News from Off-Grid Garage; Dangerous JK BMS Flaw
ฝัง
- เผยแพร่เมื่อ 1 ส.ค. 2024
- This is a short video meant to signal-boost a very important video that Andy from @OffGridGarageAustralia just uploaded. The newer JK BMSes shipped with a significant bug in their firmware that can destroy LFP batteries.
If you've not seen his video yet, please give it a watch. It's a great video on it's own, and he deserves the views and credit for this discovery.
Major safety concern with new JK Inverter BMS! Not good.
* • Major safety concern w...
0:00 - Don't Use JK BMS
5:00 - Small Channel Update - วิทยาศาสตร์และเทคโนโลยี
Several folks are pointing out that this flaw effects the new BMSes only. To restate what I said in the video; I know this. I mention this. I am upset because this is the latest in a series of problems. They're willingness to change features on the fly, sell incompatible parts, and other issues I've had over the last two years all add up. This is a company that doesn't take QA seriously, released clearly untested products on customers, and then blames the customers for problems. The BMS plays far too crucial a role to be a playground for untested development.
The lack of testing in such a critical piece of equipment is what scares me the most. They literally let their end-users test a product that can theoretically kill someone if it has a bug.
BMS= Safety device . Not proving out Safety software is an unforgivable mistake in my opinion too .
I had a good laugh when you said "recall"....from China? Surely you jest! 🤣 If the default setting was to LiFePO4 it would have only pissed the Li-ion users off but not been unsafe. In ESS use they may never get turned off, but yes, it's a rather dangerous problem
You are right, it is a fatal bug and it should not happen. But in general I'm absolutely happy with their BMS in the past. Years of positive experience, starting and running for ever without any problems. This error comes ONLY with the brand new inverter BMS and they will fix this within the next days, they said to me in person. So, everyone knows, it is brand new and following Andys channel, we all saw the problems with the older SW versions. Yes, no excuses, this should not happen. But it happened. Anyway, I'll keep on using JK-BMS, because they are still the best for our purposes. Once the community is happy with the new inverter BMS and more long term experiences will be available, it will be again a great BMS for the next years, I'm sure!
thank god I only bought one at the time :-) sell it used, take a solid BMS * n BmS then you have time for other things and everything will be fine :-).
Happy New Year
☺🤗
The Community found this problem and informed me, I just made the video to get the news out.
I had seen mine resetting but didn't pay attention to that, thought it was me clicking the wrong button or so...
Great video and summary, Madi. Thank you!
Andy, Thanks to you for your video's. I have watch so much, you must graduate me, with a degree in battery's. I have gone on and build 14 battery's to date. these is for the 4 by 4 overland community's and the offroad trailer community's here in Africa. Your comment on the JK BMS is disturbing. what do you recommend as alternative ? 4 to 16 S systems. we use 280 Ah cells.
If you do not use the newest inverter BMS, you would not have any drama.
Aye, I made a point to call out your community as well in crediting the source. It's one of the very best things with youtube; The community brings knowledge and resources to the table we couldn't get otherwise. I really hope this was caught soon enough that no one out there actually lost their cells to this bug.
You're missing the point, this isn't an isolated issue, it's just the most egregious in a long line of questionable business practices.
True community Teamwork 😊
JK fixed it! Update available on the website tomorrow. AND - default setting will be LFP ! From my view a very quick fix and good customer service. I don't know any other BMS supplier getting customer feedback and bugs addressed that quick.
They really fix the "reset" problem or they just changed the LTO to LFPO default setting after a unwanted reset ?....
@@sebydocky5080 The reset and the default values 😊
My understanding is that it resets to LTO now, not LFP, but still that's much safer. I am glad it's fixed for those who bought it, but for me, for the reasons I explained in this video, I still don't trust them anymore.
@@TheDigitalMermaid
JK note on their Website for the change:
"Change the default configuration parameters to lithium iron to improve security."
@@michaelwessel2983 Ah, ok, good to know.
Thank you for making this video. I have some older JK-BMS units for 48V, which I have not yet installed. Their intended use is in a motorhome which may or (may not) have intermittent usage with prolonged down-times. I know some have already commented that this flaw is in the newer BMS units. However, I am not willing to risk my family sleeping above 28KWH of LiFePO4 batteries that might subject be overcharging.
I will be watching both you and Andy for future info.
I have to rethink my BMSes also, I am not going to take JK onto my boat. Same as you, I know these ones are probably OK, but I just can't trust them anymore.
@@TheDigitalMermaid There aren't that many non-Chinese developed ones. I would look into a Batrium K9 and maybe still use an active balancer with it. I don't know why Andy has never at least showed support for a BMS developed in his own country.
@@davidb.5544 Because Batrium is not panacea either. They had plenty of firmware and hardware issues (and major changes to hardware without apparent benefits), and that at a much higher price point.
I had to deliberately search for your channel this weekend to catch up on videos from 4 months back because I literally haven't seen a notification of ANY new videos in that whole span.
Now youre home, working on the next phases, and I've still got some stuff to binge.
I just saw some of your comments saying people didn't sub and watch much during that time. I think TH-cam handicapped you.
Great to see you kicking ass and occasionally screwing up for our educational benefit. Keep at it. 🎉
Welcome back! I know youtube can be brutal about deciding randomly not to send out notifications, even when you select all videos. I've gotten to the habit of always clicking on the "subscriptions" tab because I don't trust YT. Haha.
it's programming 101 to store values that you do not want to reset when a unit is turned off in NON VOLATILE memory. I seriously question their testing methodology to let something like this happen.
BINGO. How the hell did they miss this flaw?!
They could at least make sure that the memory storing the value stays powered. It's a common functionality in any chip containing RTC functionality. But they could also use EEPROM or the internal program flash to store values (although over time that might kill the memory since they sometimes aren't made for many write cycles, depends on the controller)
And most important don't ever run on unsafe defaults. If the settings are missing somehow, program it to either continue with safe defaults or just stop operating altogether and sound an alarm.
I was just going to say that! Having default values be very low, or even better... Have it default to an inoperational state if a memory loss issue or etc occurs, and only becoming operational after a user accesses the settings and configures it again. Its now 2024, there is NO excuse for nonvolatile memory issues, as you said, its programming 101. Most microcontrollers now have onboard flash storage. I always say programming and writing code is like art, and many do NOT have the touch. Same with GUIs. This is why I like to develop my own, because I know the firmware and all aspects inside and out. @@Ether_Void
I'm curious if they left NVM off as a cost saving measure, or just used a capacitor instead of a battery (again to save cost).
@@john_in_phoenix It's the internal Flash on the microcontroller, so there is nothing extra to add/no need for a supply. Apparently, previous firmware releases for this BMS didn't have the issue, so it's something that was introduced in the newer versions. Should be easy to fix with an update.
Well done Andy for bringing this to everybody's attention. Whilst I totally agree that this new JK BMS is potentially dangerous and that those who are about to buy a BMS should now look at other makes, the old JK BMS has stood out as being the the best on the market (see Andy's previous videos). Unless it is shown that these older JK BMSs have a dangerous fault I will continue to use them. I have three on a land based application which is a much lesser risk than on a boat. For most of us the 5v start up procedure is only a one-off procedure during first start up so is a minor inconvenience.
Older models you've used and tested are probably going to stay fine. The reason I felt I needed to make this video is because you never know what you're actually going to get from JK when you order. Maybe you'll get a good and reliable BMS, maybe you'll get one that could destroy your batteries. It's not a risk I can recommend anyone else take.
Don't overdramatize. Jk BMS are still the best of the budget ones and the speed they fixed this flaw is remarkable. Additionally use a BMV712 as independent last resort solution and something like this won't destroy anything in your system. Yes it should never happen but did...difference they stood up and fixed immediately. Fu.. up always happen, the difference is how and if you even fix it. This compared to many other company incl expensive victron that is not fixing anything or with a huge delay. Victrons 100V range of MPPTs for example the remote steering is hot=on when the remote cable gets disconnected due to any reason. Absolute noGo in EE and a major design flaw they don't fix since 5 years, the phönix shorepower charger 50/1/1 gets over 70 degrees hot but eg cable rating in boats and RV outside the engine room is 60degrees Celsius due to ISO standards...victrons answer was installed in engine room or change cables, 75 degrees is normal..really???...the dealer I bought agreed with me and took it back in the end.
