• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

JKBMS Inverter Edition Problems/Issues | No Support / Help to fix major issues. - DO NOT BUY ! Warning (as of Oct.12.2024)

There are several issues here. Where lies the fault. The monitoring software whether it is Solar Assistant, Victron or any other software or the JKBMS firmware.

JKBMS should always provide software that allows for monitoring of the BMS. I will not buy a BMS (or battery) where I cannot via software monitor the battery. I have had to open battery cases to verify/validate what the BMS was reporting and what the software was displaying.

If/As the JKBMS firmware is updated do these updates also require third parties to update/modify their software?

Change one byte (or one bit) in any JKBMS response to any software that has not been updated and that monitors the battery and errors could exist.
 
yur43 said:
put it - it works, two buttons at the bottom have been added.
I'm testing.
100% SOC = no charge
Click to expand...
Well, thanks for chiming in too and confirming my testing with v15.30
As per usual, the coulomb counter will count up to 100%, then the inverter stopped charging.
It counts to 100% even though none of the cell voltage reaches SOC100% Volt value......

On the other hand, one cell reaching SOC100% Volt value and it resets the SOC % to 100% as usual........thus, inverter stopped charging without fully charging the battery.

ChrisG said:
Wanted to mask sure others saw this. The 100% SOC issue is not resolved according to @yur43
Click to expand...
Sigh.....how hard can it be to implement the logic, I can ask CHATGPT to come out with the coding too :ROFLMAO::ROFLMAO::ROFLMAO:
Example below on ensuring all cell reach 3.45v, preventing 100% from SOC100% Volt + coulomb counting from CHATGPT:

import time

# Constants
TARGET_VOLTAGE = 3.45 # Voltage to trigger RCV timer and eventually 100% SOC
HOLD_PERCENTAGE = 99 # Hold SOC at 99% until conditions are met
FULL_PERCENTAGE = 100 # 100% SOC after all conditions are met

# Variables
rcv_time_duration = 10 # Duration in seconds for the RCV timer (example)
rcv_timer_active = False # Flag to check if RCV timer is running
rcv_timer_start = None # Record the start time of the RCV timer

# Simulated list of active cell voltages (replace with actual sensor input)
active_cell_voltages = [3.40, 3.42, 3.45, 3.41] # Example data

# Simulated Coulomb counter SOC (this will be calculated in your real system)
coulomb_counter_soc = 98 # Example Coulomb counter SOC

def all_cells_at_target_voltage(cells):
"""Check if all active cells have reached the target voltage (3.45V)."""
return all(cell >= TARGET_VOLTAGE for cell in cells)

def check_soc(coulomb_soc, active_cells):
"""Update the SOC based on Coulomb counter and cell voltage conditions."""
global rcv_timer_active, rcv_timer_start

# Coulomb counter calculates SOC, but we need to enforce conditions
if coulomb_soc >= HOLD_PERCENTAGE:
# Override Coulomb counter SOC to 99% until the condition is met
if all_cells_at_target_voltage(active_cells):
if not rcv_timer_active:
# Start RCV timer if all cells have reached target voltage
rcv_timer_active = True
rcv_timer_start = time.time()

# Check if RCV Time has expired
if rcv_timer_active and (time.time() - rcv_timer_start) >= rcv_time_duration:
# RCV time expired, allow SOC to go to 100%
return FULL_PERCENTAGE
else:
# Timer still running, hold at 99%
return HOLD_PERCENTAGE
else:
# Not all cells are at target voltage, hold at 99%
return HOLD_PERCENTAGE
else:
# Coulomb counter SOC is below 99%, use its value
return coulomb_soc

# Simulate SOC checking process
while True:
# Example Coulomb counter updating the SOC (simulated here)
coulomb_counter_soc += 0.1 # Increment Coulomb SOC (for simulation)

soc = check_soc(coulomb_counter_soc, active_cell_voltages)
print(f"Current SOC: {soc}% (Coulomb Counter SOC: {coulomb_counter_soc}%)")

# Simulate active cell voltages changing (replace this with actual logic)
active_cell_voltages = [3.45, 3.45, 3.45, 3.45] # Simulating cells reaching 3.45V

time.sleep(1) # Delay for simulation purposes
Click to expand...
 
