, temperatures reach 40-42°C and the optimum is 25°C Thank you very much
No, that will not work. Peltier does 10° max and produces a awful amount of heat on the other side which you need to get rid of. I have the same temperatures and the cells will be fine for these few days. They can handle close to 60°C
How the Inverter decrease the Voltage ? I think it would be good to understand how & why. Amps goes inside, and to be removed need to be used or converted in heat, but the voltage reduce process......@@OffGridGarageAustralia
@@christophec252 The board controls the charging voltage CVL, so the inverter only charges to this set voltage. Like with any inverter or MPPT. With these boards, you have control over that voltage from your mobile.
Sounds like a classic case of needing VERTICAL, or nearly vertical solar panels- facing South in your case. Here in sunny Brisbane …. Should I say cloudy, showery, stinking hot and sticky Brisbane, we’re getting almost constant showers between patches of sun, coupled with extreme humidity. However my solar is still keeping my battery fairly full and running the aircon. (Couldn’t live without aircon- really disgusting weather to be honest. Despite being born and bred here, I survive our summers but absolutely LOVE our Winters. This Summer has been especially bad.)
Vertical is deceptive to winds or you need to put them lower, where shadow make them useless. I created a flimsy 45 degree setup, with plastic panels which can sway in the wind, they are not zero. But almost
There are systems to melt snow off solar panels, and ground arrays are easier to clear manually. They do also make special very long handled snow blades for clearing roof panels. But it's a pain any way you look at it.
I had a lot less snow, but it is still quite annoying when the sun is shining. How about just using an electric blanket with an extension cord, wrapped in something safe and waterproof?
Welcome to the unlimited possibilities of HA 🐸 My Battery Dashboard looks nearly the same. An i have a second dashboard showing all single cell voltages. With a few clicks i get pretty graphs with all cell voltages and can see how diffrent cells can behave during charging. The Deye Inverters have two limits for charge current. One is set via CanBus from the BMS port. This should never be touched manually for saftey reasons. The other limit can be set via the display or via ModBus from HA - manually or by automations. HA is really a great peace 👍
Yeah, HA seems pretty slick for this sort of monitoring. But would I trust it to run the whole battery shelf including charging and inverter control? What if it fails and does not run any more? Yes, we have backups, but how to recover the whole system after a crash... I may need to test this out at some stage...
@@OffGridGarageAustralia I dont use HA for safety relevant controls. But it is perfect to manage several things. For example: If sunshine is forecastet for the whole day i reduce charging current, or i heat up the boiler first. Or i can switch off one or more inverters during the night to save idle power. Or i can overheat the house a bit if the batteries are full an there is still solar power available. Or a additional colling fan is controlled by the temperature sensor of the inverters DC board. If HA fails nothing happens. As the safety relevant limits are controlled by BSC and by the BMSes. The minor settings HA is doing in the inverter can still be changed using the buttons and the touch Display on the Deye.
I was just wondering about this. Im making a 15kw system with 16 xells in series Eve 304ah and got the JK BD6A17S8P BMS with 0.6 amp active balancing amd was wondering if that would eould be enough for this build.
Home assistant seems nice :) Are the middle shelf also the 280K ones, I know the top shelf isnt a K type electrolyte, might not be the bus bars but I see a bus bar swap soon as a test - well i hope you try and prove the bus bar theory.
I like your dashboard. Today I installed a second Multiplus II in parallel to my original one and connected two additional 280 Ah batteries. This gives me now 60kWh. Enough to charge the car from the battery in summer. Unfortunately it's not summer right now.😢
Have you considered polishing the contact points in the AL buss bar setup. That should bring connection resistance back to par. Now you could know for sure why the difference. A T Burke
Yes, I have polished aluminium terminals in the past😆 The surface will corrode within milli seconds and yes, I have tested several aluminium pastes and greases. Best result for me is still nothing in between.
Thank you, Sir. I think I will stay with copper then. No matter what size bar is used to reduce resistance, it still has to make it through the connection. A T Burke@@OffGridGarageAustralia
Hello Andy wonderfull video. Homeassistant seem very simple and power full i wonder if a comparaison with venus os would be usefull. Seem to me Homeassistant is more power full and easy to use but your opinion could be great. About connexion i discover there are a lot of way to connect JK-BMS to Home assistant. I say that because Peter board is very expensive (almost bms price) naturally the work done is great and have to be paid but the gap is very hudge (hardware arround 10 euros and price more than 100 euro) The big difference i have seen is the peter board manage the charging and not only connect the devices. Thanks for all.
Is there a rule of thumb what current the integrated Balance should be capable of based on the battery size ? I can remember that many had taken the 2A JK BMS for the usual 14,3 kWh EVE LF280K batteries, but what about a 3,6 kWh LFP 16S battery (74 Ah) ? 16 x 74 Ah x 3,2 V = 3789 Wh or 3,8 kWh which would mean 27% of the big 14,3 kWh. And these 27% would equal with a 2A BMS just 0,54 A A 0,6 A balancer should have even more reserve, but the main question is still: How should the balancer be layed out to get the battery balanced (battery capacity) and when to get a 4A or a 2A or 1A or 0,6 A on
Dear Off-Grid-Garage Community! I‘m a big fan of this channel for weeks and I want to use the community to ask some questions to fine the best setup I can get from Victron Multiplus 2 + MPPT + Cerbo GX + DIY Battery + JBD BMS. Can I raise this questions in this video? Br
I like the Setup you have done. I am planning on installing 2x Orient Power 230Ah 51.2V 11.8KWh batteries in parallel with my existing Simpliphi power 75Ah 22.8 kWh Battery Bank. The only problem I will have is that the Simplphi Battery does not have an integral BMS but rather is controlled by a Schneider Electric Connect BMS for SOC control. Is there any way I can control the three battery banks' BMSs? I can see that the two Orient Power batteries' BMSs can communicate with each other by connecting the communication cables. I just don't know how they could communicate with the Schneider Electric BMS. I can set the 2x orient Power batteries BMS settings locally on their perspective BMS and then have the Schneider Electric BMS control SOC control to the combined battery bank.
Hi Andy ✌️ Is there a way to get the new options from the new JK BMS to the older ones? And at my older JK, I see all the time some heating current, but I have no current option an nothing connected. Also the heating is switched off in the settings. Someone had any ideas? Best regards from Jägermeistertown 😊
To confirm if it's the bus bars, Flip the JBD and JK BMS around next and see if the results are the same or different. Then we'd all know for sure. I'm at least curious about it.
one thing to note is the jk bms reports full pack voltage based on its individual cells voltage. if the calibration is off between two jk bmses connected in parallel the overall pack voltage is obviously the same yet the the two bmses report slightly different pack voltage number.
Are you sure about that? Why would they add all the cell voltages together for the Overall pach voltake instand of Measuaring the pack voltage directly?
Hi Andi, I am a little bit concerned: In the past you had a 2 layer of protection A) The Solar charge controller B) The BMS Both systems worked completely independent, if one fails, the other would prevent the battery from overcharging (witch can result in fire) But now, if the BMS fails, there is a chance that the system will continue charging up to very thing goes to fire. Maybe you can make a video on this safety issue
What fire are you talking about? If the BMS fails... if the circuit breaker in your house fails. That's not good in both cases... In any way, the Peter would still charge to 55.2V. So, no fire?
@@OffGridGarageAustralia You might got me wrong. In the standard setup (solar charge controller not connected to BMS) you do have a dual layer of protection. Even the SCC fails, the BMS will shut down, or if the BMS fails the SCC will have a maximum (bulk) voltage he will charge to. No issue, no potential hazard. If now the SCC lisens to the BMS, and this (e.g. measures a worng voltage due to component failure) the charging will go on upto the damage of the battery (and maybe, maybe not - a fire hazzard) That's just my point, by connecting, you remove one layer of protection.