@@chrisr819 How would the BMV712 save a cell from overcharging with this flaw? It can't see the per-cell voltages. As I explained in the video, it's not just this issue, it's a long line of problems with this flaw being just the straw that broke the back. I can understand why people will still choose to use JK, and that's fine. Notice that I didn't say "don't use it", I said "I no longer recommend them". I chose my words carefully.
Every one of us has to decide where our comfort level is, and I have mine. JK crossed it, and this being my channel, I felt I had to say something. I've shown JK BMSes in use in my project for 2 years now. If someone saw that, and figured if I used them, they should use them and then they had a problem, I would feel terrible. So for me, my project, and my risk tolerances, _I_ don't recommend them anymore.
I'm not overdramatizing, I'm speaking my opinion.
Maddie the BMV712 has 2 measure curcuits so you can set it up as midpoint monitoring that means it measures one half so 8 cells 1-8 together versus cell 9-16. So if you have a runner eg 200mV in cell 10 then half 1 is eg 8x3.4V=27.2 versus 7x3.4+3.6v 27,4V and if you set thershold to 150mV then the BMV712 triggers the relay output to shut off. I take a breaker with trip coil, you need a battery fuse anyhow so the BMV712 connects to the trip coil of the breaker and battery disconnected. This is a requirement by ABYC and if you don't have it in US then you won't get insurance or if a claim happens they deny coverage and most likely cancel your boat policy. In an event like JK BMS all cell will go to 4.2V×16 and if your bank has 50mV deviation then the worst that can happen that 3 or 4 cells hit 3.7V while others have 3,55-3.6V and then the HVC voltage of 16x3.65V is reached and BMV trips the breaker. If you have bigger deviation (which you shouldn't have in 1p16s when they are top balanced and normally taken well care by BMS) then the midpoint monitoring trips the breaker. Like this no cell will be damaged. It cannot work on cell level of a single cell but its good enough to protect the bank from desaster events. And again ypu are obliged to have an independent last resort sountion due to ABYC doesn't matter which BMS you run and that for a good reason. With breaker with trip coil you have no additional contactor in your system, just a 2nd method besides a normal fuse has to disconnect it. I mainly use ElectrodacusBMS as i don't like contactors but also no mosfets based BMS as i often use bank as hybrid starter and house and the inrush or surge currents of a motor destroy the mosfets over time, especially if they are not matched and calibrated which for the price JK offers their BMS (as nearly all other chinese mosfet BMS) is simply not possible to do. So if a motor sizes the surge can be easily quadrupled and that kills eg one fet of your 10, BMS still works but with 9 that ahve to carry the current but because unmatched they do 't switch in the exact same moment so 1 or 2 fets have to carry the whole load..and wear til they break. So you have a slow and unrecognizable slow death of the BMS as when eg 5 fets are left it can only carry half the spec. For motor and propulsion you need a current free BMS ELECTRODACUSBMS is perfect but not available in 48V, so more or less contactor based ones like REC are your only correct choice. That just as a side note. I often use the combo in house that i have a main pack mostly from winston cell thats the master pack and eg starter and windlass is connected because this cells can do 8C peak together with an ElectrodacusBMS+BMV712, then parallel battery packs 1p4/8/16S with 280 or 304AH EVE cells+JK BMS+BMV712 to get the desired capacity. So they don't get any huge surge currents. the BMV712 then brings all the battery data into cerbo too.
Thanks for the INFO LOVE your shows !!!!!!!!!!!!
You're welcome. :)
My Seplos BMS’s had some issues at the start, now at almost 3 years and working great, all networked together, enjoying the passive balancing lifestyle - I know Andy wouldn’t approve but it’s never been a issue.
Amazing how many companies do not operate proper change control procedures. So glad this hasn't affected (I assume) your project;; I couldn't live without digimaid and her whirlwind presence :) Thank you for entertaining and uplifting me (and eductating me!!) good luck!
I'm in the clear, and I am sure my BMSes are as safe as they ever have been. The problem is that "as they ever have been" is a decidedly low bar. Really glad you're enjoying the videos, thanks for watching! ^_^
@@TheDigitalMermaid I wouldn't miss your progress (along with the inevitable trials) for the world.
Andy and I had just messaged about the JK BMS about a day before he came out with this and I had said I was about to purchase them. Sure glad it came out before I bit that bullet along with several thousand bucks in cells. You can be certain JK isn't going to have any part in replacing destroyed batteries and not likely their own junk they are producing.
You're absolutely right, NO ONE should be purchasing anything from these people until they prove themselves worthy. This kind of irresponsible nonsense isn't in any form acceptable.
Happy to see that you're helping spread the word! It needs to travel far and wide as quickly as possible!
Thank you!
Glad you avoided spending the money. Getting money back from JK seems as useful as squeezing a rock for blood...
Yeah that's the problem. An ocean and a national border standing between you and them.
I came from the industrial control business. You make careless mistakes like that, people get fired and companies get sued. There's no recourse for these people. It just rolls off their backs like water off a duck. All the while you're left holding the bag.....
Thanks. Was going to connect a new Jkbms to my lifepo4s next week! Not now.
I wish I had another to recommend to you... I don't know what is a good budget BMS now.
Thanks for relaying this info. Great catch
Andy deserves all the credit.
Enjoyed the video. Thanks for sharing
Thanks Maddy. Just remember your a big part of the community as well
Thanks. :)
That was the reason why I didn't buy JK BMS, but bought JBD instead. Yes, Andy is right that JK has a better balancing current, but safety first! If JK would fail my battery, I would burn "only" my house. If you are at sea, you might risk your life!
This is the new world of rapid product and software development and going live. This is pioneered by companies like Tesla - look at all the forums about how they left customers alone with their problems for months without any help. JK is going in the same direction - their products are tempting, really interesting and really competitive - as long as they work. If you have a problem, you end up in the cold with no option to do something. In the end, it's much better to buy inferior and more expensive products that are more stable.
That's one of the reasons why I went with Victron for example, instead of cheaper products...
This is VERY important video, the awareness needs to be spread as this might seriously impact people that have off-grid installations, go for vacation, return back just to find out they killed significant part of their energy storage system :(
Given my use-case in marine, and I suspect a decent fraction of my viewers are also considering electric sailing, this approach is not acceptable. In a Tesla or other car, you can pull over and call for a tow. If you're off-shore, well, good luck! This is why _I_ can't recommend them anymore. If others decide it's worth the risk, well honestly, I get it. I understand the appeal. It's just gone too outside my comfort level.
@@TheDigitalMermaid Honestly, for such mission critical battery, I'd probably go with a pre-built battery like Pylontec or something... Exactly for the reason you've mentioned... But yiu will still have solar and sails when all else fails...
@@rklauco Pre-built batteries are too volumetrically inefficient. Even with bare cells, I'm so space constrained that I am looking everywhere to shave off a few mm here and there. Now, this isn't the case for many other boaters, and for them I absolutely agree with your pov.
@@TheDigitalMermaid Yeah, this aspect is true... The flexibility of open crlls is great.
Thanks Maddy and Andy, for bringing this public! Luckily no harm done to your cells thank god. I just used a simple 25,00 euro hardwired 100amp bms. Because of costs when i was building my rv batterys. On top of that a placed a 1amp active balancer off 6,00 euros. For the last 3 years this config has been working flawless. I also installed a small switching voltmeter per cell block, that i regularly can switch on to check the cells. Very cheap, very lowpower in standby, but it works. Yeah no fancy app controlls etc, but i just want reliability no frills, and couldnt affort it haha.
Using the DIY-BMS system with the batteries for my sailboat. Might be worth a look (it works remarkably well, especially with the built-in 'smart shunt')
I wonder if the old ones do the same thing. Food for through. I have 2 faulty ones down the shed too.
They don't, mine have been powered off for weeks without losing their configs. This is effecting the new BMSes.
i watched Andys video and appreciate yours, from his vides I understand that its only the new BMS for inverters?
Ya, seems like it's only the new model. However, as I said in the video, this is not a new thing with them. They seem to slap together crap and throw out to clients with clearly minimal testing or due diligence.
That's also what I get from his video... Recently installed a regular 4-8S JK bms and this specific problem didn't occur (turned off for a week)
Still, it is unreasonable to expect everyone to test every safety feature of every BMS, there needs to be some trust...
