Connecting JK-BMS to Inverter - Any major advantage?

kolek said:

@hwy17 Thanks a lot for taking the time to respond.

Regarding "SOC control is a lot more accurate than voltage control," for the past day I've been reading about how the JK-BMS grasp on the SoC is less than trustworthy, and if you're using a JK-BMS, it's a better idea to use a Victron shunt to monitor SoC.

If that's true, maybe it's a better idea then to get the Victron shunt talking to the inverter than getting the JK-BMS talking to the inverter. No? If I'm spewing nonsense, please let me know. ;-)

Click to expand...

Seems proper to me. For me, one bms com advantage would be to alter the charging limits of the scc or genny if one or more bms goes down. I don't know how to make this happen with schneider without a bunch of programming wizardry which is not my wheelhouse. For an all-battery cumulative soc, say from a smartshunt, that would be useful for limiting max genny soc so I leave some room for the sccs to do something if the sun comes out. It would of course be useful to adjust this soc calc to account for batteries that are offline. In summary, it seems like a good solution would be one quality shunt, with either the shunt or inverter/charger/sccs adjusting dynamically for offline batteries. This would just be a safety function, as it should not be an expected occurrence for a battery to go offline.

chaosnature said:

you have the change logs? - its all in chinese

Click to expand...

-V15.06 Upgrade logs

1.Fixed a bug where parameters were lost after BMS restart. The bug is caused by a logical error in the parameter entered by the user.
2.Change the default configuration parameters to LPF to improve security.
2.Optimized JKBMS Modbus protocol, added RCV RFV register support. The protocol must use the PC software version V2.5.0。
3.Fixed the Bug that the hardware version may be incorrect in the device information.
4.Add battery low temperature alarm to dry triggers.
-@2024-01-09 by JKBMS

-V15.09 Upgrade logs
1.Fix a bug, parameters (RFV Time) Storage failure.
2.Change the default configuration parameters to LTO to improve security.

-@2024-01-12 by JKBMS

-V15.10 Upgrade logs
1. Button shutdown logic detection and optimization.
2. Bluetooth communication optimization.

-@2024-01-17 by JKBMS

-V15.11 Upgrade logs
1. DRY Trigger source added "12 - Remote Control "option
2. The LCD V2.0 protocol is added to the protocol library, which supports charging and discharging through the touch screen.
3. The Victor protocol allows the vendor name to be changed by writing the user's private data.
4. Fix a bug, parameters (RFV Time) Storage failure.
-@2024-03-01 by JKBMS

Well, unless someone posts something compelling to convince me otherwise, I don't see a reason (at least for me personally) to bother with communications between the JK-BMS and my inverter, at least until it's a more mature solution.

If anything, I'd rather figure out a way to have a Victron Smart shunt sending my SoC info to my Deye inverter, because at least then the SoC data it was sending would be somewhat reliable.

kolek said:

@chaosnature just to be 100% clear, you're using the new JK BMS with the latest firmware and you're saying you don't yet see a clear benefit to having the communications between the BMS and the inverter. Is that correct?

Click to expand...

Choasnature has a different inverter(Solis)
I can tell you for a fact that for deye/sunsynk/sol-ark have definite advantages..
But hey, it is your setup, you do what feels best for you

houseofancients said:

Choasnature has a different inverter(Solis)
I can tell you for a fact that for deye/sunsynk/sol-ark have definite advantages..
But hey, it is your setup, you do what feels best for you

Click to expand...

As I understand it, the main point of the communications is for the JK to be communicating the SoC to the inverter so the inverter can take various actions based on that. Is that your understanding as well?

100FUEGOS said:

-V15.06 Upgrade logs

1.Fixed a bug where parameters were lost after BMS restart. The bug is caused by a logical error in the parameter entered by the user.
2.Change the default configuration parameters to LPF to improve security.
2.Optimized JKBMS Modbus protocol, added RCV RFV register support. The protocol must use the PC software version V2.5.0。
3.Fixed the Bug that the hardware version may be incorrect in the device information.
4.Add battery low temperature alarm to dry triggers.
-@2024-01-09 by JKBMS

-V15.09 Upgrade logs
1.Fix a bug, parameters (RFV Time) Storage failure.
2.Change the default configuration parameters to LTO to improve security.

-@2024-01-12 by JKBMS

-V15.10 Upgrade logs
1. Button shutdown logic detection and optimization.
2. Bluetooth communication optimization.