@@peetbraun9439 We still have the BMS and the SCC. We only add an option to modify the coms in between. The BMS will still disconnect if cell voltage goes over 3.65V even the board will go crazy and tells the SCC to charge to 5V. Same, the other way around if the 3.65V protection fails in the BMS, the SCC will still charge to the set voltage the board tells it to charge. Nothing has changed and no safety concerns.
I like the high level of Home Assistant integration. That really is the ultimate in convenience! You might want to consider some security isolation, you don't want a hack of your main network to be able to access your equipment network. The basic way to do this is to have the machine Home Assistant is running on "straddle" the two networks. Your home network on one side and your equipment network on the other side. This would require having two ethernet ports on the machine H.A. is running on though which might not be convenient for everyone. Even on an RPi4 you can always connect up a USB ethernet dongle to get a second ethernet port. You then put the firewall on the H.A. machine itself. Basically allow the equipment network access to the home network and internet, but not allow the home network access to the equipment network. A second way is to isolate the H.A. machine and the equipment network with a router and then drill a hole through the router (figuratively speaking) so your phone app can access the H.A. web server when it is sitting on your home network, but nothing else (requiring that the phone be in bluetooth range or sitting on the equipment network to access devices directly). Its a project though. Most people aren't conversant enough with LAN routing and IP spaces to do it without a ton of effort. I do think you need to begin thinking about the security of your power systems network integration, though. -- I think you can safely start balancing once the cell voltage is high enough, even if a pack still has a lot of current. The energy the balancer is moving in the battery is effective regardless of the state of charge, as long as the balancer makes the "correct" choice. That's the basic problem with trying to balance at a lower voltage... the balancer winds up making the wrong choice and taking the cells in the wrong direction due to minor differences in cell chemistry. Once a cell is at a high enough voltage the balancer should start making the correct choice, even if there is current going in or out of the pack. And honestly, there is never "zero" current.. the battery bank will be charging or discharging, but hardly ever sitting near zero. Heh. Remember the time range for absorption time that I've been hawking for the last few years? "1-2 hours". You are getting closer 🙂 -Matt
Hello OGGA community. I have a question for you. I need to find a reliable bluetooth balancer for two 48v strings of four Lossigy 300ah LiFePo batteries. What are your recommendations?
@@OffGridGarageAustralia Sensor data is purged by default after 10 days. Only if an entity is build for long term statistic, it will store one summarized value every hour.
Using a Bluetooth connection instead of UART would have allowed the Peter board to truly cut off the charge from the BMS (the UART connection only allows reading of the BMS). But is it good to use a Bluetooth link in production? I'm testing it at home with two BMS monitored via Bluetooth by the same ESP32.
I would not trust Bluetooth for such a safety related connection and function. It is far to unreliable. I guess he made a similar decision at some stage. You've got this in your own project, I guess?
@@OffGridGarageAustralia Yes for my batteries I prefer the wired connection. But I had requests to use Bluetooth connection so now I offer two different YAML, Wire and Bluetooth. The Bluetooth connection being tested at home is stable (jk-bms code from syssi). Sleeper85
@@OffGridGarageAustralia Yes for my batteries I prefer the wired connection. But I had requests to use Bluetooth connection so now I have two different YAML, Wire and Bluetooth. The Bluetooth connection being tested at home is stable.
@@OpenSolarEnergy That is great progress. Thanks a lot for your work and sharing it here. Once it's ready, maybe I can show it on my channel as well...
If you put no-ox paste between terminal and alu busbar there cannot be an oxide layer as this connection is airtight and no-ox is sealing it permanently. This means no oxide layer can happen. I assume the total internal resistance is higher then the other two banks.
I have different experience and i am dealing with Lithium nearly 20 years now, the first 10 years in competition car stereo and the first lithium we got "illegally" from a scrapped military vehicle. The best result for a short period like 2-3years without paste if the terminal are metricously even and the busbars too...but thats rarly the case. And when you sand the battery terminal by hand you destroy that. Means the connection is not airtight anymore. The paste is highly conductive and make sure the whole surface is connecting and 2nd making the connection 100%airtight so that connection cannot corrode over time which otherwise happen if you eg use tinned busbars as contact corrosion can occure over time. I use either aluminium or silver coated busbars (which eg is used on NH fuses and high current environment installation like grid stations) with No-ox paste to seal connection.
The 'Peter' board seems pretty useful. Would love if it could update quicker than every 10(?) seconds. Also, what's the logic/reason for having "charge enable" and "charge current" on the slaves. Why not control it all from the master instead since it affect global settings, not individual slaves/BMS's?
If one or more of the BMS is off-line, or has gone into protection the current will automatically reduce by that slaves current setting, this is the primary purpose of the slave current settings, each slaves tells the master how much current it is able to accept.
If Home assist is logging the data, can you export 1 or more days & compare the sum total of amps in-out of each shelf? or daily averages. Each shelf will have a typical % share of the total Ah per day. And compare this with your estimates of cell capacities.
I ran across some 1.2A / 1.3A balancers on aliexpress that work all the time and are of the distributed / headless kind and most probably the future as they are very simple and economic. The "xS Active Balancer Equalizer Board" type sits between 2 cells in a pack and only cares about those 2 and moves energy between them if their voltage difference is big enough - keeping them on the same level. So for a 8 cell battery it's: *cell1-AB1-cell2-AB2-cell3-AB3-cell4-AB4-cell5-AB5-cell6-AB6-cell7-AB7-cell8* . In contrast to that does the MCU and active balancer on a JK-BMS for example jump around all the time and needs semiconductor switches that must not fail + needs those balancer wires (with resistance and all the associated problems).
We cannot use these 'always on' balancers for LiFePO4 as they cause an imbalance when in the flat part of the charge/discharge curve. The active balancers out there including the NEEY and JK balancers seems to be very robust and working extremely well.
@@OffGridGarageAustralia ok i see. But why would you need to set a float time at all? Im the victron system you can change the charging algorithmus to switch back to bulk once the voltage dropps a definded voltage e.g. 0.1V below the set float voltage. Once there is no solar the voltage of the pack would naturaly drop. Is that not precice enoth or way is the float time needed?
Andy, ALWAYS GREAT INFO. PKEASE HELP! Can you solve JKs BLUETOOTH "failed pair" mystery? JK keeps giving me new unique bluetooth passwords, that only work for 5-15 minutes... Then we have to start ALL over. Has not mattered the version I use... going on a month now.
Andy believe it or not it's possible to get the old faulty NEEY balancers eating out of the palm of your hand with an esp. Andy I live near to you.. we should probably catch up one day. I've been custom designing BMS and SCADA systems to use with victron gear for around 8 years now....
@OffGridGarageAustralia no. Remember version one that lots bought but they had no ability to purely top balance? There's be thousands out there nearly useless. An esp makes them into magic. Btw... I'm actually a telecommunications carrier - in my experience 0.6 amp BMS should be a ok. Cheers - matt
@@OffGridGarageAustralia further to my comment just above - regarding home assistant, no, you don't need it to be running for these to work. I have arduinos and esps of my own code fully controlling victron sccs (and more....) with no home assistant - in the latter case you just use esphome. I have had this gear in the Australian desert for 8 years - regular 45 degrees heat, only just replaced my last set of AGMs, have been using lfo now for about 4 years too in the same environment. also.. no shunt, but hyper accurate soc from victron scc's via coulomb counting - how'd you do that I hear them ask - you really should come visit me 😉
The voltage difference between the packs could be due to the calibration of the BMS. They should all be exactly the same voltage as they are all on the same busbar. The charging current could be different because of the cell chemistry.
Do you also have the SOC100%Volt setting? Mine definitely does not work... I've got this on one of the next videos. Seems like some BMS just don't do it.
Mine does not have the 100% SOC value. Ive been fiddling with the OVPR voltage and UVPR voltage. Deep cycling etc... seems to be closer, but it's still off by a fair bit.