@@TheDigitalMermaid I have two of the newer style JK BMS because of Andy, they are the 7s-20s with the push button to start them, I have had zero issues with them, and they balance so much better than the previous BMS I have had
Wow, thanks. I haven't watched Andy in a while, gotta get over there and see if my several JK's are affected. Mine are the red Smart BMS's with cooling fan, 250A rated. Really gonna suck if they aren't trustworthy, I have probably close to $1000 in these.
Those sound like Daly BMSes.
LOL, you are correct. Don't know what made me think I had JK's. @@TheDigitalMermaid
This issue only impacts the new JK Inverter BMS, not the JK BMS we have all been using for years without issue. The one you’re holding in the video is not problematic.
I spoke to that in the video. My frustration is that this is a continuation in a trend of sloppy and hap-hazard approach to product dev. This is a safety device, and they clearly do the very minimal testing and QA.
@@TheDigitalMermaid I declined to watch the rest of your video as you were spreading false information in the first 10 seconds.
@@cgutowski471 Do you have evidence that it's false information? The idea that Andy would be wrong about this is quite the statement to make...
@@cgutowski471She is very clear in the video that it doesn’t affect all of their products. But she goes over other issues that do impact other products and that demonstrate carelessness by the manufacturer.
I think she makes some very good points. What specifically do you think is factually inaccurate?
@@TheDigitalMermaid you made a very broad statement vs being specific to the new BMS that has issues.
100% with you. Any competent sofware team would not let a thing like this slip past QA without expecting heads to roll. It's evidence that there is no legitimate QA.
Exactly.
China QA is a bit like China amps, they are not the same as in the west... China is only interested in selling things and have no issues selling components that could be dangerous. China isn't the only country that uses its users as unpaid beta testers, Apple, Ubiquiti Microsoft etc have been doing it for years...
Thank you .
No prob.
I just bought 5 of the new 200 amp JK-BMS's because of a off grid garage recent video recommending them, I hope they fix it soon.
I hope so, too.
I watched Andy's video before you and IMMEDIATELY went in and checked my new JM Inverter BMS based battery which was powered off for 5 days and it still has correct settings. I have not put my new battery boxes into production so can keep an eye if this issue manifests.
Aye, I think until there's a fix or replacement, making it a habit to immediately check the configs on power up is very wise.
Then your BMS is still on Firmware 14.03 - 14.10. These versions don't have this issue.
@@TheDigitalMermaid That same can and should be applied to anything amd everything manufactured with memory that can be lost or currupted not just this BMS, its foolhardy to think otherwise.
Wow! OMG! By sheer chance I’m using the JKbms on nmc li-ion, so I’ll get away with it. But in the future I’ll be on lifepo4. I really hope this gets fixed by the time I upgrade!
It should be clearly stated that it is affecting the new ones only, and should also be made clear how to check the settings, so users can learn how to check after switched off for a longer period.
This was just too much "breaking news" for me. While still a big problem, it is important not to scare everyone, and to convey useful information in a situation like this.
Still a happy follower :) This didn't sit well with me though.
I am also not sure how it will improve safety to remove the JK BMS's that you know does NOT have this flaw. It is not like other manufactureres are immune to making errors....
Pinned comment, and I spoke to it in the video, this is not an isolated incident. This is just an egregious failure in a long line of bad design decisions, lack of QA, etc. Look at Andy having two different JK BMSes blowing FETs on over-current tests. Look at how they change, fundamentally, things like supported configurations or available ports under the same model number. As I said in another reply; this is the straw that broke the camel's back, not the whole bale.
Thanks for bring this to our attention! You and Andy have a bit of influence so it's good to get the word out. Not to take away for your video, you/we are coming up against a manufacture cultural issue. Not specifically a Chinese culture issue. Their hasn't been much feedback applied to these manufactures and in China there isn't the robust certification compliance that there are in EU and Canada/US, so their only feedback in monetary and it's been a boom for most of these products coming out of China. Definity caveat emptor, and with exception of very very few, they will not take care of the customer.
For me, my biggest fear is that someone chooses to use JK for their build, because they saw me use it in my build, and the BMS they get damages their equipment. I would feel horrible. Even if JK changed completely their business and development practices, it would take a long time to regain trust, imo.
One military organization had devices to use a GPS receiver, a laser rangefinder and a compass to be able to generate coordinates for a distant target, to pass to friendly aircraft to attack. The batteries in this device died, were replaced, and on reboot the device displayed its _own_ coordinates. Those were passed to the aircraft and it was only because they were using a practice weapon that they got a lesson instead of casualties. These mistakes matter.
O_O - holy sh!t
Not defending the bms flaws, but wanted to chip in something weird I've found online. 3.65V is not really the limit voltage for LiFePO4, it's more of a recommendation. It seems that just like Li-ion, 4.2V is the limit where dangerous things start happening. It's just that for LiFePO4, charging them to 4.2V is just too much of a sacrifice to longevity for too little gain (same reason we choose to not even charge to 3.65V).
The specs for the batteries say to not charge over 3.65v, so this is the limit as far as I'm concerned. Above that might work to some degree? It's not worth the risk though... I've seen what happened to overcharged LFP batteries...
Best setting to use for LiFePO4 is 3.55v and best setting for Lithium Ion is 4.15v
I have been using those max limits for years with no issues.
I charge it to 3.54 v per cell
There's a video on TH-cam showing what happens if you overcharge just to 3.85V. The entire cell becomes seriously bloated.
If you charge to 4.2V you will have a bloated cell (and it will bloat well before getting to that voltage).
Something I don't think was mentioned...
All the solar charge controllers etc.. and my new (bought a month ago) mains 7 stage battery charger..
When first used or disconnected etc for long enough..
ALL revert to the LOWEST ie voltage setting.. or AGM rather than flooded Lead Acid etc.
NOT HIGHEST!
Exactly, where the hell are the fail-safes in these things?!
To all the people saying 4.2v is fine for lithium iron phosphate. Well, it used to be in the early days, because they used very similar electrolyte composition to the lithium ion cobalt nickel manganese type. For the last 10 to 15 years though, the electrolyte has changed to make the cells last longer and the new electrolyte has a smaller ‘voltage window’. This means that it can dissociate into gaseous byproducts if the voltage goes too high. As far as I am aware, most modern LiFePO4 cells use an electrolyte that will begin gassing at around 3.8V, and will create uncontrollable gassing above 4V. So no, it’s definitely NOT FINE to allow them to charge to 4.2V, and will most likely result in high gas pressure, bloating and rupture. Several people have made videos and posts online showing this happening. Some cells bloat like a balloon at 3.85V. As always, read the datasheet that comes with the cells, and respect that maximum charge voltage! If you can’t get a datasheet for a cell with proven tested parameters, don’t buy them!
It's not the voltage, per se, but the heat generated above 3.2V. Basically, everything above 3.2v is just making the cell hot, not getting stored as usable charge. Don't believe me? Test it for yourself, cycle test 3.2 to 3.6 in 0.05v steps and see how much more power you get back out each step.
saw you first and then andy... comented to him too that i already ordered one 100amps.. you both might saved my upcoming cells .. if they re good!!! . still hope its an easy fix like the bad easy failure they shipped to me 40 hours ago
I'm glad to hear you hadn't put it into production yet.
@@TheDigitalMermaidthanks to the whole youtubers community and the others that trying to help us all with all those alarming videos... !!!!! i have my tons of troubles with other brands so i wont judge yet this one ... didnt try any yet!!!! . you all keep the beat.
I race rc cars and one of my classes used a one cell lipo. The charger can auto set to NiMh and cause the battery to swell explode and catch fire. When you said it switches to metal I was shocked that something like that is not a first order safety protocol in the programming.
When you're building a device SPECIFICALLY designed to protect your batteries and your equipment, something as simple as ensuring the settings survive being turned off is pretty damn basic. Also, where's the fail-safe? Why not make the default configuration one that sets the lowest max charge and the highest min discharge?
i toasted a 4 cell lipo and the ensuing fire from such a small battery was pretty scary.I DO NOT CHARGE THEM IN MY BASEMENT ANYMORE.
@@davidkettell6236 Li-po scares the hell out of me...
When I was a teenager {way back in the 70's} my first car was a 62 VW. I learned quickly that I needed to bring the VIN number when getting parts because VW would make changes throughout the year.
That would drive me insane...
