JKBMS Inverter Edition Problems/Issues and dealing with JK to get them fixed - (BUY at your Risk/Peril) UPDATED Aug.16

[Steve_S]

EDIT: AUGUST 16.
Today is / was cutoff day for response. None received and no solutions are forthcoming.
DO NOT EXPECT Support or Help from JiKong !
There was going to be an additional Thread to NOT BUY or DO SO AT YOUR RISK but this thread will serve as it contains the thread of issues etc.
There will be no further follow-up from myself. I am washing my hands of it.

V15.24 is a relatively stable Firmware 15.23 Kills displays DO NOT USE. (note they still have 15.23 as most recent firmware om their site.)
V15.27 BORKS SOC and more and CAN brick BMS.
PC-APP 2.9 Works reasonably
DROID App 4.25 also works reasonably well and adds 3 decimal point accuracy where appropriate.
!!! BMS Does operate reasonably well in Stand-Alone Mode not using COMMS for interface & controls.

GOOD LUCK, Hope you don't get further phookered.

---

AS OF Monday, August 5 @ 22:00 hours Shenzhen Time, JKBMS Company has NOT responded to Vendors/Suppliers who messaged them about issues, fixes & support nor Individuals direct messages.

If they do respond, it will be updated in here for others to see & make their choices.

BUY with the understanding that there may in fact be little to no support or fixes/patches for issues beyond the current state. Do NOT have "Faith" (or false hope) that they will step up & do the right thing and/or complete the current product lines.

Tuesday August 6 as of 06:30 Shenzhen Time one supplier got a response & expecting similar. You'll love it ! Says a lot.
"Helen (the saleswoman I contacted) replied to me. She said that she had not received such feedbacks on the BMS issues."
Really ? Video's, Emails by many, WeChat attempts & more... Hmmm HHHHoooookay !

 

[....]

I got around to implementing the ghost battery using a dedicated microprocessor to spoof the battery. I created a new thread for it at https://diysolarforum.com/threads/j...er-ghost-battery-open-hardware-project.88554/


So far it has been working well, I was out of town last week so the system sat around 100% SoC but the ghost battery worked flawlessly and the RCV (55.2v)-> hold for 1 hour ->RFV (53.6v) charge controller worked perfectly as shown in the plot below:




At least for now I consider the system stable and unless I find something else wrong will heat shrink up the ghost battery and call this project done

 

[Steve_S]

A very good creative solution that will help many in the interim. EXCELLENT !!!

!!! QUESTION FOR MEMBERS Help yourselves
We need a comprehensive list of "problems, Quirks, Issues and NEEDS/WANTS.
Please post the issues being observed in plain form with minimal detail. EXAMPLE BELOW

* 100% SOC is triggered when 1st cell reaches setpoint and cuts off Charging with connected Inverters (Make/Model experiencing this)
* SOC Drift Correction & resets must be addressed and lower resolution for Amperage in/out needs improvement to show at "least" 0.1 accuracy.
* Errors within Battery Packs in parallel are not identified by the pack to Victron UI.
* BMS cannot disable T3 & T4 if BMS is used in 8S Config and generates a persistent ERROR if the sensors are not present.
(could be auto-config based on # of cells or with a switch.)
* PC-APP cannot download BMS Profile and upload same to other BMS', requiring manual setting of Each BMS parameters separately.
* PC-APP should also be able to run on Linux / Mac. LINUX more likely preferred.
* Phone-APPs install inconsistently on Phones with some versions installing while others do not on the same phone. There are also different sources from PGYER or Google-Play or JKBMS.NET which are often different versions as well. APPLE likely the same ?
* A LACK of Update Notes / Change Logs on Patches (firmware) and on Software which is critical to know what has been addressed/fixed/updated.
* DEFAULTS for LFP & Other chemistries MUST BE CORRECTED to safe values (Temp Cutoffs & Voltages etc)

Then specific Protocol Stack issues for Victron, PylonTech, Deye and others as YOU THE OWNERS have experienced and noted.

WHY ?
I would like to have a Comprehensive List put together that we can ALL SEE and Consider and provide to JiKong in a clean simple format so that they can address the issues & problems with a checklist.

PLEASE - Keep It Simple & let's not go wandering off with chatter. Let's get this list together FAST so it can be forwarded up to JK for actioning & to see IF THEY WILL actually address these in a timely fashion.