-@2024-01-17 by JKBMS

-V15.11 Upgrade logs
1. DRY Trigger source added "12 - Remote Control "option
2. The LCD V2.0 protocol is added to the protocol library, which supports charging and discharging through the touch screen.
3. The Victor protocol allows the vendor name to be changed by writing the user's private data.
4. Fix a bug, parameters (RFV Time) Storage failure.
-@2024-03-01 by JKBMS

Click to expand...

thanks for this....

did you send the firmware link?

kolek said:

As I understand it, the main point of the communications is for the JK to be communicating the SoC to the inverter so the inverter can take various actions based on that. Is that your understanding as well?

Click to expand...

And cell voltages and pack voltages and temperature and charging voltage and much more

Here's an limited overview of what is send to the inverter by the BMS




But as someone much smarter than me once stated " this is my opinion, and opinions are like assholes, everyone has one"

chaosnature said:

thanks for this....

did you send the firmware link?

Click to expand...

Nah, I am just a random guy that stumbled upon the change log and saved it just in case it contained useful information.

@houseofancients
As always, thank you for your suggestions.
Are you recommending based on what you're seeing on the new firmware from JK-BMS, or from the independent project that's developing this, or both?

Because if it's the former, I don't have the new JK-BMS, I have old ones (still new in the box and yet to be installed) which don't support the new firmware. I'd need to sell the ones I have and buy new ones to implement the new JK-BMS/firmware.

And I thought we just agreed that new firmware was too new to trust.

Either way I'm completely open to being wrong, could you please point me in the direction of the posts or information you're seeing that this is a great option to implement? I've read hundreds of posts on this and I'm apparently not seeing what you are. Much appreciated!

kolek said:

@houseofancients
As always, thank you for your suggestions.
Are you recommending based on what you're seeing on the new firmware from JK-BMS, or from the independent project that's developing this, or both?

Because if it's the former, I don't have the new JK-BMS, I have old ones (still new in the box and yet to be installed) which don't support the new firmware. I'd need to sell the ones I have and buy new ones to implement the new JK-BMS/firmware.

And I thought we just agreed that new firmware was too new to trust.

Either way I'm completely open to being wrong, could you please point me in the direction of the posts or information you're seeing that this is a great option to implement? I've read hundreds of posts on this and I'm apparently not seeing what you are. Much appreciated!

Click to expand...

Again, cannot comment at all on the jk, other than that it isn't mature enough for me..
Can only comment on BMS - inverter communications, give my comments based on my inverters and BMS's and their behaviour and recommendations from both inverter and BMS manufactorer I have spoken too/researched

houseofancients said:

Again, cannot comment at all on the jk, other than that it isn't mature enough for me..
Can only comment on BMS - inverter communications, give my comments based on my inverters and BMS's and their behaviour and recommendations from both inverter and BMS manufactorer I have spoken too/researched

Click to expand...

Ok fair enough. Thanks again.

The JKBMS Inverter Edition software & firmware + can be obtained directly from JK at the link below.
NOTE: If site comes up in Chinese, go to the Top Right of the page, choose the British Flag for English.

I've been running JKBMS for quite some time, even before I edited their manuals. I've got quite a few of them out there in service, and my bank currently uses those. I am however doing upgrades & enhancements to t my bank & will be using the new Inverter Models BMS' in all of my packs in the bank. While they cannot interoperate with my specific inverter directly the bank can be connected to my management system which will allow me to actively monitor & control them as I need. With a small NodeRed Patch, the inverter will be able to interact.

JK presently does support a few of the more popular protocols and is adding more support while refining both the firmware & software. Also note that they are one of the FEW who have firmware upgradeability.

@Steve_S Thanks much for your contribution. How are you mainly monitoring your SoC? Shunt or JK-BMS?

kolek said:

@Steve_S Thanks much for your contribution. How are you mainly monitoring your SoC? Shunt or JK-BMS?

Click to expand...

I use Midnite Solar gear so my primary at this time is the WizbangJr Shunt. It is also accessible via Midnites ModBus stack.

Steve_S said:

I use Midnite Solar gear so my primary at this time is the WizbangJr Shunt.

Click to expand...

And I assume you're using that because the JK-BMS generally isn't as accurate at monitoring SoC.
And since SoC is probably the most important external metric the inverter needs to know, would you agree it makes more sense to prioritize getting the more accurate shunt SoC data into your management system/inverter versus focusing on communicating the "less accurate" data generated by your BMS (assuming you had one compatible with the JK-BMS)?