One thing to point out is that LiFePO4 cells won't generally bloat with the risk of venting and fire until they are well over 4.2v. 4.2v used to be the charging cutoff for LiFePO4 back in the early days before they recognized that they didn't need to charge that high and the cells last more cycles when not charged to 4.2v regularly. Even the 3.55v, 3.65v 3.7v charging limits are all just going off best practices to be easier on the cells and allow them to last longer with less chance of electrolyte degradation.
That being said, I remember when some of the DALY BMSes had a similar issue and that was why I stuck with ANT-BMS at the time and still use them at this point.
The actual max usable cycle range of a LiFePO4 cell is 2.0v-4.2v but the majority of manufacturers stick to 2.5-3.65or3.7v since it is far easier on the cells and so little capacity is available outside the range we use today. That is actually one of the main reasons why LiFePO4 offers so many more cycles than NMC and other similar variants.
I am interested to see what JK BMS decides to do since I still consider them to be very new to the scene only being around for a few years at this point whereas I have been using the ANT-BMS stuff for about 6 years on.
They do need to figure this out or a lot of manufactured batteries that are currently being manufactured will be running outside of manufacture spec as soon as customers turn them on for the first time and some of the vendors don't even let the customer change settings.
Interesting about the voltage range... it would help explain why the cell I had that dropped to 2.0v doesn't appear to be damaged. That said, the EVE specs say to maintain between 2.5 ~ 3.65v, so that's what I have to recommend others do of course. Interestingly, I've seen overcharged batteries that puffed up like marshmellows... I wonder now what voltage they actually hit.
@@TheDigitalMermaid Yes, when LFP first hit the market, recommended charge and discharge voltages were very different. Many early "hardware" BMS (not programmable) used what would be very out of spec voltages today. The more reasonable voltages manufacturers spec today is the primary reason they are rated for so many cycles. Of course minor tweaks to chemistry are also a part.
@@TheDigitalMermaid I had a charger fail once and it took a few cells to a bit over 5.5 volts and they puffed up after being held at that voltage for several hours but surprisingly they never vented. It had to be close but LiFePO4 is surprisingly tolerant to abuse even when they vent it usually doesn't result in fire but we all try to keep the chances that a cell ever gets overcharged to a minimum where possible.
When I noticed the cells I discharged them down to below 1v and discarded them the next time I brought cells for recycling.
I definitely understand the need to follow the manufacturers recommendations and make the same recommendations. I too follow similar charging parameters to what Andy at offgrid garage is doing since it makes since to not abuse the cells any more than necessary.
5.5v!? Good lord...
I saw that video this morning, thanks for getting the word out,do you think a 3rd party could reflash this to take the dangerous bug out? Unfortunately after watching Andy's videos I ordered one yesterday,so not sure what I am going to do now
See if you can cancel the order, or return it? There's a chance this bug can be fixed with a firmware update, so long as it's not a problem like having no non-volatile RAM on the system. The bigger problem in my mind is the proven track record of being careless, in all the senses of the word. I just don't trust them anymore.
Question... what breaker or fuse (model would be appreciated) should I use with these BMS's
Less about the BMS, as I put the fuse immediately after the main pack positive, but any T-Class fuse rated higher than your expected loads will work. Personally, I like (and accidentally tested) A LittleFuse JLLN-200.
the comment i received from manufacturer "Hello, please do not close BMS. Closing BMS will restore the default settings regardless of what it is.
This is factory set.
You can choose to use the sleep mode in the app, which will not reset your parameter settings." yeah..
Holy shit, if that's true (and it would be in-line with my experiences with them), then they're making it even worse. None of their previous BMSes reset to default on reboot. The idea that they could make a change like that and hand-wave it as being normal is ... wow. Could I ask you to email me? Address is in the about page, with a copy of that response?
Hi
good decision , with new cells you don't need an active balancer !
I can recommend the BMS2 from Seplos .
I also use BMS3 but I'm not completely convinced yet.
The 16s flagship from JDB might also come into question, if it is as good as the 4s from JBD in my OstBlockBatterie, then it is a real alternative.
I'll be testing the 16s soon.
very hard :-)
Ps.: and with Seplos you don't have to beg for a guarantee
I'm going to need some time to think about what I'll recommend, I've not paid close attention to the BMS world in the last year.
@@TheDigitalMermaid. maybe focus on the important things, voltage readout accuracy, SOC determination, accuracy, failure frequency, connection to your monitoring
..The Bms should be as simple and robust as possible for your boat
... waterproof could not hurt (although :-) the seplos is already running 2years on a racing boat in Bulgaria)
....less important , bluetooth, active balancer
Ps.: you will know how to proceed, the second
try will be a hit :-)
all the best :-)
Ευχαριστούμε!
Σας ευχαριστώ πολύ ευγενικά, εκτιμώ πραγματικά την υποστήριξη! ^_^
My old jk bms discharge switch is no more working, even it is off my bms is still passing the current, due to that my inverter is draining the battery to the lowest volt...😢
Well every lithium installation with or without BMS require due to ISO for Europe (mobile lithium usage) and ABYC (American boating and yachting organisation) an independent last resort solution. The standard one meanwhile used in RV and boats is to use a battery monitor with an remote relay. I always use the victron BMV712 that does LCV and HCV and with midpoint monitoring also cell deviation vut off when deviation between packs 1+2 compared to 3+4 is above 0.5V. BMS is at 3,55V overvoltage and 2.8V LCV and victron BMV712 at 4x3,65V=14,6v and off at 4x2,5V=10V. So if BMS fails due to any reason like eg here with wrong default after storage the BMV would prevent from overcharge or deep discharge. I highly recommend this in EVERY lithium installation. saved my ass and big junks of money already several times plus the monitoring the BMV712 provides is also very helpful. Especially in combination with victron cerbo if your BMS is not integrated yet or as eg total of several BMS. I always use one BMV712 per battery and BMS also in multiple batteries in parallel and one for the system overall as last resort. For eg winston cells you can skip a BMS in 4 or 8S just with a BMV712 doing LVC and HVC and deviation and with saving on BMS the winstons are not that expensive anymore compared to EVE. The BMV712 is super reliable, don't know one in 20years dealing with lifepo4 that failed.
Victron's approach to their BMS and LFP setup is pretty unique, and only viable in their environment where _everything_ is talking to each other. They wanted to be able to handle the same kind of inrush currents like starting diesel engines, making them actually drop-in replacements for LA (LFA/AGM/etc). The problem with externally disconnecting charging in parallel battery installs, like I have, is that if one battery needs to stop charging, and turns off the charge from the Quattro/Multiplus/etc, then all batteries stop charging. If you have a contactor-based BMS, then the BMS can ask the specific battery's contactor to open, but contactors have their own trade-offs.
There's no universal solution, but in the end you somehow must have something that can see the per-cell voltage and cut off charging/discharge reliably. I know some people run LFP without a BMS, and that's their choice, but it's not one I'd ever endorse.
I'm a bit surprised you've stuck with JK so long. But as far as chinese jellybeans go, they're pretty much all like that. We're assuming JK is actually making these things and not sticking their name on stuff. (which is very common!) Sounds like they shipped LiPo firmware with LFP settings, and when the internal cap/battery discharges, it loads the LiPo settings. That's a nasty error anyone should've known would not fly.
There's not been many options to replace them. Even now, I'm not sure what I'll switch to. I've not hidden my frustrations at JK, this was just the last straw for me.
I am sorry you both are having problems. Chin I believe will be a future military problem and with that said consider future problems . Fuses both master and branch are necessary . Future and current supply chains need to be changed immediately. Thanks for all your content. Out smarting electrons 😳🤔 well you know the rest of the story .❤❤❤
Just bought one a month ago I don't plan on turning it off so should not lose it's setting hopefully by the time I'm ready to use it full time they have a firmware upgrade.
I hope so. In the meantime, if the BMS does reboot for any reason, be sure to check the config when you turn it back on.
@@TheDigitalMermaid Wonder if there is a way to setup an alert with Solar Assistant. Run it off a Raspberry Pi. I have so much work to do haven't gotten to that point .
I have the old version with the update. Up to now they are ok. I am happy that i dont have the new version 😊. I have seen also the video of Andy. Also was a bit shocked to.
I'm sure my old ones are fine, too, but as I said it's more about recommending them. You just can't know what you'll get, even when you order a specific model number. That's madness to me.
Any changes in hardware yet? What I understood from Andy is that data is stored in RAM instead of Flash.
i was temped to buy five of them for my 80 Cells (5 battery strings) but now I am nit shure anymore.