Quick question, wasn't it always said that Lifepo cells don't need a float mode? If the cells are reduced from 55.2 to 53.6 volts, they are discharged and are no longer at 100% SOC. Wouldn't it be better if they are not actively discharged, but can simply rest without discharging? The voltage would still be falling slow. Kurze Frage, wurde nicht immer gesagt, dass Lifepo Zellen kein Floatmode benötigen? Wenn die Zellen von 55,2 auf 53,6Volt reduziert werden, werden diese doch entladen und sind nicht mehr bei 100% SOC. Wäre es nicht besser wenn diese nicht aktiv entladen werden, sondern nur einfach ruhen können ohne entladen? Die Volt Spannung würde trotzdem fallen, aber viel langsamer und natürlichen? 😊
Dazu müsste man die Batterien vom Busbar trennen. Das geht bei einem OffGrid setup aber nicht. Die WR müssen ja immer die volle Leistung können, auch wenn die PV mal z.B. wegen einer Wolke das nicht liefern kann. Die Batterie muss da immer als Puffer dienen können. To do this you would have to separate the batteries from the busbar. However, this is not possible with an off-grid setup. The inverters always have to be able to provide full power, even if the solar panels cannot deliver this because of a cloud, for example. The battery must always be able to serve as a buffer.
@@Juergen_Miessmer Wieso trennen??? Als Andy den Floatmode aktiviert, wird doch aktiv entladen bis die 53,6Volt erreicht werden. Hat nichts mit Offgrid zu tun sondern wie der Floatmode immer funktioniert. Die Frage war, ob einfaches Ruhen nicht besser ist. Why disconnect??? 'When Andy activates the Eloatmode, the battery is actively discharged until the 53.6 volts are reached. Has nothing to do with off-grid but how the Eloatmode always always works. The question was whether simple resting is not better.
@@derfreiemensch Weil an der Busbar nicht nur Laderegler und Batterien angeschlossen sind, sondern auch die Wechselrichter. Diese brauchen ja weiterhin Energie um das Inselnetz zu versorgen. Und diese Energie muss irgendwo herkommen. Ruhen ist technisch bei einer Inselanlage nicht möglich. Ohne float würden die Batterien sofort stark entladen und die PV nicht mehr genutzt werden. Because not only charge controllers and batteries are connected to the bus bar, but also the inverters. These still need energy to supply the load. And that energy has to come from somewhere. Resting is technically not possible with an offgrid system. Without floating, the batteries will be discharged much stronger an no more solar power is used.
@@Juergen_Miessmerwieso? Einfach im Absorption Mode bleiben, dann wird doch nichts getrennt oder abgeschalten. Und der mppt stellt weiter die Energie für die Verbraucher zur Verführung ohne die batterie zu entladen.
Hi Andy is it possible to remote login to the jk inverter bms when no victron stuff is being used.If so can you do a video for us stupid people that just use a basic system. Many thanks.
No, that's not possible as far as I know. You cannot login to a JK BMS remotely unless you have some sort of remote Bluetooth system... don't know if something like this would exist.
Andy if you can work out how to fix that new JK Please let me know as that is the same problem i have here. I just installed a new JK i brought on ebay, Has hardware ver: V11.xw and software ver: V11.288 version v4.17.0 The other problem with it is that it will say it is 0% Soc when you hit a cell voltage at 3.284 v I'm using the same settings as you. As i use you as my bible 🤣🤣 Thanks for the great video
With BSC on an ESP32 you can read all values from a JK via Bluetooth. I think this should also work with the new Inverter JK. BSC can send all values via mqtt and this could be read by HA. Or do you think of controlling the charger? This depends on the charger or inverter with integrated charger. Many brands can be controlled by HA
I'm working on a solution to hardwire the JK-BMS, like my other boards, so you could use the New JK-BMS as a standalone or master/slave with other BMS eg older JK or JBD, or other CAN Batteries.
@@Juergen_Miessmer I don’t have HA I’m using solar assistant but HA looks so much better. I can control my inverter already using SA however not as easy as this dash board this looks great
I'm not saying anyone should buy it! I'm just testing and showing it here on the channel. You didn't buy the Multiplus either, just because I showed it in a few videos, right?😉
Create a set of automations that control the max charge current. The. You don’t need to worry about manually setting it. I have one on temperature, one at 95% that starts slowing, another when max cell voltage goes over 3.6, and another at 100% charge. They change the charge current to have a soft landing. The. I have ones going down that reverse that. Also for everyone else, a $3 esp32 you can do all of what rheee boards do for free otherwise using the open source library from which this is based.
Probably never. The guy building the Peter Board took the free and open sorce work from many others, simply added the Bull, absobtion and float Logic and sells it for crazy smount off money. Since the communication from seplos to esphome is not allready implement by someone else it will not happen. Thats why only the few bms's are Supported because someone took a lot of time and Energie to implement it.
Hab kein Edelstahl ausser dem Gewindebolzen. Aluminium Terminals mit Aluminium Bus bars. Hab soviel verchiede Pasten probiert. Am besten laeufts ohne irgendwas ist meine Erfahrung.
@@OffGridGarageAustralia ich dachte in dem video sagst du dass es probleme gibt diesbezüglich? die wago Paste hat abrasives material drin. sehr fein aber es fühlt sich rau an. das reisst die oxidschicht auf und die paste hält den sauerstoff nachher fern. das ist nicht irgend ein hokuspokus produkt sondern wird bei elektroinstallationen angewendet wenn alu leiter zum einsatz kommen. das nicht ohne grund! ich verwende immer ein klein wenig von dieser paste. auf den batteriepol draufgeben und dann die busbar drauf und etwas kreisen lassen um die oberfläche anzuschleifen, dann anziehen so fest es die alu gewinde zulassen
@@OffGridGarageAustralia Oh man, I didn't think you'd respond to my unintelligible comment. Lol. But hey, thank you for you and your videos. I was able to DIY my solar setup mostly because of your vids. Almost a zero-bill setup, not quite. I will get there, eventually.
I dont think more than 1p is practical with these cells. A 1p System is always safer. And how much more those a 3 × 16s system cost compared to 1 x 16s3p? Only the bms and a few cables which is well worth the safety in my few.
Most of the off-grid users don't need any of this... What we need is that if we have sunny day to change our batteries, and supply the home application, and your victron system is not possible to do that because BMS is control usage from the PVs. ...so reason why this float is useless in your system is,if some moment house need more power, and there is cloudy for the moment, it will use the energy what we want to keep for the night time... After the cloud get away, there is no chance to charge that float , and we are missing that at night... Better is to use some high voltage mpp system... Or simple Growatt or Deye... This is not necessary to control remotely, but if you have to...then something is wrong with your system... For the moment i install over 430 systems, and most of them are missing more battery capacity... and some more solar panels... Anyway... just we need to learn to live with solar system... Good luck guys... have a sunny day... (Sorry for the folks covered by snow at this moment, hope that sunshine will come back soon... just try to clean up your PV to get some light on it...)
That's not how floating Lithium works. We charge to say 55.2V absorption voltage and stay there for 1h to balance. After that we drop the voltage to 53.6V. That means, we are still 99.7% charged. And the controller will keep this voltage for the rest of the day. It always brings the battery back to that voltage (if there is enough sun) even if we had a big load for a while. We will NOT lose any energy through that and have 100% available for the night. Sorry, but what you're saying is not correct. I have shown this many times in my videos.
What would happen if you decreased the max charging current in the master and the two slaves had charging mode enabled? So both slaves would have charging set to 100A with charging enabled, but the master max charging was at, lets say ... 50A. Sorry ... off Topic: For those looking for fairly cheap flexible 1/0 and above cabling. ... look for Locomotive Cable at an electrical supply shop. $4 USD per foot, and you could make buss bars from it, it's that flexible.
I am glad that Andy is finally seeing the value of HA. I have been running HA for several years now, it is my "PMS" (Powerplant Management System). It gathers all of the data from my Inverter, BMSs, etc. and provides me with a LOCAL human interface to monitor and control everything. I built several boards to monitor & control JK & JBD BMSs, MPP & PowMr Inverters, as well as PZEM Power monitors (Shunts and Current Transformers), and automatically populate all the data points into Home Assistant. The net result is that HA is the "Brains" of my system, not the BMS.