PS: Do not worry about making waves with JK, some are chicken to do so because they do not want to offend the Goodie Providers as they fear being cutoff by others if they take a Hard Stand against a company who is screwing up. You can join in and send the Complete List of issues directly to them if you want, but I WILL PUSH IT myself in any case. I don;t do reviews / vids etc and have nothing to risk other than pissing them off for highlighting their shortcoming and Outting them for it. Everyone has a Right to Know what the screw ups are !

 

[JJJJ]

We need a comprehensive list of "problems, Quirks, Issues and NEEDS/WANTS.

This is a terrific idea. Will help communication with JK so they can prioritize needed fixes.

 

[AshleyL]

For the Dry Relay contact, one cannot enter decimal value.
For example, "Cell voltage over" and "Cell voltage below". If we can enter decimal value, I can connect the Heltec 5A capacitive balancer on/off switch to the relay.

 

[Steve_S]

AGAIN, JK Personel state they have received no comments. complaints or issues as of 2024-08-06 07:37 Shenzhen Time. n This was at yet another vendor carrying the product... That's TWO they've denied knowledge to.

COME ON FOLKS, Chime In it is The Opportunity and if you don't add your voice to the issues. problems etc that must be resolved, then go mutter about it somewhere else and do not ask anyone to help you.

THIS IS THEIR LIST OF FOLKS ! E-Mail & Harrass them into action if you want.
Helen@jkbms.net
Seven@jkbms.net
Sandy@jkbms.net
Miranda@jkbms.net
Ross@jkbms.net
Nami@jkbms.net

 

[Steve_S]

Continuing this here, thanks for the move Supervstech.
Throwing a PING in here for @Nami of JKBMS.NET / Hankzor so it get's noticed (big maybe there). not holding breath for that to happen now.

 

[yur43]

Continuing this here, thanks for the move Supervstech.
Throwing a PING in here for @Nami of JKBMS.NET / Hankzor so it get's noticed (big maybe there). not holding breath for that to happen now.

A lot of JK-BMS ads on YouTube.
Full of negative comments about the JK inverter.
And they pretend that they did not hear anything?

 

[kolek]

Andy from Off-Grid Garage (the well-known YT channel) talked quite a bit in one of his videos about working with JK and said they had been basically unresponsive to any of his bug and issue reports despite him trying to work with them over a period of... was it 1 or 2 years? I don't remember exactly, and I don't remember exactly which video he said that. If someone can post the specific video where he talked about that, great.

In my own experience, I have messaged them a few times and never received any response.

Would be great tho if they suddenly changed, but I wouldn't get my hopes up.

Just my 2 cents worth.

 

[Steve_S]

ASSEMBLING a LIST from YT Vid Collections (boy there's a lot to muddle through)
JKBMS Errors Issues & problems
-----------------------------------------------
JK Master BMS shows an alarm but the slave with the alarm doesn't. Master does not indicate which Slave Pack and what the alarm is.
No Alarm shown on VIM nor Victron Status screen

A Single Pack can block charging ALL packs if there is an Alarm/Error.

JK Firmware V15.23 "Kills LCD Display" and is still the Latest Update available on JKBMS.NET when 15.24 Fixes it and while 15.27 is flying about the web (unofficially)

Victron: Battery Cell Status reports ONLY the High & Low Cells but NOT what pack(s) they are within. This is effectively useless as it is currently reported to AIO / Victron Interface.

VICTRON: low temp temperature, low charge temperature indicators says Ok. BUT the BMS is not reporting what pack is STOPPED nor does it set the flag properly to indicate that error. Works OK when ONLY a Single Master Battery Pack is online, not for Paralleled battery packs.

Inverter BMS: if there is any protection any alarm, over voltage, current, temperature, ETC there's no communication between the parallel battery Packs and the master is not reporting this issue back to your inverter or to the victron system. There is no way to know what errors are occuring. They are NOT reported.

SOC 100% will report that Pack is FULL when only 1 cells reached set Trigger Point, For many devices like Voltronic, Victron etc, that terminate charging which is NOT Correct and other packs in Bank will also stop charging as a result. THIS IS TERRIBLY WRONG ! This ends Absorb/Bulk, and Float thereby effectively terminating Balancing & Levelling of cells & packs. THE FIX should be to set 100% SOC Value and trigger FLOAT once reached. Other options are possible.

Another SOC 100% Error. Often the BMS will report 100% SOC even when NO CELL has reached the 100% SOC voltage as set ! ABSOLUTELY WRONG ! This is likely a result of Bad coulomb counting and poor math within the firmware.

BMS does NOT COLLECT & REPORT Consumption and does not report this at all to anything.
BMS does not register Amperage in/out above 0.3A and often times not until 0.5A This obviously causes serious SOC drift which is made worse with multiple packs on a bank. Within 24 hrs there can be as much as 20% drift between identical packs with identical BMS'.