I have to agree, this sort of flaw is not acceptable on any level. And I've even had 3 of the BMS's like you have, to find that running more than 1 in parallel caused issues which was strange. I too had lack lustre responses from JK. While their hardware is ok, it's nothing without the software to back it up. Hopefully someone gets kicked in the butt for this problem, and they have a change of procedure, but we won't hold our breath.
Another commentor quoted a reply from JK saying "well, just don't turn it off". If that's a genuine reply, and it would be in-line with how they've handled my complaints in the past, then no one is getting kicked in the butt. My core argument is the company culture over there... Sure, when you get the BMS with the features you expected and that variant performs as expected, it's good for the money. The problem is their willingness to change shit on the fly, without model number variation, and pass the buck to the end user for getting caught out. JK clearly does not give one damn about the users of their _safety_ devices.
What or how do I find out what BMS I have if I purchase the battery from seller without opening the case and voiding warranty?
Email the manufacturer. Almost certainly though it's not one of the new ones.
Provision a system with the new JK BMS. Ensure the settings are correct and then use the new system. As long as you don't shut it down for several months, then bring it back online without confirming the settings, there is no danger. It's kind of like winterizing a boat or an RV, then getting it ready to use again the following spring. I'm not saying it's a GOOD thing their newest BMS boards don't retain those settings. It really is a problem, but for anyone who has bought one, it's not like you can't use them, especially if you're like 99% of people who but a system to work...and it just keeps working.
It's not "several months", it's "a few days". The idea that a restart of the BMS, regardless of the duration, could result in the damage or destruction of your cells is not a condition I would ever be comfortable with.
@@TheDigitalMermaid - The settings shouldn't change, that's for sure. I'm glad I've stuck with Overkill Solar and the JBD BMS's, thus far.
Andy's next video says they fixed the issue within a couple of days. I understand the weight of this issue so I am not discounting it as 'no bid deal'. I just think that their response is worth mentioning. I have not gone through more of his videos, so I don't know if there are more updates other than the one update he posted a few days after his initial video. I am still on the fence about JK. The price seems good, but then I wonder about this sort of thing.
JK is the best of the "budget" brands still, if you really want the active balancing (I do), but I don't actively recommend them anymore. I don't have a better alternative, I suppose JBD I would consider safer, but then I don't know them well enough to say if that's actually true (the JKs being popular means they get a lot more scrutiny). I'm working on an alternative, but it'll be a lot more $$. =/
I watched andys video earlier and was going to post about it but the nap had my name on it. pretty spooky stuff. we will see what happens at JK.
JKs response would need to be 180 degrees different from how they normally react to problems if they want _any_ chance to recover trust, in my humble opinion.
If you’re hybrid inverter or solar charge controller is not set to max voltage charging, your cells are good quality and balanced, you’d never reach 4.2V anyway. I love JK. 250kWh and all doing great.
The solar MMPT will protect the pack from pack over voltage, but the BMS does individual cell protection, the BMS gives some redundancy.
Wonder if there is a cell monitor/alarm device that could give redundancy in case of BMS failure? Not that hard to design, should be fairly inexpensive, cheap if it protects the invest in batteries.
Also live in Ontario, looking for cells, did you have them delivered here, vender you recommend?
The MPPT, and any external charger, can't see the internal cell voltage. Say you set the charger to stop charging at (3.55 x 16 =) 56.8v, and 15 of the cells were at 3.3v average (3.3 x 15 =) 49.5v, leaving 7.3v for a single high cell to go up to. The issue with the BMS resetting the high voltage disconnect would still occur. If you've got something monitoring per-cell, well you have a BMS of some sort. :)
I ordered the cells from QSO (Qishou) on Alibaba and had them shipped from China, took about 2.5~3 months iirc. Cheers, neighbour!
@@TheDigitalMermaid
Thanks i look at the failure modes of a BMS, worst case would be a shorted Hexfet, it that happens your expensive battery's have no cell level protection. I always like belt and braces protection.
As far as I know, this problem is only with the new BMS. The other ones are ok. I disconnected my 3 for 7 days. After reconnecting, all settings were there. No problem at all.
This specific issue is limited to the new inverter models, I believe.
Thanks for warning us. I got the older black one (200A). Did not connect it yet.. I'll try how it goes.
A good charger should pervent that from happening also, right?
Chances are, the older BMSes like I have are fine, just test carefully when you build. My complaints and frustration are at how JK operates as a business, with this latest flaw being just the worst example.
Hi,
If you use a charger that does not communicate with the bms....is there still a problem? (So a dumb blue victron charger set for lithium iron)
Other Scenario
A jk bms, with no charger attached cannot start actively balancing to a value higher than the highest charged cell... as in, can it generate the 4.2v without a charger attached?
Bought 2 older ones last year and was about to finally use them 🤔
@@gerardhealy764 Yes there's a problem... A charger can't see the internal per-cell voltage, so if you set the charger to 57.6v (3.6v/cell), but one cell is higher than the rest, you could have a situation where 15 cells are averaging 3.3v (49.5v) and that leaves 8.1v that could go into the cell. So if the BMS was set to not cut off until 4.2 volts, you could damage a few cells and the charger would still see under 57.6v.
@@TheDigitalMermaid thanks for that. I was left with question in my head even after watching andy's video. A hint on the alternatives you are considering would be great (unless it would spoil things). Keep up the good work.
P.s. I just blew the capacitor bank on my kweld spot welder by not restricting the inrush current, was too impatient to wait for my new bench power supply too arrive and used a battery!.. so learning the hard way here too!
@@gerardhealy764 Learning the hard way is the DIY life! Haha. Sorry, I can't really speak about the alternative plans just yet.
Andy has also updated that this issue has been resolved by JK in an updated software patch. So, in approx 24 hours from his video coming out, it's sorted.
This particular issue has been resolved, yes. I am happy for people who own this model. The general argument I made about JK's business practices stands though.
@TheDigitalMermaid agreed. I'm not advocating for the sorely lacking QA and customer treatment attitude either. Unless you are Andy or similar with their ear, it's a great big middle finger. I think the hard pill to swallow is this. They manufacture for their domestic market.
Though Andy has now pointed out that 14.7 will fix the flaw in the new Inverter BMS. I will say that I support your position. In a mission critical (life dependent) environment, you don't take chances and trust is easily lost. If my BMS gives my some hiccups in my setup, I can fix it without risk to life, you can not. I have both the new inverter BMS on 14.5 and a couple of the older JK 16s BMS (with a start button that works). I will stick with mine though.
This is just it; It's the culture of how they release products and how they (obviously don't) care about the customers. I'm glad this is fixed for people who own them, but that doesn't rehabilitate their reputation in my mind.
@@TheDigitalMermaid I am a software engineer also. I would never release a software patch or program change without a full unit test and QA cycle. I don't like "call back"
HOWdy T-D-M, ...
WOWSER = THANKS for the WARNING !
yep I do watch Andy's CHANNEL but YOURS always comes 1st - Thank ANDY too
Thanks again for the HEADS-UP & I'm so glad that YOU suffered no extreme
Thanks, and this is why I made the video. I know people may have chosen JK because I use them, and I would feel horrible if one of my viewers lost their batteries because of their BMSes.
That sucks. I was already subscribed to Andy, he makes good videos. I try to never make the BMS protect a battery. The more control I have, the better.
Having your charger set to max out below the max voltage of the BMS is always wise. That said, external chargers have no view inside the batteries so they can't see imbalances in the cells. It's entirely possible that even if you the charger set to max 57.6v, if the cell average is 3.3v, and one cell goes high from imbalance, that's still 8.1v that can go into that one high cell. A BMS plays a cruicial role in safety, imo.
The part from the 4s-8s i got one around the same time and in the video where you test the 4s i don't think you connected the last positive wire to battery+, the last wire is the power supply wire for the bms, could you retest this please i am really intresed if it will work, you just need main battery negative on the ballance wires and on the B- connector, and power on the most positve wire as this is the powersupply lead. So you don't even have to build a pack to test this just connect it to a 12v psu or one of those lead acid batteries.
I did, it didn't work. It was a fairly narrow window where some BMSes were sold under that model number without 4S support.
@@TheDigitalMermaid Thank you for testing it, they really should held accountable, problem is they will only do something about it if it start hurting there business. SO the only thing we can do is not promote them and don't ever buy there products again. Real bummer tough i finally found something good.