I hope the system will still safely work if HA does not for whatever reason?
I'm using HA only for monitoring atm. Even if the BMS fails, the MPPTs will only charge to a certain voltage...
Thanks for the update video. Love the build process!! Question/Observation: I don't think the "Charging Enabled" toggle is a good option as it implies charge is ON/OFF while it is not. It is basically a toggle that reduces the CCL (removes the battery from the CCL calculation in Victron). Maybe call this CCL include or exclude.
Thanks again Andy!
Yes, that's what I said and explained, the charging enable toggle has only an effect on the CCL. In Home Assistant, you can call it whatever you like, this is just the default name of the toggle. I know what it does, so it's OK for me, but thanks for your advice and explanation for others again.
It was very nice to see how you made the home assistant work. Maybe you could show how you did it on another video a good day. I have trouble to make connect to the victron system😮
Yeah, that goes a bit away from the solar and battery topic though and only a few people would be interested i such a video. I'll have it on my list...
I have been using Home Assistant for 2 years to monitor and control all aspects of my home powerplant. Having no Victron components, it was the best choice for me. I have designed a few "Dave" boards that interface my Inverter, BMSs, and Shunts to HA.
What if you HA fails though?
@@OffGridGarageAustralia What if the BMS Fails? What if the RasPi fails? What if a circuit breaker fails? After 2 years of daily use, I have very strong confidence in HA, and I have no dependency on the internet to update a web site in some foreign country (that I have no control over) and I have full control of the data display and control logic. I have WAY more confidence in HA's development approach than I do in JK's, based on several years of direct experience with both.
@@DavidPrueThanks for sharing, I appreciate it.
Excellent I am glad that you finaly use Home Assistant to display / control your system. So much nicer so much powerful 😊 Would also be nice if you could recap how data are exchange in your complete system ( ve bus, cerbo gx, can ...)
Yes, I will make a separate video about the coms in my system. For this test, everything has changed...
I want to point out that all the software that makes the peter board possible is open source and done by a lot of people in their free time. Esphome is the key word if you want to look it up. Some other guy did all the work implementing the comunication from the bms ti esphome.
He took all this work from others, added some charing logic, then makes it cloesed source and sells it for an insane amount of money.
Keep the source open and simply sell your designed hardware with your firmware pre installed, so people which want a plug and play version can but it from you, but taking an open source Projekt modify it and not letting the whole comunity profit from it but just yourself is just wrong and you know it.
another entertaining and informative video great Sir! Love from the states
Thanks a lot my friend from the states!
It's always nice to see your videos. You are taking all the information on the dashboard from Victron or from the PeterBoards
The data come from the Peter boards from the single BMS (2x JK and 1x JBD). Some Info also comes from Victron. All data is coming together in Home Assistant.
Awesome channel Andy.
Off the SPAT and doing Jim Beam 0 degrees and 4" snow on the ground 😅.
Woohoo, enjoy 🍸
I like the Dashboard Andy, big improvement. At 15:45, I as you have a good milliohm tester and a baseline on your you pack with the Aluminum busbars, I might suggest testing an aluminum anti-oxide compound to see if it adds any resistance and prevents corrosion and resistance build-up.
Many years ago we used a lot of aluminum conductors in the oil and gas industry at a remote location and found 'Penetrox A' (DXE-P8A) an industrial Aluminum Conductor Anti-Oxidant indispensable, and that it effectively prevented aluminum to aluminum and aluminum to copper connection corrosion if used a liberally. The stuff is relatively cheap and a bit messy, but worked well on connecting aluminum to copper buss bars and even using copper compression lugs not rated for aluminum, on aluminum wire.
Thanks Jack. I have tested three different paste, greases and compounds in the past. They all seem to work well for a while but eventually have made the situation worse. I had the best experience with NO grease whatsoever.
I also have a JK BMS 60A / 0.6A Balance and its just working fine for a 16S 48v system w 9.6kWh.
Waiting for my Peterboard and will also start Home Assistent later this year.
I got 80amp / 0.6 amp BD6A17S8P with 16s 15kw should still be fine?
@@nodelayfordays8083 That must be 300/320 Ah cells. Looks ok to me. When all cells are almost the same voltage, balancing with 0.6A is enough I guess.
DIY Battery banks I like more than ready made Storage products as they are a lot cheaper. And I like JK BMS. So I have ordered the so called Peterboard from Uksa 007 to be able to connect to my Deye Inverter over Canbus .
Awesome, I really liked it, it will be very inspiring for many people.
Automatic Plug & Play Home Assistant integration should be part of all smart solar devices nowadays.
I hope solar device manufacturers will wake up and start putting this in their devices.
It would be great to follow up with a video showing the process of integrating the whole Victron System into Home Assistant.
I have all Victron Gear in HA for monitoring purposes. I still like the VRM better, so will only have a small dashboard in HA with some selected numbers I'm interested in. For now at least😆 HA is a can of worms!
This is going to be super helpful in smartening up my system, thanks jeeps Andy n Peter 🤙
It has been great since installation. I love the monitoring.
Thanks Andy - love your work as always (though way beyond my level of expertise).
Stay cool, and stay charged.
I’m very happy to have 2A balancing with my JK BMS’s, and would actually like even more - I occasionally need to extend the absorption charge time to fully balance if it’s been a while since they’ve been fully charged.
Is this for your 48V system? I wonder why they are so out of balance that 1h is not enough.
Because they are grade B cells not matched what you often get from alibaba and aliexpress plus a lot of vendors. Or people think they safe money when buying garde b labeled cells from reliable sellers when the save 20 Euro on a 130Euro cell and think they just get a bit less capacity....
I agree with him its better to have more balancing current then you actually need right now as the cells aging over time and not evenly so the delta the balancer has to cope is getting bigger over time. So 0.6A might be ok now but in 5 years not anymore. And if i am building a system its designed to last the battery life.
Forgot to add the more cells you have in series the more the whole cellpack will deviate over time, means a 48V pack needs always a BMS and a capable balancer while if you have high quality cells in 4S a battery monitor with an external relay like Victron BMV712 that does LVC and HVC is more then sufficient. Know over 40 installs (half i done myself) in cars older then 7 years that have no BMS with 4x40AH till 4x100AH Winston cells as only battery, no BMS. Took some appart and deviation after 7 years were below 5mV. And they often see 5C peak use from the engine starter surge. Forget doing that with 24 or 48V as the production differences of each cell add up when you put them in series. And that added differences which grows over time need to be covered by the balancer.
And if you use that in eg boats where the bank capacity is often twice or more of the solar total wattage these banks are massivly oversized so getting seldomly to 100%SOC. Means when they do the active balancer needs to be bigger to balance the bigger deviation as a result of that.
@@chrisr819 Yes, the more cells the bigger the problem, as only two cells can balance at any one time.
I have a 2 x 18s system, and often go many part charge cycles before filling and getting a chance to balance. I usually need every bit of the 2A balance current if it’s been a long time since the last full charge and balance.
Enjoyed the video. Thanks for sharing
Thank you.
That all in one monitoring system and so on.. as you have it sure looks useful ... pat
First attempt to have all three batteries one page. This is also check on the Peter boards if they are doing the right thing...
Hi Andy, that is so cool!!! I really need to work on the connectivity in my workshop, it's a total black spot.
You will get there over time. It wasn't a priority for me either. But now the installation for the shelf is done and I can look a bit further...
Andy, have you had a problem with firmware update failing and then the JK BMS locking up? I tried to update the BMS, got a failed message and now I can't see the BMS in the app or computer.
Yes, I had one as well. Fix is in this video here: ua-cam.com/video/KznZtqpoles/v-deo.html
@@OffGridGarageAustralia thank you I saw that video got mine fixed as well you are the best!
Having the ability to set a float and absorb is something I really miss having. Thanks Andy
Yes, especially with the older JKs and JBD BMS. They don't have these features built-in like the new JK Inverter BMS.