Steve_S said:
QUICKY UPDATE 15.30 Fireware !

OK, Took Bricked BMS, pulled the cover and disconnected power line, restarted with holding Power Button (and NOT letting go), using PC-App (already to go with firmware etc) and entered the FORCE UPDATE code and forced it. It took it and after a BMS restarted everything looks fine. SETTINGS NOT CHANGED !

Quick look through and everything "appears" to be fine.
Noted that the Forced Update cleared the logs - not a great loss but noteworthy if someone is keeping them.

Did 2nd 15.27 BMS and updated to 15.30 (no force required) and that went well.
Logs also seem to have been purged out.

SOC ! NOT RIGHT !!!!
They are NOT at my designated 100% (3.425), yet the BMS says they are at 3.395. The BMS' came up saying they are at 100%. UGH !
I am leaving 3 Packs at 15.24 (fairly stable) and will have 3 Packs running 15.30. As they cycle and do their thing today, I should get a better picture tomorrow morning with regards to SOC as dispensed out. Given they did not read "actual cell voltage in relation to my 100% SOC setting and flubbed it, I doubt this is sorted yet.

FYI: I am using the 2.9.0 PC-App which is still apparently MIA.

I will report back in this thread IF anything new or interesting pops out of this 15.30 update.
Click to expand...
From changelog: Added the ability to count discharge energy and charge energy.

Is this not the new way of calculating the SOC? Instead of voltage measuring... Or am I talking crap :P
 
prosier said:
From changelog: Added the ability to count discharge energy and charge energy.

Is this not the new way of calculating the SOC? Instead of voltage measuring... Or am I talking crap :P
Click to expand...
Yesterday all my packs hit FLOAT about noon and sat floating till around 6:30 when they started to slowly provide power/ All perfectly topped off and balanced with the latest differential being 0.003.

Now 05:35 hours and so just "overnight" low power usage (really quite minimal) and my Smartshunt says 98% which by the Voltage is actually quite right. The 3 Packs running 15.30 now say 97%, 99%, 95% yet Each Pack AVG Cell Voltage & Differential is identical and still within 0.005 differential. The 3 that are still 15.24 are 97%, 98%, 98%.


SOC = STILL BORKED !
 
Steve_S said:
Yesterday all my packs hit FLOAT about noon and sat floating till around 6:30 when they started to slowly provide power/ All perfectly topped off and balanced with the latest differential being 0.003.

Now 05:35 hours and so just "overnight" low power usage (really quite minimal) and my Smartshunt says 98% which by the Voltage is actually quite right. The 3 Packs running 15.30 now say 97%, 99%, 95% yet Each Pack AVG Cell Voltage & Differential is identical and still within 0.005 differential. The 3 that are still 15.24 are 97%, 98%, 98%.


SOC = STILL BORKED !
Click to expand...
I think 15.24 is still the most stable firmware with the least bug. Is there any reason to even consider .27 or even 30 at this point?

In the end with the wide use and adoption of the JK inverter BMS. I think the community would be better served by "jail breaking" the hardware and flashing an open source community supported firmware. There is precedent for this. Voltronic inverters have a notorious charging bug which has been on for years and which victronic refused to fix. The community released a firmware to fix the bug and continues to maintain same firmware today.

If we have to create a gofundme.. I think it would be worth it.
 
Steve_S said:
QUICKY UPDATE 15.30 Fireware !

OK, Took Bricked BMS, pulled the cover and disconnected power line, restarted with holding Power Button (and NOT letting go), using PC-App (already to go with firmware etc) and entered the FORCE UPDATE code and forced it. It took it and after a BMS restarted everything looks fine. SETTINGS NOT CHANGED !
Click to expand...
I had the same issue in update from 15.24. But I didn't have to open the box and switch power off/on BMS . Press and hold power on/off button and press reset has been enough. Force update started
My PC Software 2.9.0
 
bigbrovar said:
In the end with the wide use and adoption of the JK inverter BMS. I think the community would be better served by "jail breaking" the hardware and flashing an open source community supported firmware.
Click to expand...