Sorry for your bad luck, but we have had many our customers complain about the 8s version as well. They also have very poor Ram / Over volt balancing where they ram voltages of 4-5v into a cell to try to get it to balance faster. This is not good as incoming voltage should never exceed the full charge voltage of a cell such as 3.65v. You can verify this with a volt meter to see the incoming voltage. The same is if you up the voltage of a charger to get a cell to charge faster.
If I recall correctly, their supercaps they use to store the power during balancing are under-rated as well. I would need to crack a BMS open to confirm my memory, but ya. They're cheap for a reason...
It's definitely a failure mode worth considering. Failing safe should be one of the leading principles.
Absolutely
what kind of BMS should I get ? I'm about to billed a Wheelchair ? I have JK K-B2A24S20P.
I wish I had a good recommendation... I suppose in a pinch? The JBD (Overkill Solar) that I used early on in the build (I gave to Andy who did more testing on it). Do a good top-balance before you build the battery and you should be OK. Otherwise, consider adding an external active balancer. If you're building a 12v battery, there's a variant Andy experimented with a few months ago that is a rebadged JBD with direct support for an external balancer, that one looked good, though I've not played with it myself yet.
Huge flaw but
Your charger will not keep changing after it reaches its maximum voltage
So at worst you overcharge a few cells that are out of ballance
But you need to babysit these bms's to notice the change
"at worst you overcharge a few cells" - which could destroy them. I'm going to do a follow-up video in the next day or two to address the idea that the charger can prevent damage, because I don't think that is a safe assumption. If you set you charge voltage to 57.6v, 3.6v/cell, and all your cells are perfectly balanced than sure, that would be fine. However, we know how incredibly flat the voltage curve is on LFP... If you had one cell above the others, and the average across 15 was still 3.3v, that's 49.5v, leaving 8.1v overhead for that 16th cell. If the BMS doesn't cut the charge, the cell will be destroyed in this scenario. This 4.2v flaw would easily lead to the destruction of a couple cells.
Oh WOHW! YOU are blowing things up. Andy is talking about the INVERTER JK-BMS. You do NOT use the inverter BMS's so you could NOT blow up your investments!
I'm guessing you didn't watch the whole video, I explained why my frustration goes beyond this single bug.
@@TheDigitalMermaidYou should not be that frustrated. JiKong still builds the best BMS's imo. But...... another safety hazzard has been released. The "hacked" master password.
also booting some BMS first time is simply conect short B- to B-P bms
I can add a story here as well. I bought a version with a 2A balancer. When opening it, the PCB was labelled with 1A. I sent a question about that to them. After asking where I got that from I didn't get any answers anymore.
Now this BMS is throwing errors about the current sensor. Will probably switch to something else.
Do you have pictures or records of this? If so, would you mind sharing them with me? My email is under the About tab.
Here is another point of view… (This counts for all influencers.)
You, and also Andy, and many many more are promoting stuff (in this case JK bms’es) without proper testing.
A lot viewers, including myself, have bought things because TH-camrs (in other words; people with huge influencing capabilities), where promoting stuff. In these cases faulty stuff.
You could also say that nobody of the influencers ever considered proper testing -before- promoting it to their communities.
Anyway, I’m glad you and Andy made this go public.
To be clear, there's a difference between actively promoting and showing what we use. I've never taken a sponsorship and I've pair retail prices for everything I've bought. I always explain why I chose what I chose, and I've often spoken to the issues about JK that annoy me (as I state in this video as well).
As for proper testing, speaking for myself, my channel has never been a review channel, it's a channel to document my project and my process of learning. I've never claimed to be an expert .That all said, I know some people will choose JK because I used them, and that's exactly why I felt compelled to make this video.
Great, my inverter style is in customs. Luckily it is just to try the new style, have enough of the normal style for my 3 48v and 24v pack.
If you can't return them, then fingers crossed they release a firmware to fix this. At the very least, make it a habit to check you settings any time the BMS is turned on after being powered off.
I have programmed the "advanced settings" ... in 2 JK BMSs (from late 2023 purchase /recent hardware); and I think they are the best choice in market place for my BMS. ... You seem to have criticism of the default LiFePO4 BMS settings ??? that are fully adjustable. Please correct me if I am missing something. If advance settings changed at a boat doc, I wonder if you created a unique new password? Re setting charge voltage: that is done at charger level, not the BMS protection and cut off triggers level. So ... 🤔
The problem is that JK changes features and capabilities without changing model numbers (hence why I had two BMSes that were 24v only). They add and remove ports on a BMS without changing the model number. This flaw could have been mitigated with fail-safe defaults, and prevented with even basic QA testing. Andy has had FETs blow at least twice because the JK BMS didn't cut the FETs fast enough in over current tests... It's clear to me now that JK can't be trusted, despite having some versions and releases that are OK overall.
@@TheDigitalMermaid ... Re Andy's FET blowouts w smoke on JK BMS. I remember seeing 300 second delay for an over amp limits cutoff trigger (possibly a ridiculous default setting? / vs my choosing a way safer 15 seconds for trigger timing, ... that is easily adjustable.
@@CaptBill69 The time to disconnect wasn't the problem, the timing of the FET shutdown was. There are many FETs in parallel carrying the load, if one shuts down slower than the others, then the load of the shut-down FETs starts to go through the last FET that is still up, overloading and popping it. It's poor /sloppy design.
I've used & installed over 50 of the older JKBKS version, 1 and only 1 had an issue.
Maddie, chillax have a PUFFY and consider your reaction (overly so) reflect on the "Mass Bash you just did, seriously.
Last Point, over a WEEK AGO the first reports of the RESET TO DEFAULT ISSUE on DIYSolar and another Solar Forum as well....
The devil's lettuce gives me a vicious headache, I'll pass. I'm glad you've had good luck with your BMSes. Mine, the ones I got that actually worked, have kept working fine. I thought I explained pretty clearly in the video why I am upset, and it's not because of this one flaw. It's a long pattern of behaviour from JK that upsets me. If someone bought JK BMSes because they saw I use them, and they got unlucky and got a bad BMS and it destroyed their cells, I would feel horrible. I had to go on the record with my concerns.
oh wow, this is just the kind of thing that scares me, purchasing the wrong kit to work with other equipment, what a nightmare
i guess this is an example of a company producing something and marketing it, find all the mistakes and learn from it and produce an upgrade, a white box black box design - personally i don't agree with it, i understand why its done, but really adequate testing should be done, equipment needs certifying, things in Europe have to be marked / certified before being aloud to go in the market place - like toys with pins in them, not aloud, too dangerous, so no go, all stores have to abide by law that these things are checked
i remember a consumer program many years ago, who would bring such things to the publics attention, now-a-days we have the internet and can learn pretty much instantly whether something is a good purchase or not, so a big shout out to that off grid garage fella for pointing out this matter, as you say, what if had blown all your batteries and other kit
this is one of the many reasons why i wish people would write a book on the subject, explaining what they have, kit wise, how it works, how its wired up, a complete idiots guide to doing it correctly, the big issue is, is that the kit is soon obsolete, superseded and with it prices change
i follow a chap here in the UK, who lives off grid, he used to install lifts, so has a good understanding of electrics and power (i don't) he has lived off grid for around 3-4 years, got his house, a workshop and now building larger barn, his power comes from both solar and wind turbine as well as having his own water supply from a stream, now runs a water mill giving him an abundance of power at his fingertips, all free, apart from the kit and storage batteries which like anything over time will wear out and need upgrading, he has don this already a few times, but never the less, its interesting to watch, but i would pay handsomely for a book - there are many examples out there possibly, but finding something that you trust and have seen working such as on here at You Tube can be a bit bewildering
I can understand the "build quick, break, iterate" model, but that has to be kept in-house. Anything sold to end users, particularly safety equipment, needs to be well tested. Clearly, the JKs are not.
Writing a book would be a great idea, but the real problem there is just how damn fast things changed... By the time you got through the publishing process, it'd be out of date.
@@TheDigitalMermaid It's interesting that the JK "inverter" BMS version is now listed as "pre-order" only now even though they have already sold many and people have them! What are they going to do change hardware on the fly again to fix issues like they've done with the old JK BMSs? 🙃
@@davidb.5544 Almost certainly that's exactly what they'll do. If they screwed up working with Andy, I have to believe they Just. Don't. Care.