Hello Can you tell me if I can use Peltier cooling of the battery enclosure, as is the case with you? In summer, in Romania, in my garage, temperatures reach 40-42°C and the optimum is 25°C Thank you very much
No, that will not work. Peltier does 10° max and produces a awful amount of heat on the other side which you need to get rid of.
I have the same temperatures and the cells will be fine for these few days. They can handle close to 60°C
@@OffGridGarageAustralia Thank you
Andy, The float mode reduce the voltage, how it works ? where the V are going ? I can understand Amps are inside the cells but the Volts...
Don't know exactly what you mean?
How the Inverter decrease the Voltage ? I think it would be good to understand how & why. Amps goes inside, and to be removed need to be used or converted in heat, but the voltage reduce process......@@OffGridGarageAustralia
@@christophec252 The board controls the charging voltage CVL, so the inverter only charges to this set voltage. Like with any inverter or MPPT. With these boards, you have control over that voltage from your mobile.
My solar panels are covered by 9 inches of snow and ice. For 7 days my production has been zero. Welcome to winter in Oregon.
Even in winter, down south in Australia, you may get snow once for a day in 50 years. lol Not a fan of snow.
Sounds like a classic case of needing VERTICAL, or nearly vertical solar panels- facing South in your case.
Here in sunny Brisbane …. Should I say cloudy, showery, stinking hot and sticky Brisbane, we’re getting almost constant showers between patches of sun, coupled with extreme humidity.
However my solar is still keeping my battery fairly full and running the aircon. (Couldn’t live without aircon- really disgusting weather to be honest. Despite being born and bred here, I survive our summers but absolutely LOVE our Winters. This Summer has been especially bad.)
Vertical is deceptive to winds or you need to put them lower, where shadow make them useless. I created a flimsy 45 degree setup, with plastic panels which can sway in the wind, they are not zero. But almost
There are systems to melt snow off solar panels, and ground arrays are easier to clear manually. They do also make special very long handled snow blades for clearing roof panels. But it's a pain any way you look at it.
I had a lot less snow, but it is still quite annoying when the sun is shining.
How about just using an electric blanket with an extension cord, wrapped in something safe and waterproof?
Did the jbd balancer turn on? At the end of the video it still showed off.
It's a charge balancer (which is turned off). It only turns on if there is 0.0A showing. Stupid design from JBD.
Welcome to the unlimited possibilities of HA 🐸
My Battery Dashboard looks nearly the same.
An i have a second dashboard showing all single cell voltages. With a few clicks i get pretty graphs with all cell voltages and can see how diffrent cells can behave during charging.
The Deye Inverters have two limits for charge current. One is set via CanBus from the BMS port. This should never be touched manually for saftey reasons.
The other limit can be set via the display or via ModBus from HA - manually or by automations.
HA is really a great peace 👍
Yeah, HA seems pretty slick for this sort of monitoring. But would I trust it to run the whole battery shelf including charging and inverter control? What if it fails and does not run any more? Yes, we have backups, but how to recover the whole system after a crash... I may need to test this out at some stage...
@@OffGridGarageAustralia
I dont use HA for safety relevant controls. But it is perfect to manage several things.
For example:
If sunshine is forecastet for the whole day i reduce charging current, or i heat up the boiler first.
Or i can switch off one or more inverters during the night to save idle power.
Or i can overheat the house a bit if the batteries are full an there is still solar power available.
Or a additional colling fan is controlled by the temperature sensor of the inverters DC board.
If HA fails nothing happens. As the safety relevant limits are controlled by BSC and by the BMSes.
The minor settings HA is doing in the inverter can still be changed using the buttons and the touch Display on the Deye.
@@Juergen_Miessmer Yes, that is perfect and what I like to do as well. Atm I look outside and change the current accordingly.
I was just wondering about this. Im making a 15kw system with 16 xells in series Eve 304ah and got the JK BD6A17S8P BMS with 0.6 amp active balancing amd was wondering if that would eould be enough for this build.
Yes, it will be no problem. Actually, this battery is now my best-balanced pack in the battery shelf.
Home assistant seems nice :) Are the middle shelf also the 280K ones, I know the top shelf isnt a K type electrolyte, might not be the bus bars but I see a bus bar swap soon as a test - well i hope you try and prove the bus bar theory.
Top is LF304
Middle and bottom are 280K cells.
@@OffGridGarageAustralia ah, thought only 1 of the 280s were k electrolyte. There goes that theory.
I like your dashboard.
Today I installed a second Multiplus II in parallel to my original one and connected two additional 280 Ah batteries. This gives me now 60kWh. Enough to charge the car from the battery in summer. Unfortunately it's not summer right now.😢
This depends on where you come from 😜
It's very much summer here!😉
Great, 60kWh sounds like a good capacity, especially when charging a vehicle.
Have you considered polishing the contact points in the AL buss bar setup. That should bring connection resistance back to par. Now you could know for sure why the difference.
A T Burke
Yes, I have polished aluminium terminals in the past😆
The surface will corrode within milli seconds and yes, I have tested several aluminium pastes and greases. Best result for me is still nothing in between.
Thank you, Sir. I think I will stay with copper then. No matter what size bar is used to reduce resistance, it still has to make it through the connection.
A T Burke@@OffGridGarageAustralia
Hello Andy wonderfull video. Homeassistant seem very simple and power full i wonder if a comparaison with venus os would be usefull. Seem to me Homeassistant is more power full and easy to use but your opinion could be great.
About connexion i discover there are a lot of way to connect JK-BMS to Home assistant.
I say that because Peter board is very expensive (almost bms price) naturally the work done is great and have to be paid but the gap is very hudge (hardware arround 10 euros and price more than 100 euro)
The big difference i have seen is the peter board manage the charging and not only connect the devices.
Thanks for all.
I trust Victron more than Home Assistant though, so I use HA only for monitoring purposes.
That sounds good！🤩
Is there a rule of thumb what current the integrated Balance should be capable of based on the battery size ?
I can remember that many had taken the 2A JK BMS for the usual 14,3 kWh EVE LF280K batteries, but what about a 3,6 kWh LFP 16S battery (74 Ah) ?
16 x 74 Ah x 3,2 V = 3789 Wh or 3,8 kWh which would mean 27% of the big 14,3 kWh.
And these 27% would equal with a 2A BMS just 0,54 A
A 0,6 A balancer should have even more reserve, but the main question is still:
How should the balancer be layed out to get the battery balanced (battery capacity) and when to get a 4A or a 2A or 1A or 0,6 A on
Dear Off-Grid-Garage Community! I‘m a big fan of this channel for weeks and I want to use the community to ask some questions to fine the best setup I can get from Victron Multiplus 2 + MPPT + Cerbo GX + DIY Battery + JBD BMS. Can I raise this questions in this video? Br
Thanks a lot for your question. I have made a ton of videos about exactly such a setup. You can leave your question under any video, no problem.
I like the Setup you have done. I am planning on installing 2x Orient Power 230Ah 51.2V 11.8KWh batteries in parallel with my existing Simpliphi power 75Ah 22.8 kWh Battery Bank. The only problem I will have is that the Simplphi Battery does not have an integral BMS but rather is controlled by a Schneider Electric Connect BMS for SOC control. Is there any way I can control the three battery banks' BMSs? I can see that the two Orient Power batteries' BMSs can communicate with each other by connecting the communication cables. I just don't know how they could communicate with the Schneider Electric BMS. I can set the 2x orient Power batteries BMS settings locally on their perspective BMS and then have the Schneider Electric BMS control SOC control to the combined battery bank.
Interesting. I've never heard of a simpliphi without a bms, nor a schneider bms. Do you have a link for either of these?
The question is if you need any SOC control at all for your setup or can just use the battery voltage for determine CVL.
Hi Andy ✌️
Is there a way to get the new options from the new JK BMS to the older ones?
And at my older JK, I see all the time some heating current, but I have no current option an nothing connected. Also the heating is switched off in the settings. Someone had any ideas?