Edited for clarity

kolek said:

The newer version of the JK-BMS will talk to inverters, and there's a project to allow the JK-BMS to talk to the inverter.
It would mean my JK BMS could talk to my Deye inverter.
Sure it's cool, I get it. Maybe you can implement some cool charging strategies, etc., but beyond things like that, is there any real major advantage?
There's really only 2 advantages that I personally would be interested in:
- additional safety
- additional battery longevity
If it's not delivering either of the above, I probably won't bother.
Thoughts?

Click to expand...

A bit more detailed than in the original post, but ...

I used to think it wasn't necessary ,.. until one battery started getting more and more unbalanced and I hit OVP all the time, 5 disconnections an hour each day, etc. Long term that can damage BMS ... So it's better if inverter reduces load (current) as well as voltage near the top of the charge (or discharge) rather than the BMS cutting out "at full load".

1. First and Foremost ... PREVENTION instead of PROTECTION .
Avoid OVP and UVP as much as possible. Let cells time to balance out. DERATE instead of TRIPPING.

Your BMS should be the last line of defenses (PROTECTION), not the first and only one.

2. SOC Control. This of course in itself is a double edged sword. Coulomb counting is highly inaccurate (integrate 0.1A offset over several hours ...), so you must as part of your operation periodically let the cells reach 100% SOC and stay there at a "high" voltage until the balancer has done its job.

Unless and until you do that, then SOC control will do more damage than good.

The SOC is only as good as the operation and BMS reports it. I had BMS reporting 80% at 52V for a 16s LiFePo4 System ... I just didn't reach the top of the charge curve and let them stay there for a while. So over the course of several weeks coulomb counting did its thing ...

100FUEGOS said:

-V15.06 Upgrade logs

1.Fixed a bug where parameters were lost after BMS restart. The bug is caused by a logical error in the parameter entered by the user.
2.Change the default configuration parameters to LPF to improve security.
2.Optimized JKBMS Modbus protocol, added RCV RFV register support. The protocol must use the PC software version V2.5.0。
3.Fixed the Bug that the hardware version may be incorrect in the device information.
4.Add battery low temperature alarm to dry triggers.
-@2024-01-09 by JKBMS

-V15.09 Upgrade logs
1.Fix a bug, parameters (RFV Time) Storage failure.
2.Change the default configuration parameters to LTO to improve security.

-@2024-01-12 by JKBMS

-V15.10 Upgrade logs
1. Button shutdown logic detection and optimization.
2. Bluetooth communication optimization.

-@2024-01-17 by JKBMS

-V15.11 Upgrade logs
1. DRY Trigger source added "12 - Remote Control "option
2. The LCD V2.0 protocol is added to the protocol library, which supports charging and discharging through the touch screen.
3. The Victor protocol allows the vendor name to be changed by writing the user's private data.
4. Fix a bug, parameters (RFV Time) Storage failure.
-@2024-03-01 by JKBMS

Click to expand...

That's what Andy from Off Grid Garage Observed, where the BMS would restore to Li-Ion parameters in case an invalid parameter was entered (and no warning was issued). <= 15.06 according to what I see here at least

First Rule of Crisis Management is Prevention !
@kolek
I use that Shunt because it is an essential component of my Midnite setup, Classic Solar Controllers, EPanel etc and the software is on my desktop watching it out in the powerhouse. It is "reasonably" accurate but I do look at the overall voltage because I have several packs.

I do not use an AIO, I use a Samlex EVO 4024.
The Solar Controllers (Midnite Classics) have a ModBus protocol stack which can be interacted with by using software.
My Inverter/Charger has advanced programming and features. Their protocol stacks are not public. I have the stack info but with an NDA so I cannot share it.

My current JK's are JK-B2A24S20P, accurate and not a problem in fact !
The inverter models have many more settings and tweaks that can be applied to each battery & the bank as a whole. Given that my bank will consist of 6 280AH batteries, I am tying everything together. With charge limiting capabilities, that can come in really handy with a bank of batteries... A lot of the current feature set is stuff a LOT of us have been asking for.

Steve_S said:

My current JK's are JK-B2A24S20P

Click to expand...

I have 3 of the same model, sitting new in a box waiting to be installed. I'm debating whether to use them or sell them and instead install the new firmware upgradeable JK model. What do you think?

kolek said:

I have 3 of the same model, sitting new in a box waiting to be installed. I'm debating whether to use them or sell them and instead install the new firmware upgradeable JK model. What do you think?

Click to expand...

Keep also in mind that at the moment the new jk inverter bms is only in beta in syssi esphome-jk-bms repository. Maybe not even that. So maybe you will NOT be able to monitor it using home assistant etc.