Although this is a serious safety flaw, I'm not certain but dont you need to enable the charge and discharge on first use?
It's always a good idea to regularly check your parameters. Hopefully JK take this seriously and use some kind of non volatile storage for the bms settings and at least have the default settings set so any connected cell chemistry will remain in safe voltage limits
In my experience, no, the BMS charges out of the box. I've not specifically tested this, but I don't remember ever having to do it, either.
It's a firm wear problem its defaulting to Lithium Ion - It should have been spotted at the development stage - Fortunately most folks solar or grid chargers will prevent overcharging as long as the cells are in balance - it needs resolving ASAP as the BMS is the primary safety system
The problem is that the voltage curve is so flat, that it doesn't take much for one cell to spike it's voltage before the others. Unless the BMS can request the chargers to stop, I don't think it's safe to trust the chargers to save your cells. I think I'll do a follow-up video, because this has come up a lot.
I hear your frustration and understand it, but, what gets to me is that the BMS does not just reset the values set, it resets the entire chemistry.
This is the issue, for those using liion it is not to much of an issue, but for the rest of us that use LFP, this is one hell of a problem.
Again, one should not just trust the BMS, this is where knowing your system and having someone who knows how to program these systems comes into play, if your mppts are correctly set up you should not be able to over charge your batteries at all.
If one fails, and you have not checked your system then, this is very much able to happen, as I said on Andy's channel, how this was able to get out into the wild without being tested......
That is the question, at least, it is a simple coding fix, but, IMO, should never have happened.
This should be the one part of the coding that should always be front and center, clearly marked, and should not be changed.
MPPTs, chargers, etc can't see the internal voltages of the cells. I wrote this as a reply elsewhere, but consider a case where you decide you want to cut charging at 3.6v/cell, so you set your chargers to stop charging at 57.6v. If you have one high cell, given how flat the charge curve voltage is, you could have an average of 3.3v/cell for 15 of the cells, and the 16th would be allowed to climb all the way to 8.1v (15 x 3.3 = 49.5v). You need the BMS to be able to cut charging...
More though, when combined with all the other problems I've had, and Andy has had, it's clear that JK just doesn't give a shit.
I thought your 1st BMSs were Overkill Solar that you gave to Andy? I've been running (2) 16s flawlessly for almost 2 years. I haven't turned my system off once.
Good memory, it was. I was lured in by the JK's 150A over the OKS's 100A and the 2A active balancer. The OKS / JBD BMS itself never gave me a hassle.
can you show us what are the "good" one's and the bad new one's ?
I have limited experience with OverKillSolar (JBD), and they seemed OK (though passive balancing). My experience is otherwise with JK, which I am not happy with. I don't have an alternative I can recommend yet.
I bought my JK BMS through an EU importer in Holland, thus having them confirm I got the right product knowing the 12V /24V models confusion etc. My delivery was a success. These BMS are apparently not supposed to be shutdown for prolonged times, only for allowing service.
When buying unspecified products as end users we have to confirm ourself the product is safe and fit for our intended use.
REC bms has active balance models that by my opinion are the most stable and reliable piece of equipment. Though expensive. JK is the best budget version .
Until now, I recommended JK for being the best feature set for the price. I don't know what I'll recommend now, I have to think about it. None of the cheaper BMSes seem good to me... I had the get this out though, I couldn't bear someone buying JK because they saw me using them, and then have a disaster happen.
@@TheDigitalMermaid I still recommend the JK - there is nothing else out there that comes closes especially at this price point for DIY applications. It's part of the 'cost' of DIY: cheap, available, bugfree: pick two. And even if you go for available+bugfree, your options are, maybe, one (REC).
@@upnorthandpersonalDude, I get it, I do. It's why I held on and forgave so many of their... Quirks. This though... Such a simple thing to check and something that would destroy a person's batteries... It's just not worth it. Where are their fail safes? Maybe I'll try JBD with an external balancer? I need to think about this more...
@@TheDigitalMermaid JBD doesn't have a function to power itself off after a cell low voltage event for example, so it can completely drain a battery well below safe cell voltage levels.
@@upnorthandpersonal oh FFS, ok, that's a deal breaker. I honestly don't know what to do next... I'll figure something out
I have used those 4-8S 200A with 2A balancer what have been ok and i still do checks.
One 8-20S 100A BMS on my 8S Lifepo4 UPS battery.
I dont lose faith in the manufacturer in question because they have product with model flaw, its good that flaw have been found and hopefully people will get their money back or new product with the flaw fixed on those new inverters BMSes.
You can also find flaws in cars, which can be dangerous, but I have not left the brand because of those, it is important that the manufacturers fix flaws and replace those, dont kwno what JK plans to do so remain to be seen.
I manufacture custom batteries and service batteries for living and also hobby, luckily we havent used these new inventory models.
I mentioned in the video that I could forgive the occasional problem from a company. I'm a software developer and lord knows I've screwed up in my days. The problem here is that this isn't in isolation. On top of the various problems I've had with product consistency, app crashes, etc, Andy's has (twice I believe) blown FETs in JK BMSes in overcurrent tests. Clearly they build to minimum spec and throw products out into the market with little thought to safety or consistency.
@@TheDigitalMermaid Yeah, my point is not to belittle your opinion, the problem is serious when it comes to cell safety.
Im also do software development, im not sure is this BMS fault in software or in hardware area like memory gets reset when power is off too long.
Never had app crashes or heard that any customer have fault BMS yet but like i said that we have not used any those new inventer models what Andy have, i need knock on wood right now :D
Sun is finally starting to shine here in Finland, the solar panels are starting to produce something.
Man... I have a 3p8s battery that runs my home with a jk BMS. This worries me. Mine is 24 volt but dang this is super bad. If my battery blew up it would ruin me as it took over a year of saving everything I could just to buy the cells
It JUST the Inverter Edition, NOT the previous versions....
What other so-far-undiscovered flaws are there out there? I just can't trust this company anymore. Now, I will still use the JK BMSes I have until I can find an alternative, but the shaky trust I had in the company and their products is gone now.
@@TheDigitalMermaid I'm only home a couple days a month so my JK is off almost always but so far it's not had an issue. I bet jk prices go way down now that Andy kinda blasted them . I'm glad he did tho for sure.
@@TheDigitalMermaid I wonder about Chinese electronics and if we were to go to war with them I wonder if they have some way of destroying our systems we've built with all the Chinese products lol
i could imagine the hassard and damage that BMS could cause,
This bug could destroy your batteries, full stop. No real risk of fire, but I wouldn't be surprised if it caused a cell to vent, and certainly swell.
0:25 you say that you had to apply 5v across B- and P-.
Where this statement comes from? Is this your own finding or it's kind of officially stated method?
Maybe it's more outdated now? As I know currently BMSes have a button to start. Maybe earlier revisions of PCB had only this method to start the BMS ?
It was "the way" back then. If you look at the earliest videos I did with JK, I go into detail on it.
They probably fired qa as "cost savings". Works for quarterly profits but fails on lawsuits
But what about people who don't use heavy low density LiFePo4 batteries but rather Li-ion?
My channel has never been about NMC/Li-po or other chemistries, so I can't speak to that. I guess with this specific bug, they'd be OK, but again for me it's more about how JK approaches product dev and release that bothers me.
If I ever build my own Lifepo batteries, I going with DIYbms. If anything goes wrong, I know i can reprogram/repair it myself. I dont see these Chinese bms's going the long haul.
I have been looking for a mid-range quality BMS for a long time. The market is flooded with really scary devices. I'm an electrical engineer that designs power electronics (for different applications).
I have looked at these low-cost BMS units and totally unwilling to attach them to any of my packs.
Maybe I should design and build my own. There is probably enough frustrated people in the market to pay a little more so I could cover the development cost.
Developing a modern BMS is surprisingly hard, but by all means, I'd welcome anyone to try new designs. One thing is clear though; These cheap BMSes are not the right ones for safety critical installs (which by the nature of batteries, most are).
@TheDigitalMermaid Agreed...
I did a basic project definition for a BMS a couple of years ago. It was not a trivial product design cycle by any means.
What I wanted in the end is a system that is aerospace reliable and ensure that failure modes are benign.
Single points of failure in a BMS is scary. The consequences can be severe, as you clearly already know.
Is it the same reset issue with the type of BMSes you're using too?
No, it's just the newer "inverter" BMSes. I am not personally effected by this flaw, which is why I said in the video that I was signal-boosting Andy's video. He did confirm it, and given all the problems I've had with JK in the past, I have no reason to doubt him.