Best regards from Jägermeistertown 😊
If the BT app does not show the new option, the hardware of the JK cannot handle it. No way to update. Hence the Peter Solution for older JKs...
To confirm if it's the bus bars, Flip the JBD and JK BMS around next and see if the results are the same or different. Then we'd all know for sure. I'm at least curious about it.
Yes, I have to do this at some stage... I want to know myself.
one thing to note is the jk bms reports full pack voltage based on its individual cells voltage. if the calibration is off between two jk bmses connected in parallel the overall pack voltage is obviously the same yet the the two bmses report slightly different pack voltage number.
Are you sure about that?
Why would they add all the cell voltages together for the Overall pach voltake instand of Measuaring the pack voltage directly?
JK uses coulomb count, not cell voltage as far as I know to display SOC
Which pack of batteries had the damaged cells replaced ? with newer cells, would that affect charging?
That has no effect. It was cell 15 in the bottom shelf battery.
Hi Andi,
I am a little bit concerned: In the past you had a 2 layer of protection
A) The Solar charge controller
B) The BMS
Both systems worked completely independent, if one fails, the other would prevent the battery from overcharging (witch can result in fire)
But now, if the BMS fails, there is a chance that the system will continue charging up to very thing goes to fire.
Maybe you can make a video on this safety issue
Lifepo4 does not burn up due to overcharging. It will swell, degas, and go kaput. Other chemistries run the risk of fire on over charging.
What fire are you talking about?
If the BMS fails... if the circuit breaker in your house fails. That's not good in both cases...
In any way, the Peter would still charge to 55.2V. So, no fire?
@@OffGridGarageAustralia You might got me wrong. In the standard setup (solar charge controller not connected to BMS) you do have a dual layer of protection. Even the SCC fails, the BMS will shut down, or if the BMS fails the SCC will have a maximum (bulk) voltage he will charge to. No issue, no potential hazard.
If now the SCC lisens to the BMS, and this (e.g. measures a worng voltage due to component failure) the charging will go on upto the damage of the battery (and maybe, maybe not - a fire hazzard)
That's just my point, by connecting, you remove one layer of protection.
@@peetbraun9439 We still have the BMS and the SCC. We only add an option to modify the coms in between. The BMS will still disconnect if cell voltage goes over 3.65V even the board will go crazy and tells the SCC to charge to 5V.
Same, the other way around if the 3.65V protection fails in the BMS, the SCC will still charge to the set voltage the board tells it to charge.
Nothing has changed and no safety concerns.
Would this work with my 12KWgrowatt inverter?
Again Andy great vid. Has anyone bought the Peter board for selectronic so-pro?
The Peter boards support the BYD, Deye, Goodwe, Pylon and Victron CAN protocols for now. Hard to tell, if nobody has ever tried it...
I like the high level of Home Assistant integration. That really is the ultimate in convenience! You might want to consider some security isolation, you don't want a hack of your main network to be able to access your equipment network.
The basic way to do this is to have the machine Home Assistant is running on "straddle" the two networks. Your home network on one side and your equipment network on the other side. This would require having two ethernet ports on the machine H.A. is running on though which might not be convenient for everyone. Even on an RPi4 you can always connect up a USB ethernet dongle to get a second ethernet port. You then put the firewall on the H.A. machine itself. Basically allow the equipment network access to the home network and internet, but not allow the home network access to the equipment network.
A second way is to isolate the H.A. machine and the equipment network with a router and then drill a hole through the router (figuratively speaking) so your phone app can access the H.A. web server when it is sitting on your home network, but nothing else (requiring that the phone be in bluetooth range or sitting on the equipment network to access devices directly).
Its a project though. Most people aren't conversant enough with LAN routing and IP spaces to do it without a ton of effort. I do think you need to begin thinking about the security of your power systems network integration, though.
--
I think you can safely start balancing once the cell voltage is high enough, even if a pack still has a lot of current. The energy the balancer is moving in the battery is effective regardless of the state of charge, as long as the balancer makes the "correct" choice. That's the basic problem with trying to balance at a lower voltage... the balancer winds up making the wrong choice and taking the cells in the wrong direction due to minor differences in cell chemistry.
Once a cell is at a high enough voltage the balancer should start making the correct choice, even if there is current going in or out of the pack. And honestly, there is never "zero" current.. the battery bank will be charging or discharging, but hardly ever sitting near zero.
Heh. Remember the time range for absorption time that I've been hawking for the last few years? "1-2 hours". You are getting closer 🙂
-Matt
Thanks for your ideas, Matt. Appreciate your feedback and inspiration!
Hello OGGA community. I have a question for you. I need to find a reliable bluetooth balancer for two 48v strings of four Lossigy 300ah LiFePo batteries. What are your recommendations?
You would need a 12V balancer like this one from Ali: s.click.aliexpress.com/e/_DFnmJnn
It is also good to have InfluxDB and Grafana running in HA, because HA only keeps the last ten days of Data.
Yep, that's what I do. Have a separate pi with nodered, influxdb and grafana.
I have data in my HA since day 1 when I installed it. I can go back months...?
@@OffGridGarageAustralia
Sensor data is purged by default after 10 days. Only if an entity is build for long term statistic, it will store one summarized value every hour.
@@Juergen_Miessmer It still gives you the data before the 10 days. I'm happy with that.
Using a Bluetooth connection instead of UART would have allowed the Peter board to truly cut off the charge from the BMS (the UART connection only allows reading of the BMS). But is it good to use a Bluetooth link in production? I'm testing it at home with two BMS monitored via Bluetooth by the same ESP32.
I would not trust Bluetooth for such a safety related connection and function. It is far to unreliable. I guess he made a similar decision at some stage.
You've got this in your own project, I guess?
@@OffGridGarageAustralia Yes for my batteries I prefer the wired connection. But I had requests to use Bluetooth connection so now I offer two different YAML, Wire and Bluetooth. The Bluetooth connection being tested at home is stable (jk-bms code from syssi). Sleeper85
@@OffGridGarageAustralia Yes for my batteries I prefer the wired connection. But I had requests to use Bluetooth connection so now I have two different YAML, Wire and Bluetooth. The Bluetooth connection being tested at home is stable.
@@OpenSolarEnergy That is great progress. Thanks a lot for your work and sharing it here. Once it's ready, maybe I can show it on my channel as well...
Andy, which home assistant are you referring to? Alexa, Victron, peter?
HomeAssistent is the name for a package of software running as service on a linux host.
@@Juergen_Miessmer Thanks Juergen, so if using Peter boards with a Victron Cerbo GX this cannot be done!
@@stoptheirlies
Peter board can be used and configured standalone.
But the nice dashboard Andy is showing here is a HomeAssistant frontend.
If you put no-ox paste between terminal and alu busbar there cannot be an oxide layer as this connection is airtight and no-ox is sealing it permanently. This means no oxide layer can happen. I assume the total internal resistance is higher then the other two banks.
Tried that and some other pastes as well. Best result was without anything though...
I have different experience and i am dealing with Lithium nearly 20 years now, the first 10 years in competition car stereo and the first lithium we got "illegally" from a scrapped military vehicle. The best result for a short period like 2-3years without paste if the terminal are metricously even and the busbars too...but thats rarly the case. And when you sand the battery terminal by hand you destroy that. Means the connection is not airtight anymore. The paste is highly conductive and make sure the whole surface is connecting and 2nd making the connection 100%airtight so that connection cannot corrode over time which otherwise happen if you eg use tinned busbars as contact corrosion can occure over time. I use either aluminium or silver coated busbars (which eg is used on NH fuses and high current environment installation like grid stations) with No-ox paste to seal connection.
hello Andy, I just sent to you to your e-mail a few screen shoot from all my datas to home Assistant
Thank you, I have seen it and replied.
The 'Peter' board seems pretty useful. Would love if it could update quicker than every 10(?) seconds. Also, what's the logic/reason for having "charge enable" and "charge current" on the slaves. Why not control it all from the master instead since it affect global settings, not individual slaves/BMS's?