DEYE/Sunsynk still uses a Force Request Charge to charge from GRID. Parameters are not settable on BMS. Protocol Implementation is incomplete. A Temp Fix is to use Pylon Protocol 002. A HACK !

----------------------
ADDED Aug.08
BMS data not correctly displayed on 4.3" Screen with V15.24 & V15.27 Firmware. Indicated Balancing is ON when not actually on (below setting value) and Phone App verified it is off. Amps In/Out values also do not update with BMS - it is sometimes delayed or just off by a bit.

 

[Sun in Lithium]

The main issue for me with the Jk inverter BMS and Luxpower inverter, is the fact that once the BMS reports 100% it sends a flage to stop charging. This is with 15.27 and latest updates on the 6000xp.

One way to fix it is to add an option. To hold the soc at 99% until after requested absorb time is up.

I believe it was Andy's request to set the 100% soc based on voltage.

Which works fine for Victron and SMA SI

But not so much for AIO's.
So an option to choose would be great.

2nd thing is the heater logic turns on only after low tem charge disconnect. And off righ after low temp charge recovery.
(also there needs to be a charge potentiall before it turns on which is good.)

Which will result in a constant cycle. At low temperatures.

It should have its own parameters.
For when it turns on and off.

 

[Steve_S]

A lot of JK-BMS ads on YouTube.
Full of negative comments about the JK inverter.
And they pretend that they did not hear anything?

Yes, I have found a few but I'm not going through comments cause sometimes it's a LOT ! Andy's Last vid was a good source for the Victron issues but as Sun In Lithium above, notes issues with Luxpower as well and offers possible solutions.

A GENERAL NOTE FOR THE THREAD
There are several different protocols in play with these BMS', including other brands like Seplos. The Communications information for developers are sometimes completely "open" and the entire "current" information is readily available while others are not or are partially "closed source" making it difficult. Then of course there are versions (whole other ball of ... ) And last but not least it also depends on the Inverter Manufacturers implementation for the comms protocols.

At present, it appears that the PylonTech implementations have the least issues "in general" just from what I have read & watched on YT. I have no way of knowing myself, so...

 

[Steve_S]

Andy from Off-Grid Garage (the well-known YT channel) talked quite a bit in one of his videos about working with JK and said they had been basically unresponsive to any of his bug and issue reports despite him trying to work with them over a period of... was it 1 or 2 years? I don't remember exactly, and I don't remember exactly which video he said that. If someone can post the specific video where he talked about that, great.

In my own experience, I have messaged them a few times and never received any response.

Would be great tho if they suddenly changed, but I wouldn't get my hopes up.

Just my 2 cents worth.

He mentions it in the vid below. I pulled some of the info above from that & other vids, trying to assemble it all in one place.
A few others who helped JK with various things "served" their purpose and all of a sudden ghosted. Its just Bad JuJu for a company to do that, so time to either get off the pot or ... suffer the consequences whatever that might be.

 

[kolek]

Hi @Steve_S thanks for posting that video. Actually it wasn't that video that I saw, but this has been a long-standing problem, and so obviously he's mentioned it in multiple videos.

I really applaud your efforts to try and get JK to fix their issues, but IMHO they have already been given many chances already and proven they have no intention of working with the community to address their issues. Or maybe they just don't have the manpower to respond. I have no idea.

Your time is precious and I'd hate for you to be wasting it on a (frankly) crappy Chinese company that has no intention of listening.

In my own experience, having worked with so many Chinese companies in the past, JK's behavior is very representative of what you can expect from Chinese management. They are only interested in responding when you are going to make a purchase. After the purchase, you don't exist. A notable exception being Amy at Luyuan.

To fix this problem permanently, we'd need a BMS where the software is open source and on Github so that we can fix our own bugs. I guess there are open source BMS projects? Don't know anything about that. In this case, the hardware could still be Chinese.

 

[Sun in Lithium]

To fix this problem permanently, we'd need a BMS where the software is open source and on Github so that we can fix our own bugs. I guess there are open source BMS projects? Don't know anything about that. In this case, the hardware could still be Chinese.

I very much agree with this. Even if Some of us needed to sign a NDA, preferably not as I'm very much pro open source.

I find it very frustrating when you know how to code, and it would be such an easy fix. But you're helpless because some Chinese company is ignoring you.

For me the jury is still out. But they can only deliberate for so long.