@@TheDigitalMermaid I'm too a follower of Andy. I watched his video before yours. I just asked to get confirmed.
Thanks!
Eu sou do Brasil. Agora estou preocupado. Isso pode acontecer com qualquer pessoa. Obrigado pela troca de experiência e ensinamentos. Vou ficar atento porque pode acontecer comigo
You will like the Riden power supply, be sure to use the green battery charger banana jack for charging cells or batteries. I am not a fan of JK BMS, but everyone has different priorities. I understand that Andy wants to get that last 0.1% of usable capacity for his (uncommon) use case of going months between full charges, but his testing methodology really annoyed me. I watched him change balance parameters at least 5 times in a single video and then declare that passive balancers don't work. Meanwhile I go six months without changing parameters and have no problems (I suspect some of his cells have wildly different self discharge rates). He "shadow banned me from commenting on his channel when I pointed out recommending a $500 active balancer simply to get a cut wasn't what I wanted to see. Anyway, enjoy the Riden, it is a very versatile power supply for working with LFP, reliable for a single cell or an entire pack. I own 5, so I can top balance 4 cells simultaneously (set it to 3.65v and wait for it to finish). I use a 12v BMS to get 4 cells up until cutoff, then top them up to 3.65v individually. I use Anderson connectors to connect the cells so they are easy to work with and you only have to hook up positive and negative correctly once. The Riden actually told me what was wrong with my 4 worst cells, since all were at the same resting voltage when received (3.3v). 12 of them took 180 amp hours to charge to 3.65v, and these 4 took less than 5 amp hours. That correlates to 12 cells measuring about 274 amp hours or above, and 4 that measured 264ah (approximately). Best I can figure is that the 4 were fully charged and sat somewhere for approximately 2 years (I knew they were "grade B" when I purchased them).
Anyway, don't charge to 3.65v on a regular basis, and in my experience an active balancer gets an extra half an amp hour after a year or so. I use JBD, since they started life making BMS for medical equipment and they have never given me a problem. Daly is like JK, random undocumented changes, and the very first one I tested wouldn't show charge cycles because (you guessed it) it wouldn't mark it full until 4.2v.
Yes, I used to design manufacturing test sets for power supplies 40 years ago, so I cringed everytime Andy changed parameters in a single cycle.
Glad to hear about the Riden! I've not had a chance to play with it, it was very much an unexpected gift! Having a 60v bench PSU is going to be super handy, now and on the boat.
The thing with Andy is that, for very different reasons, it's similar to my targets... I am trying to eek out every last mile of range I can out of my pack, and I don't plan to hit 0 or 100% SoC very often.
I'm going to set the top of charge to 3.5 or 3.55v, and stop the discharge at 2.6 or even 2.65v. You're right, there's so little power beyond there, so why not? I'm starting to think JBD, which I also started with, it going to be the new one to play with. I still need to think about this some more.
@@TheDigitalMermaid One think with riden if you planning go straight into 24A constant current you will need add a fan at the side of the power suply. it works flawles about 1h than inside temp sensor hits 75C and shut down itself(working safety), if you lower to 23A or put powersuply on the side(side vents now are at top and bottom) so hot air go out throught to vent and works nice again , small isue just put fan on side and problem solved
They could make an update so that it was not 4.2v but 3.5 or 3.3v and discharge 3v, then you would see that something was wrong and fix it instead of a fireman having to tell you that the batteries are gone.
If the defaults are hard-coded then a firmware update won't help, but an update could solve the problem. Another commentor mentioned it could be a bug that failed to write the new config to non-volatile memory, and if that's what it is, it can be fixed. However, a reset to defaults should always be a reset to fail-safe values, and this is not the case, clearly.
Hi I was told that the JK factory claims they have eliminated the problem through some updates, can anyone confirm this?
It's true, yes. I can't remember the specific version that fixed it though. Andy covers it in detail on his channel.
@@TheDigitalMermaid so does this BMS have your vote of confidence now?
@@alexberov6428 no, this was not the only problem, as I mentioned on the video. It was just the last straw. I knew it could be fixed, but they have so little concern for quality that these kinds of mistakes can happen again with safety critical parts of the system
@@TheDigitalMermaid thank you, I think Ill use the Seplos 2.0 version BMS
What is the model BMS number. Is all jk bms have this problem.
This specific problem is limited to the new inverter model from JK. Older BMSes appear unaffected, thought as I explained in the video, this is a sign of poor corporate culture.
I doubt this affects anything other than the brand new design BMSs that Andy has been covering. (I also think that a simple firmware update will more than likely fix the problem).
Personally I think JK are generally way ahead of the pack, and we might get some teething problems from time to time (bugs) simply because they’re so proactive at design and production of new gear. Long gone are the days of the Daly BMSs.
HOWEVER I’ve had one just like your’s sitting in a cupboard, unpowered, for about six months. I’ll get around to powering it up again (maybe later today) when I have a chance. I believe it will be fine, and will have retained it’s settings. (I’ll get back to you on this.)
Cheers again from hot and sticky Oz🇦🇺
It only effects the latest BMSes, but it's yet another example of their corporate culture and product development ethos. Slap it together and throw it into production! The role of the BMS is too crucial to use the "build it fast and break things", at least not in safety products going out to customers.
that may be but i have never had a problem with a Daly BMS ,they be be old school but they work ok.
It’s not an issue even with the new ones. Because you simply don’t disconnect them and put them on a shelf after configuring them like Andy does. This is such a remote and unlikely scenario to come by for any user. The product is meant to put in a battery box, configure it and use it. Not take it apart and leave it disconnected and imagine you don’t need to set it up again when putting it in a new battery. I can guarantee you have to go change some settings no matter if you put it in a different battery box.
@@pederb82 The funny thing is that in his video, Andy says he hasn't had this experience. I think maybe on a community post he said he had been able to verify it. It was reported by a few of his subscribers.
That said, when you configure parameters, the values should be written to (an) EEPROM, and they should be read back when the BMS first receives power. That's just good product development.
The fact that this is happening says that they are either NOT writing config to EEPROM, or they're not reading it on power up. That's a simple mistake, but can cause some not so safe situations. That's just NOT good product development. It's a pretty fundamental part of the design. And yes, you should be able to take it out of one battery box and put it in another (your example) and have the setting retained. Of course, you MAY have to change some settings, if the new box isn't 'similar'
@@CollinBaillie let this be clear: I agree that it should be that way. It’s clearly an error. If I got my way an BMS would only work on a specific chemistry period. Nothing beats consumer proofing things. What I’m reacting to is the doom of using the product. I want to know more about what is going on in stead of just hearing it’s dangerous - coming from my component level experience when it comes to electronics. I don’t see the BMS loosing its volatile memory from simply turning it off, without disconnecting it from the cells. It just isn’t logical. IF it does - then we can bring out the warning signs.
no no. only the JK Inverter BMS. ITs new tech and there will be bugs. I have had 4 JK BMS's and only issue I have had was when I had a bad crimp on a balance lead. I ordered the ones with handle bar displays and not regular display which can also start them because I didn't need big display for each or want a big display. I know you have had your issues but They are the only ones with active balancers that work.
Did you watch the full video? I explained why I no longer recommend JK, and it goes well beyond this one flaw. This flaw was just the worst, and the last straw for me.
@@TheDigitalMermaid its only the new inverter bms not the others versions like we have. That flaw will also be fixed with a Rom adjustment.
On a different note/topic. Seeing how you are a software programmer, you might be interested in fellow cruiser SV Rover project for a DIY project for 15" chartplotter with all the bells and whistles for around $500. He, also a software programmer, chose 15" touchscreen for cat helm bulkhead but viable for mono and smaller size monitors. The recent posting is on This Rat Sailed blog.
Ooooh, I am very interested! I tried searching for 'SV Rover' but all I got was a lot of car videos. Can you tell me his TH-cam handle (@...)?
Not a Ytube channel. He does written blog with pics on thisratsailed.blogspot.com@@TheDigitalMermaid
JBD all the way for me
I need more time to think, but at this point, that's currently the one I'm thinking of. I had the JBD Andy's got now, and I liked it, I only switched for the JK active balancing.
@@TheDigitalMermaid that's a point. I have external balancers and a voltage detection circuit to turn them on. Bit of wiring, but it works nice.