If one or more of the BMS is off-line, or has gone into protection the current will automatically reduce by that slaves current setting, this is the primary purpose of the slave current settings, each slaves tells the master how much current it is able to accept.
Niiiice 👍
If Home assist is logging the data, can you export 1 or more days & compare the sum total of amps in-out of each shelf? or daily averages. Each shelf will have a typical % share of the total Ah per day. And compare this with your estimates of cell capacities.
Yes, you can do this directly in Home Assistant and have three graphs showing the data of the batteries.
Three phones 😂 well we do what we can. It is a work in progress 😊 ✅
Im pretty shure the reason are not the busbars.
I think the cells are different : chemistry, machanical tolerances, charge curves...
Charge curve could be possible... I now have the data to compare.
@@OffGridGarageAustralia
I can barely wait for your results 🐸
I ran across some 1.2A / 1.3A balancers on aliexpress that work all the time and are of the distributed / headless kind and most probably the future as they are very simple and economic.
The "xS Active Balancer Equalizer Board" type sits between 2 cells in a pack and only cares about those 2 and moves energy between them if their voltage difference is big enough - keeping them on the same level. So for a 8 cell battery it's: *cell1-AB1-cell2-AB2-cell3-AB3-cell4-AB4-cell5-AB5-cell6-AB6-cell7-AB7-cell8* . In contrast to that does the MCU and active balancer on a JK-BMS for example jump around all the time and needs semiconductor switches that must not fail + needs those balancer wires (with resistance and all the associated problems).
We cannot use these 'always on' balancers for LiFePO4 as they cause an imbalance when in the flat part of the charge/discharge curve.
The active balancers out there including the NEEY and JK balancers seems to be very robust and working extremely well.
So why would you want to go into float mode at all and not stay on Absorption until the battery goes down to 80% or so abd then switch to bulk again?
That's not possible to stay in absorption and let the battery go down to 80%. Also makes no sense with LiFePO4 to go 80%
@@OffGridGarageAustralia ok i see.
But why would you need to set a float time at all?
Im the victron system you can change the charging algorithmus to switch back to bulk once the voltage dropps a definded voltage e.g. 0.1V below the set float voltage.
Once there is no solar the voltage of the pack would naturaly drop.
Is that not precice enoth or way is the float time needed?
@@RFDarterI think that has been explained in the latest video now: ua-cam.com/video/eZX7CkEEq_8/v-deo.html
Ty Andy, compare tepretures with "Me", Insane (another earth side)..🤧
Andy, ALWAYS GREAT INFO.
PKEASE HELP!
Can you solve JKs BLUETOOTH "failed pair" mystery?
JK keeps giving me new unique bluetooth passwords, that only work for 5-15 minutes...
Then we have to start ALL over. Has not mattered the version I use... going on a month now.
Has the BMS ever worked on BT normally? Tried a different phone?
Andy believe it or not it's possible to get the old faulty NEEY balancers eating out of the palm of your hand with an esp. Andy I live near to you.. we should probably catch up one day. I've been custom designing BMS and SCADA systems to use with victron gear for around 8 years now....
The NEEY Balancer. Don't hey just work once programmed?
@OffGridGarageAustralia no. Remember version one that lots bought but they had no ability to purely top balance? There's be thousands out there nearly useless. An esp makes them into magic. Btw... I'm actually a telecommunications carrier - in my experience 0.6 amp BMS should be a ok. Cheers - matt
@@OffGridGarageAustralia further to my comment just above - regarding home assistant, no, you don't need it to be running for these to work. I have arduinos and esps of my own code fully controlling victron sccs (and more....) with no home assistant - in the latter case you just use esphome. I have had this gear in the Australian desert for 8 years - regular 45 degrees heat, only just replaced my last set of AGMs, have been using lfo now for about 4 years too in the same environment. also.. no shunt, but hyper accurate soc from victron scc's via coulomb counting - how'd you do that I hear them ask - you really should come visit me 😉
just add victron integration to home assistant and use victron dvcc settings as same slider
I have the Victron integration but have not seen a DVCC slider anywhere...
What is the entity called?
The voltage difference between the packs could be due to the calibration of the BMS. They should all be exactly the same voltage as they are all on the same busbar. The charging current could be different because of the cell chemistry.
The Voltage is only exactly the same if no current is flowing.
... and all BMS are correctly calibrated.
Whats your JK model?
JK-PD6A20s10P
Do you sell any 3.2v 100-150 ah cells;
No, I have links of reliable sellers on my website where you can buy: off-grid-garage.com/batteries/
@@OffGridGarageAustralia thank you for your time and for your videos
3 of 6 my JK’s will not reset to 100%. Same settings as you. Been like it for 18 months.
Do you also have the SOC100%Volt setting?
Mine definitely does not work... I've got this on one of the next videos. Seems like some BMS just don't do it.
Mine does not have the 100% SOC value. Ive been fiddling with the OVPR voltage and UVPR voltage. Deep cycling etc... seems to be closer, but it's still off by a fair bit.
Quick question, wasn't it always said that Lifepo cells don't need a float mode? If the cells are reduced from 55.2 to
53.6 volts, they are discharged and are no longer at 100% SOC. Wouldn't it be better if they are not actively discharged, but can simply rest without discharging? The voltage would still be falling slow.
Kurze Frage, wurde nicht immer gesagt, dass Lifepo Zellen kein Floatmode benötigen? Wenn die Zellen von 55,2 auf 53,6Volt reduziert werden, werden diese doch entladen und sind nicht mehr bei 100% SOC. Wäre es nicht besser wenn diese nicht aktiv entladen werden, sondern nur einfach ruhen können ohne entladen? Die Volt Spannung würde trotzdem fallen, aber viel langsamer und natürlichen?
😊
Dazu müsste man die Batterien vom Busbar trennen.
Das geht bei einem OffGrid setup aber nicht. Die WR müssen ja immer die volle Leistung können, auch wenn die PV mal z.B. wegen einer Wolke das nicht liefern kann.
Die Batterie muss da immer als Puffer dienen können.
To do this you would have to separate the batteries from the busbar. However, this is not possible with an off-grid setup. The inverters always have to be able to provide full power, even if the solar panels cannot deliver this because of a cloud, for example. The battery must always be able to serve as a buffer.
@@Juergen_Miessmer Wieso trennen??? Als Andy den Floatmode aktiviert, wird doch aktiv entladen bis die 53,6Volt erreicht werden. Hat nichts mit Offgrid zu tun sondern wie der Floatmode immer funktioniert. Die Frage war, ob einfaches Ruhen nicht besser ist.
Why disconnect???
'When Andy activates the Eloatmode, the battery is
actively discharged until the 53.6 volts
are reached. Has nothing to do with off-grid
but how the Eloatmode always
always works. The question was whether simple
resting is not better.
@@derfreiemensch
Weil an der Busbar nicht nur Laderegler und Batterien angeschlossen sind, sondern auch die Wechselrichter. Diese brauchen ja weiterhin Energie um das Inselnetz zu versorgen. Und diese Energie muss irgendwo herkommen. Ruhen ist technisch bei einer Inselanlage nicht möglich.
Ohne float würden die Batterien sofort stark entladen und die PV nicht mehr genutzt werden.
Because not only charge controllers and batteries are connected to the bus bar, but also the inverters. These still need energy to supply the load. And that energy has to come from somewhere. Resting is technically not possible with an offgrid system.
Without floating, the batteries will be discharged much stronger an no more solar power is used.
@@Juergen_Miessmerwieso? Einfach im Absorption Mode bleiben, dann wird doch nichts getrennt oder abgeschalten.
Und der mppt stellt weiter die Energie für die Verbraucher zur Verführung ohne die batterie zu entladen.
@@Juergen_MiessmerDas ist nicht richtig was Sie schreiben.
Oh good, I need a distraction. Just started it. 😊
Welcome Gaz! 🥈
Is this using Modbus or MQTT?
RS485 from the BMS to the Peter Boards and CAN communication to the inverter/Venus OS...
