Do you need an additional battery monitor with JK BMS?

[pekka8]

If there is a system with Victron Multiplus II inverter/charger and JK BMS v1.9, is there any need for additional Victron battery monitor (shunt)?

Is battery monitor only for viewing status info, which JK can provide, or does Multiplus or Victron MPPT utilize BM data somehow? If does, is it possible to provide the data from JK BMS?

Only use I see for BM is a backup and sanity check for seeing that JK BMS works correctly, and I don't see point in that.

 

[1ufvlsi6x]

I find them a luxury, in the sense that if you oversized the batteries a little then reading 25% instead 30% doesn't matter; you'd aim to rarely end up 0% anyway; and you should oversize the batteries a little if the system is quality so getting high quality shunts is less important the more quality the system is so it becomes a circular fallacy.

It's way more critical to charge correctly; it would get at 100% SoC (or whatever the goal is) properly and safely; then even the JK BMS' more inaccurate coulomb counter would be more than enough because it's not drifting much in a day anyway assuming it charges fully frequently.

 

[kommando]

The JK PB BMS suffers from SOC drift, not a problem if you regularly trigger the 100% reset. Over winter I never get a reset so my JK SOC drifts downwards, so I have a Victron Smartshunt and using a Rasp Pi I feed its SOC to my inverter. My inverter switches off when presented by a low SOC regardless of the measured V so I needed this intervention, if your inverter reacts off measured V and ignores SOC then no need.

 

[JanisD]

In case You have plenty of sun (which I doubt) and use Your system intensively, info from BMS SOC could be enough.

I my case (a bit to south from You which is good and having small system which is bad) I have periods when Wh collected covers only self consumption plus some light loads.
The JK BMS I have (B2A8S20P) simply ignores small charge/discharge currents and have measurement steps of 0.22A, so at such periods it drifts far away from real SOC. And there is no way to forcefully set some SOC value to it (I know, there are different opinions of that) even if I think I know what I'm doing.
The Victron shunt (300A), having steps of 0.1 A, is more precise and it is possible to correct the drifted SOC to one closer to reality.

So, not having data from Your system there could not be easy answer - if You are up to challenge, You could start without the shunt, leaving the place for it to be added later and monitor Your system closely.

Just my 2c.

 

[DC_Ripple]

The JK PB BMS suffers from SOC drift, not a problem if you regularly trigger the 100% reset. Over winter I never get a reset so my JK SOC drifts downwards, so I have a Victron Smartshunt and using a Rasp Pi I feed its SOC to my inverter. My inverter switches off when presented by a low SOC regardless of the measured V so I needed this intervention, if your inverter reacts off measured V and ignores SOC then no need.

I have a Victron Smartshunt and using a Rasp Pi I feed its SOC to my inverter.

If you don't mind can you please explain this?
Are you using the Smartshunt SOC to input to your inverter? Does Victron Connect send the data out?
I have a few Raspberry Pi Zero Ws laying around unused and would be interested to try this.

Thanks for any explanation of the connections.

 

[kommando]

My Rasp Pi is using Venus OS with a custom Sunny Island driver so I doubt it would work for a non SI setup.

I have this as a backup, tested and working using release 2.5, I could not get 2.6 to work using ESP firmware loader but you may have luck using Visual Studio. It will take any pylon compatible inverter.

GitHub - sijones/DiyBatteryBMS: Integration of Victron to Pylontech supported inverters with DIY Batteries

Integration of Victron to Pylontech supported inverters with DIY Batteries - sijones/DiyBatteryBMS

github.com

 

[DC_Ripple]

Thanks.
VE.Direct port is RS232 for my SmartShunt.

I am using RS485 Pylon communication from the batteries to the inverter.

Not sure If I can use the inverter with RS485 and CAN at the same time, I think it is an Either / Or option, but not both.
Eco Worthy Hybrid 5000W Inverter. 3 E-W 100Ah batteries.

I am running a Pi5 with Solar Assistant.
It will either communicate with the batteries [RS485] 'OR' the VE.Direct [RS232]. Again, not both.
I have emailed Solar Assistant about adding the ability to read both at the same time.
They say they might add it in the future.

RPi has RS232 hardware since day 1.
Seems like an easy add for SA to implement.

 

[pekka8]

Thank you all for the advice. I already have Victron BMV 712 from previous setup with dumb battery, so I'll keep it. I have periods of little or no sun and some occasional very small loads (Multiplus is also always on), so won't risk with JK SOC drift. Though there is still the problem where BMV might show correct SOC and JK not - what to do then?

 

[ConnerLabs]

Though there is still the problem where BMV might show correct SOC and JK not - what to do then?

Ignore the JK SOC? Does the JK actually perform any action based on its SOC reading?

I thought all actions taken by the BMS were on the basis of its cell voltage readings and the SOC is just for the user’s information. I could be wrong.

 

[pekka8]

Ignore the JK SOC? Does the JK actually perform any action based on its SOC reading?

I thought all actions taken by the BMS were on the basis of its cell voltage readings and the SOC is just for the user’s information. I could be wrong.

This is good question.

 

[kommando]

I am not aware of the JK PB itself doing anything based on its SOC but my issue was the SOC is sent via canbus and my SI inverter does act on low SOC's by ultimately turning itself off, even if the V is high. My first trick was to drop low SOC protection to 4, 3 and 2% but with a freezer and going away for a few days I went for a more robust solution.

Perversely the SI ignores the 100% SOC messages and unlike other inverters carries on charging until the target V is met.

If my inverter ignored the JK SOC then I would have left it alone.

 

[740GLE]

Jk doesn’t track less than around .5 to 1a so SOC drift is huge if you go more than a few days of getting full, so yeah a shunt is a must imo

 

[1ufvlsi6x]

Jk doesn’t track less than around .5 to 1a so SOC drift is huge if you go more than a few days of getting full, so yeah a shunt is a must imo

Not an issue for everyone, especially for people near the equator. Their panels give massive solar per area.