Only if you can also build the hardware, otherwise you'll chase a moving target and if they decide to close shop or something you lose all the work.
 
upnorthandpersonal said:
Only if you can also build the hardware, otherwise you'll chase a moving target and if they decide to close shop or something you lose all the work.
Click to expand...
How though? there are still thousands of this BMS in the wild. an open firmware with a community behind it means it continues to get support and update even if the hardware is DOA. There are tons of old linksys wireless routers with totally EOL hardware which still gets latest update from the openwrt open wireless router project. Same thing on the mobiles.. lots of dated hardware still continue to run on community software from android open source project.

I think the beauty here is that, an open firmware supported by community would breath life into the JK inverter BMS and ensure it continues to get timely update.
 
bigbrovar said:
How though? there are still thousands of this BMS in the wild. an open firmware with a community behind it means it continues to get support and update even if the hardware is DOA. There are tons of old linksys wireless routers with totally EOL hardware which still gets latest update from the openwrt open wireless router project. Same thing on the mobiles.. lots of dated hardware still continue to run on community software from android open source project.

I think the beauty here is that, an open firmware supported by community would breath life into the JK inverter BMS and ensure it continues to get timely update.
Click to expand...

Because it's quite a major undertaking to reverse engineer the firmware. Personally I would not want to do this for a dead-end piece of hardware (assume JK stops making it tomorrow) just for the ones in the field. One of the reasons: you'll probably discover a bunch of issues with the hardware you'd want to fix, improve (think the coulomb counter, the entire AFE, etc.) which you now just have to work around in software instead of fixing them properly.
 
HarveyDent said:
I had the same issue in update from 15.24. But I didn't have to open the box and switch power off/on BMS . Press and hold power on/off button and press reset has been enough. Force update started
My PC Software 2.9.0
Click to expand...
When the BMS bricks in Mid-Update the Power Button becomes NON-Functional and you cannot turn off the BMS with the switch. Your only option is to disconnect the BMS Power (Red Plug-in wire).

Not my first rodeo !
 
OMG ! This is PRECIOUS !!
Msg sent to me via Alibaba Messenger this morning from a JK REP !
! NB: Keep in mind this is after the "Appeasement/Pacification" attempt by posting 15:30 Firmware & not bothering with the rest AND still being Borked no less.

1724847649685.png



YUP, I did, LONG AGO !



<--- REALLY ! "Great Adjustments" ??
ROFLMAO... Yup, indeed - adjusted without a doubt, we ALL see it, feel it and have to suffer for it...

"COOPERATE" Hmmm I wonder what dictionary is being used because they do NOT cooperate with people, not by any definition I am familiar with.

EDIT: I Gave this "Sandy" person a link to this thread.. Oi
EDIT2: LMAO !!! She responded with "I won't bother you" with an AOK emogi Ohh My how telling it is...
 
Last edited:
Steve_S said:
Yesterday all my packs hit FLOAT about noon and sat floating till around 6:30 when they started to slowly provide power/ All perfectly topped off and balanced with the latest differential being 0.003.

Now 05:35 hours and so just "overnight" low power usage (really quite minimal) and my Smartshunt says 98% which by the Voltage is actually quite right. The 3 Packs running 15.30 now say 97%, 99%, 95% yet Each Pack AVG Cell Voltage & Differential is identical and still within 0.005 differential. The 3 that are still 15.24 are 97%, 98%, 98%.


SOC = STILL BORKED !
Click to expand...
Can you printscreen and post your jk bms settings?
 
prosier said:
Can you printscreen and post your jk bms settings?
Click to expand...
All of my BMS are 200A Model with 2A balancing. Even sequential serial numbers as they all came from one batch.
All are Votage Calibrated to 3 decimal accuracy taking the readings at Pack Terminals and they are accurate.
Entire system is calibrated.

Just Looked: the 15:30 Packs now report 95%, 98%, 93% and they are ALL et exactly the same voltage even down to the Cell Averages with only 0.002 deviation.

1724852835320.png

1724852372119.png
 
Last edited:
Steve, have you logged and can graph exactly how many amp hours each BMS has had through it in that time frame?
 
Quattrohead said:
Steve, have you logged and can graph exactly how many amp hours each BMS has had through it in that time frame?
Click to expand...
At this time I do not have facility to do so. I am not using an AIO, No "Home Assistant" or the like nor Victron SW (no Victron here :-( ).
My SmartShuint is a Midnite WizBangJr due to my being a Midnite Guy and it does not log such.
 