I've already started to order the seplos 10e bms again. And I'll just add a neey active balancer if needed. And connect the positive power leed of the neey to the inverter side of the bms. That way if the battery shuts down do to low voltage the balancer won't continue to draw power.

As that is an issue I've noticed with the neey it has no low voltage shutdown setting.

 

[houseofancients]

I very much agree with this. Even if Some of us needed to sign a NDA, preferably not as I'm very much pro open source.

I find it very frustrating when you know how to code, and it would be such an easy fix. But you're helpless because some Chinese company is ignoring you.

For me the jury is still out. But they can only deliberate for so long.

I've already started to order the seplos 10e bms again. And I'll just add a neey active balancer if needed. And connect the positive power leed of the neey to the inverter side of the bms. That way if the battery shuts down do to low voltage the balancer won't continue to draw power.

As that is an issue I've noticed with the neey it has no low voltage shutdown setting.

There is a fully open BMS in hard and software that works really well..
Stewart pittaway's diy bms

DIYBMS v4

Share your videos with friends, family, and the world

youtube.com

GitHub - stuartpittaway/diyBMS: Do it yourself battery management system for Lithium ion battery packs/cells

Do it yourself battery management system for Lithium ion battery packs/cells - stuartpittaway/diyBMS

github.com

 

[Norwasian]

Perhaps some effort toward communication with JK in Chinese would be worthwhile. Does one of the members here have sufficient fluency to correspond in written Chinese?

I speak basic Chinese, but do not write or type it; and I certainly lack technical vocabulary. I could say "electricity" and that's about as technical as I'd get!

Having worked with Chinese, I know that Google Translate is capable of massacring it. Messages may not survive translation, and the folk at JK likely rely on machine translation. If they know some English, they are not likely fluent in idiomatic English.

If communicating with JK workers in English, my advice would be to avoid all humor, all sarcasm (Asian languages do not understand Western sarcasm well, not having it in their own language), all unnecessary technical terms, and all idioms. Use simple language and grammar that can more easily survive translation, and keep your sentences short. Even simple language, if it happens to be in a confusing order, may not convey what one thinks. For example, the word "thing" in Chinese translates to "East-West," i.e. the two words together create the concept of "thing." Imagine if one happens to be talking about the orientation of one's solar panels and mentions some "thing" related to them....it could create an ambiguity in the language that would be up to the reader to interpret. This is why it is very important to keep the sentences and their grammar simple when communicating across languages.

I kid (an avoidable colloquialism right there) you not, I have seen examples of machine translation that converted a positive message to a negative one or vice versa, e.g. lost the "not" in the sentence.

 

[MrSparkle]

What @Norwasian said!
(A simple "like" is not enough.)
There is a very good saying "Eloquence (expressiveness) is to only say what is needed".

 

[kommando]

There is a fully open BMS in hard and software that works really well..
Stewart pittaway's diy bms

DIYBMS v4

Share your videos with friends, family, and the world

youtube.com

GitHub - stuartpittaway/diyBMS: Do it yourself battery management system for Lithium ion battery packs/cells

Do it yourself battery management system for Lithium ion battery packs/cells - stuartpittaway/diyBMS

github.com

This is a good project, Stuart is very responsive and capable. However it is not a BMS that actually has its own internal switching using mosfets to disconnect the battery, most builds using this do not add one and so there is a risk involved unless you add a switch/mosfets.

Designing a MOSFET switching PCB for diyBMS, low temp, low and high voltage shutoff?

I know there has been talk about relays for under and over voltage control, I know Stuart is designing a shunt to measure current, I don’t know to much about mosfet switching, I have had limited success with mosfets on a solar HWS directly connected to solar panels, BMS like Daly use mosfets in...

community.openenergymonitor.org

 

[Steve_S]

Perhaps some effort toward communication with JK in Chinese would be worthwhile. Does one of the members here have sufficient fluency to correspond in written Chinese?

I speak basic Chinese, but do not write or type it; and I certainly lack technical vocabulary. I could say "electricity" and that's about as technical as I'd get!

Having worked with Chinese, I know that Google Translate is capable of massacring it. Messages may not survive translation, and the folk at JK likely rely on machine translation. If they know some English, they are not likely fluent in idiomatic English.

If communicating with JK workers in English, my advice would be to avoid all humor, all sarcasm (Asian languages do not understand Western sarcasm well, not having it in their own language), all unnecessary technical terms, and all idioms. Use simple language and grammar that can more easily survive translation, and keep your sentences short. Even simple language, if it happens to be in a confusing order, may not convey what one thinks. For example, the word "thing" in Chinese translates to "East-West," i.e. the two words together create the concept of "thing." Imagine if one happens to be talking about the orientation of one's solar panels and mentions some "thing" related to them....it could create an ambiguity in the language that would be up to the reader to interpret. This is why it is very important to keep the sentences and their grammar simple when communicating across languages.