Their focus must be mainly for solar power, so that batteries are almost exclusively a night time thing.

 

[1ufvlsi6x]

And let's be honest here: the entire thing is a luxury at a general systemic level anyway.
You can guess after a point what will remain even if it drifted a lot anyway.

Maybe more important if you're not the user of it but someone clueless.

 

[740GLE]

Not an issue for everyone, especially for people near the equator. Their panels give massive solar per area.

Their focus must be mainly for solar power, so that batteries are almost exclusively a night time thing.

people in the north can have massive solar panels with equal output to those in the equator

IMO it’s a blind spot that may factor in the BMS SOC drift.

A smart shunt is cheap compared to the rest of system equipment, IMO there’s no good reason not to have one.

 

[1ufvlsi6x]

people in the north can have massive solar panels with equal output to those in the equator

Not if irradiance is almost zero in the winter. Here's a quick summary from a chat bot (probably a bit off in terms of accuracy but it's for the gist of it:

Numbers (rounded)​

Day / Location	−4h	noon	+4h
Winter solstice (≈ Dec 21)
Equator	558.7 W/m²	1017.5 W/m²	558.7 W/m²
Crete (35.2°N)	245.6 W/m²	620.4 W/m²	245.6 W/m²
N. Finland (69.0°N)	0.0 W/m²	0.0 W/m²	0.0 W/m²
Equinox (≈ Mar/Sept)
Equator	600.0 W/m²	1100.0 W/m²	600.0 W/m²
Crete	508.6 W/m²	917.1 W/m²	508.6 W/m²
N. Finland	279.2 W/m²	458.4 W/m²	279.2 W/m²
Summer solstice (≈ Jun 21)
Equator	558.7 W/m²	1017.5 W/m²	558.7 W/m²
Crete	704.2 W/m²	1079.0 W/m²	704.2 W/m²
N. Finland	635.8 W/m²	800.2 W/m²	635.8 W/m²

The summer is great for everyone but winter is terrible in edge cases (also weirdly the summer can be higher irradiance than the equator).

PS I guess that last bit makes sense between of the tilting geometry, but the equator always wins on the year average obviously.

 

[740GLE]

No one is arguing the north or south sucks at total generation compared to the equator.

IMO a shunt should be a part of every system no matter the size or total Kwhr/year, it’s a fact that JK BMS can’t track low current flows.

Sure you don’t “need” one to make the system work, but you “should” have one.

 

[tinyt]

The JK PB BMS suffers from SOC drift, not a problem if you regularly trigger the 100% reset. ....

Pardon my ignorance, how do you do this? Do you remove power from the BMS or something else?

 

[kommando]

Pardon my ignorance, how do you do this? Do you remove power from the BMS or something else?

To get the JK to trigger 100%

 

[pekka8]

Just making sure: Victron shunt is also only for display? It cannot communicate with Multiplus II without additional expensive hardware such as Cerbo?

 

[ConnerLabs]

To get the JK to trigger 100%

Is it mandatory to clutch your head and grimace like that to trigger the SoC reset? I thought it went to 100% when any cell voltage exceeded some threshold.

Just making sure: Victron shunt is also only for display? It cannot communicate with Multiplus II without additional expensive hardware such as Cerbo?

I think it's the other way round. The Victron Smartshunt and BMV712 can communicate with many of Team Smurf's MPPTs and DC-DC chargers over Bluetooth. It's the Multiplus II that has limited comms functionality and needs a Cerbo.

 

[pekka8]

Is it mandatory to clutch your head and grimace like that to trigger the SoC reset? I thought it went to 100% when any cell voltage exceeded some threshold.

I think it's the other way round. The Victron Smartshunt and BMV712 can communicate with many of Team Smurf's MPPTs and DC-DC chargers over Bluetooth. It's the Multiplus II that has limited comms functionality and needs a Cerbo.

I forgot the bluetooth network between BM and MPPT so that's right.

 

[ConnerLabs]

I forgot the bluetooth network between BM and MPPT so that's right.

I always wondered why Victron didn't put more comms options into their inverters. They only have VE.Bus which I think is some kind of RS485?

The inverter firmware just does the basics and they want you to use their DVCC algorithm running on the Cerbo for battery energy management in complicated systems.

Smart Shunt communication with MultiPlus II 3000/2x120 - VictronEnergy

communityarchive.victronenergy.com

You can use a Raspberry Pi in place of a Cerbo. A massive geeky rabbit hole that I've resisted going down so far.

 

[hardtop]

I use a Pixel tablet on a portable dock and the JK App. Thats all I've needed. 218 cycles so far in 1Y173D. Been thinking about connecting a Linux box to the JK hard wired though serial. Wireless isn't that great when there is interference and I'm planning for an overland DIY with a permenant install.

My charger is hot plugged into the pack with an anderson connector and generator fed. I could waste a lot of time in the Victron sw and sensor options I'm sure. My phoenix 12/1200 might have some options too but no time for it unless I had a remote start and stop generator. Then I could potentially create a program to fire the generator to fire up at x% remaining and shut off at 100% SOC.

 

[740GLE]

I forgot the bluetooth network between BM and MPPT so that's right.

You can get a VE.bus BT dongle for a MP, and it’ll share info to anything on the same VE.Smart, granted it’ll just be AC usage which you already kinda have with the shunt.

I thought I’d be happy with a VE.smart network of the shunt, SCC and MPII all talking together, but it was just a gateway to a Cerbo/Venus which I picked up 5 months into my system.

If anyone wants a BT VE.Bus dongle hit me up I have one that could go to a good home.