Steve_S said:
All of my BMS are 200A Model with 2A balancing. Even sequential serial numbers as they all came from one batch.
All are Votage Calibrated to 3 decimal accuracy taking the readings at Pack Terminals and they are accurate.
Entire system is calibrated.

Just Looked: the 15:30 Packs now report 95%, 98%, 93% and they are ALL et exactly the same voltage even down to the Cell Averages with only 0.002 deviation.

View attachment 239810

View attachment 239805
Click to expand...
Did one cell here, hit 3.449v ? Because at that point (OR all cells >3.449?) the bms should reset to 100%.
 
When it comes to BMS settings Steve_S wrote the book. Literally. The previous rendition of the JK manual was updated to English by him so he knows his Bms material. If something is not working correctly it is the unit. A key reason he started this thread.
 
JJJJ said:
When it comes to BMS settings Steve_S wrote the book. Literally. The previous rendition of the JK manual was updated to English by him so he knows his Bms material. If something is not working correctly it is the unit. A key reason he started this thread.
Click to expand...
I surely agree, I am new to the stuff, just thinking :)
 
FWIW guys I have a couple of Hantek 365s bought 5 years ago, worked ok a data logger for V and I used for battery monitoring. Warning the graph s/w only works on Win XP (maybe Win7?). Nice thing is you can open 2 instances for 2 units (for V and I) and display separate graphs in own windows and arrange the to display side/side on screen. Mine wer £50 ea Runs off usb supply
1724857747927.png
 
@Steve_S Hi I am very uncomfortable putting myself in the hands of JK for total battery control. I get the it can do some things well, like Battery protection and cell display monitoring and alarms UVP and OVP. I havent attempted any comms to extract data.

I get also that active balancing must take place in the knee region ca 3.550 - 3.600. It seems that this phase should only be needed say 1/week to 1/month YMMV. Normally cells should stabilise ca 5-10mV region and having been top balanced then a normal target 100% = 3.45V is desirable. For long life you dont want to be hitting the ceiling all the time. Tesla aims for 80%SOC (for his Lion cells). Switching to balance mode can be controlled separately by your own timer system. IMHO forget JK rubbish SOC, its never gonna work. Arrange your own coulomb counter (Smart Shunt seems favoured) and keep you own SOC records. Important thing is to shut down charge when needed
 
Might I suggest a search on “working range” as there is a great deal written in other threads about this. In my cell testing it was eye opening how few amps are above and below that range.
 
Boron said:
@Steve_S Hi I am very uncomfortable putting myself in the hands of JK for total battery control. I get the it can do some things well, like Battery protection and cell display monitoring and alarms UVP and OVP. I havent attempted any comms to extract data.

I get also that active balancing must take place in the knee region ca 3.550 - 3.600. It seems that this phase should only be needed say 1/week to 1/month YMMV. Normally cells should stabilise ca 5-10mV region and having been top balanced then a normal target 100% = 3.45V is desirable. For long life you dont want to be hitting the ceiling all the time. Tesla aims for 80%SOC (for his Lion cells). Switching to balance mode can be controlled separately by your own timer system. IMHO forget JK rubbish SOC, its never gonna work. Arrange your own coulomb counter (Smart Shunt seems favoured) and keep you own SOC records. Important thing is to shut down charge when needed
Click to expand...
Have a look here at the post below
Steve_S

Post in thread 'study shows capacity loss with high average charge state (LFP)'

Many of us are aware of this report BUT you really have to read it to see what is actually going on.
Appreciate that LFP like all chemistries has 2 Voltage Ranges.
Allowable which is where no harm/damage is caused to cells if kept within said range (2.500-3.650)
The Working Range which is where the Deliverable AH come from (3.000-3.400)
* VPC = Volts Per Cell

This testing process worked THROUGH THE ALLOWABLE RANGE of 2.500 Bottom to 3.650 absolute top.
Reality is that NO ONE - Not an ESS Maker or EV Maker using LFP will ever allow for Max Charge to goto 3.650Vpc ! At most they will go...
  • Like
  • Love

Not going to go round & round the mulberry bush again.
No longer going to respond unless it is On Topic and directly related to a specific point.
 
You must log in or register to reply here.

diy solar

diy solar
Back
Top