I kid (an avoidable colloquialism right there) you not, I have seen examples of machine translation that converted a positive message to a negative one or vice versa, e.g. lost the "not" in the sentence.

I have been dealing with Chinese Vendors since the late 1980's (Computer related) and learned a few lessons. KISS Applied in language is essential.

PLEASE EVERYONE - LETS KEEP ON TOPIC !
NB otherwise I will have to ask an Admin to Clean Off-Topic wanderings out.

I need the LIST of ISSUES Found with various AIO's ETC interfacing with these BMS.
Problems with Settings or Functions that are not working or need adjustment.
Software / APP Issues (functional & Technical) - Feature Requests to a point (IE like downloading BMS profiles)

 

[kolek]

I need the LIST of ISSUES Found with various AIO's ETC interfacing with these BMS.

Rather than overwhelming them with a long list of issues, which they are much more likely to ignore, how about trying to get them to fix even ONE issue? You can get someone to write the issue in gramatically correct Chinese. Then after they ignore you, you can be sure that sending them a long list is going to produce no results.

I'm already 99% sure they will ignore your request anyway. We already know they did nothing for Andy, and he's not dumb, nor is he a poor communicator. There's no point in having lots of people get involved before you have verified they have the intention of working with the community, and they have proven that intention by actually fixing ONE issue.

 

[JJJJ]

I love DIY however building my own BMS is not in my list of abilities. I don't think I am alone in this.

I think the matter at hand is how to resolve the communication issues with JK. They have a product which is on the market, is widely sold and had made improvements in software development. The concern as I understand it is that support has appeared to wane leaving key issues yet to be resolved.

This thread was at 247 views early this morning and has jumped to 323 at this time so it seems to be a topic many are interested in. I recall a while back where Will mentioned the number of members and visitors to this forum. I can't recall the exact number and I don't want to hijack the thread but it was an impressive number.

The intent of this thread as I understand it is to encourage readers to communicate their concerns to JK. If JK decides to turn a deaf ear then customers have the opportunity to vote with their wallets and purchasing decisions. Many have already spent their hard earned money however and are hopeful the product continues to develop.

A comparative analysis of BMS units would make for an interesting read if one wanted to start a new thread about it.

 

[Quattrohead]

I've already started to order the seplos 10e bms again.

Yikes this says a lot, they are not exactly easy to use and bug free either.
And yes communications with them, KISS.

 

[AshleyL]

I have been dealing with Chinese Vendors since the late 1980's (Computer related) and learned a few lessons. KISS Applied in language is essential.

PLEASE EVERYONE - LETS KEEP ON TOPIC !
NB otherwise I will have to ask an Admin to Clean Off-Topic wanderings out.

I need the LIST of ISSUES Found with various AIO's ETC interfacing with these BMS.
Problems with Settings or Functions that are not working or need adjustment.
Software / APP Issues (functional & Technical) - Feature Requests to a point (IE like downloading BMS profiles)

Fine fine......
Using Growatt SPF 3000, 5000 and 6000 ES.
Have both V14 and V15 hardware with latest firmware available (v14.20 and v15.27)
Tried Pylontech and Growatt protocol on JKBMS, no luck with SOC100% stop charging.
*Growatt inverter only support Pylontech and Growatt own protocol*

Critical issue with JKBMS coulomb counting, it doesn't take account of voltaic and coulombic efficiency during charge and discharge cycle.
It will always take more power to re-charge a discharged battery.

From what I can see, JKBMS actually expects 100% efficiency in charge and discharge cycle, duh.......assuming if the 100% is not triggered by single cell voltage, the coulomb counter gonna trigger 100% without actually fully charging the battery.
Example, after discharging a 280ah battery pack by 200ah, it requires 201ah or 205ah to fully recharge the pack again.

Charging and discharging several times in a day gonna cause the SOC % to drift negatively for certain.

 

[Norwasian]

Not a list here, just one item, the only problem I have personally experienced so far with JK BMS (I believe the model is: JKPB2A16S15P): The screen suddenly changed from English to Chinese language, without user input or intervention. The precipitating factor, if there was one, appears to have been the occurrence of a shorted circuit past the inverter which flipped a breaker. The 3kw toroidal (LF) inverter, however, did not miss a beat, yet the screen on the battery's BMS changed to Chinese.