Hi Andy is it possible to remote login to the jk inverter bms when no victron stuff is being used.If so can you do a video for us stupid people that just use a basic system. Many thanks.
No, that's not possible as far as I know. You cannot login to a JK BMS remotely unless you have some sort of remote Bluetooth system... don't know if something like this would exist.
Andy if you can work out how to fix that new JK Please let me know as that is the same problem i have here. I just installed a new JK i brought on ebay, Has hardware ver: V11.xw and software ver: V11.288 version v4.17.0 The other problem with it is that it will say it is 0% Soc when you hit a cell voltage at 3.284 v I'm using the same settings as you. As i use you as my bible 🤣🤣 Thanks for the great video
That's all sorted in the last video: ua-cam.com/video/KznZtqpoles/v-deo.html
Sorry Andy I meant the one on your bottom shelf of 2.0, Not the new one you can do an up grade on @@OffGridGarageAustralia
7:57 can you do this with home assistant with the new jkbms?
With BSC on an ESP32 you can read all values from a JK via Bluetooth. I think this should also work with the new Inverter JK.
BSC can send all values via mqtt and this could be read by HA.
Or do you think of controlling the charger?
This depends on the charger or inverter with integrated charger. Many brands can be controlled by HA
I'm working on a solution to hardwire the JK-BMS, like my other boards, so you could use the New JK-BMS as a standalone or master/slave with other BMS eg older JK or JBD, or other CAN Batteries.
@@Juergen_Miessmer I don’t have HA I’m using solar assistant but HA looks so much better. I can control my inverter already using SA however not as easy as this dash board this looks great
@@uksa007keep us posted!!
The difference in voltage might be the accuracy of the voltage measurements with different BMS's
That wouldn't explain the difference in charging amps at higher SOC though.
@@davidpenfold
One of the Murphy's laws may apply. Under the most controlled conditions the pice of electronics will do dam well what it feels like.
@@universeisundernoobligatio3283 Quite possibly. But the fact it's still taking in current does tend to point to a more physical/resistance issue
Yes, absolutely, we now have three different BMS measuring so there will be some inaccuracy..
Heya, yes that's a very nice system. If you don't mind I would like to "copy" that system for the use to bild my own batterie self.
Absolutely, go ahead!
💙🙏😎
Thanks man!
Peter board for 4 BMS is 810USD....sorry for 1 person show with 30USD for hardware this is far over the top. Sadly the inverter BMS are only 8S...
I'm not saying anyone should buy it! I'm just testing and showing it here on the channel. You didn't buy the Multiplus either, just because I showed it in a few videos, right?😉
People buying it because of your recommendations so you should give them the whole picture.
@@chrisr819 What is the whole picture???
Create a set of automations that control the max charge current. The. You don’t need to worry about manually setting it.
I have one on temperature, one at 95% that starts slowing, another when max cell voltage goes over 3.6, and another at 100% charge. They change the charge current to have a soft landing.
The. I have ones going down that reverse that.
Also for everyone else, a $3 esp32 you can do all of what rheee boards do for free otherwise using the open source library from which this is based.
Yes, you can do that. At the moment, I'm only monitoring to see how much trust I have in the Peter Board solution and HA.
When is it posible to buy the Peterboard for Seplos???
Probably never.
The guy building the Peter Board took the free and open sorce work from many others, simply added the Bull, absobtion and float Logic and sells it for crazy smount off money.
Since the communication from seplos to esphome is not allready implement by someone else it will not happen.
Thats why only the few bms's are Supported because someone took a lot of time and Energie to implement it.
Thanks Andy
@@derfreiemensch ?
@@RFDarterwhat did you mean with???? It would be nice to have the same options with seplos bms and many would buy it😊
@@derfreiemensch because you posted "Thanks Andy" without him repilying here.
Wago Alu-Plus-Kontaktpaste Art.-Nr. 249-130 hilft gegen Probleme mit Alu Kontakten ,-)
Das stimmt, Alu und Edelstahl passt nicht so zusammen. Ich kenn das mit Kupfer und Alu, da haben wir auch so eine Paste für in der Arbeit.
nun, hab sie auch verwendet. billiger als NOALOX, aber fühlt sich ein bisschen sandig an. hab auch noalox, gibts beim buerklin in D
Hab kein Edelstahl ausser dem Gewindebolzen. Aluminium Terminals mit Aluminium Bus bars.
Hab soviel verchiede Pasten probiert. Am besten laeufts ohne irgendwas ist meine Erfahrung.
@@OffGridGarageAustralia ich dachte in dem video sagst du dass es probleme gibt diesbezüglich? die wago Paste hat abrasives material drin. sehr fein aber es fühlt sich rau an. das reisst die oxidschicht auf und die paste hält den sauerstoff nachher fern. das ist nicht irgend ein hokuspokus produkt sondern wird bei elektroinstallationen angewendet wenn alu leiter zum einsatz kommen. das nicht ohne grund! ich verwende immer ein klein wenig von dieser paste. auf den batteriepol draufgeben und dann die busbar drauf und etwas kreisen lassen um die oberfläche anzuschleifen, dann anziehen so fest es die alu gewinde zulassen
🐸🐸🐸
JBD not balancing? 22:28
Nope, it's a charge balancer, it only balances if the current is 0.0A. Totally BS
Another Andy ism Masterpeter! 😂😂
I'm running out of names for all these devices now😆
Hey Hey!
Hey, third comment, so 🥉
@@OffGridGarageAustralia Oh man, I didn't think you'd respond to my unintelligible comment. Lol. But hey, thank you for you and your videos. I was able to DIY my solar setup mostly because of your vids. Almost a zero-bill setup, not quite. I will get there, eventually.
@@gamedex Hey, that is great. I'm always glad to read about others being inspired from my videos. Good on ya!
0.6 amps / 16 cells.
pls test 2p16s / 3p16s
I dont think more than 1p is practical with these cells.
A 1p System is always safer.
And how much more those a 3 × 16s system cost compared to 1 x 16s3p?
Only the bms and a few cables which is well worth the safety in my few.
Still the same shit with the JK-BMS: It does not calculate SOC correctly!
Once it is calibrated it works!
Like and Comment #1
Yes, will do a full video on the coms site. Thanks!
If it actually balances at .6a than yeah. If it's only .6a at a one volt difference like most balancers 😂
Most of the off-grid users don't need any of this...
What we need is that if we have sunny day to change our batteries, and supply the home application, and your victron system is not possible to do that because BMS is control usage from the PVs.
...so reason why this float is useless in your system is,if some moment house need more power, and there is cloudy for the moment, it will use the energy what we want to keep for the night time...
After the cloud get away, there is no chance to charge that float , and we are missing that at night...
Better is to use some high voltage mpp system...
Or simple Growatt or Deye...
This is not necessary to control remotely, but if you have to...then something is wrong with your system...
For the moment i install over 430 systems, and most of them are missing more battery capacity... and some more solar panels...
Anyway... just we need to learn to live with solar system...
Good luck guys... have a sunny day...
(Sorry for the folks covered by snow at this moment, hope that sunshine will come back soon... just try to clean up your PV to get some light on it...)
That's not how floating Lithium works.
We charge to say 55.2V absorption voltage and stay there for 1h to balance. After that we drop the voltage to 53.6V. That means, we are still 99.7% charged. And the controller will keep this voltage for the rest of the day. It always brings the battery back to that voltage (if there is enough sun) even if we had a big load for a while. We will NOT lose any energy through that and have 100% available for the night.
Sorry, but what you're saying is not correct. I have shown this many times in my videos.
What would happen if you decreased the max charging current in the master and the two slaves had charging mode enabled? So both slaves would have charging set to 100A with charging enabled, but the master max charging was at, lets say ... 50A.
Sorry ... off Topic: For those looking for fairly cheap flexible 1/0 and above cabling. ... look for Locomotive Cable at an electrical supply shop. $4 USD per foot, and you could make buss bars from it, it's that flexible.
100A each from the two slaves + 50A from the master = 250A total charge current from the inverter.