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

[agtcovert]

The % deviation of one day not getting full charge is 3% max between top and bottom.
3% deviation every day, 3 days close to 10% deviation. I don't think JK can ever fix the SOC caused by the cheap shunt.

My experience has been about the same with 15.33. With a full charge, at a higher state of charge, there's maybe 2-3% deviation between my SmartShunt and the BMSes. As the discharge deepens -- in a single day -- the deviation increases. It gets worse with a partial charge, especially on days (like this whole week, ugh) where I get just a bit of solar in.

I also have no idea how exactly it interacts with my AIOs (SRNE ASPs). In testing the other day, I basically forced charging (I've got grid tied in, so when it's cheapest, I use it currently) by lowering the thresholds for 'stopping discharge'. I noticed when the cycle started the inverters were told max charging current of 55A; they're capable of 120A grid and 200A max, and I have two in parallel. With 5-280Ah batteries, maxing that out is barely 0.25C. So...I changed the master BMS max charge current from 200A to 180A and poof, full throttle. I really have no idea why, that's just as likely a bug in the inverter firmware, which SRNE seems to guard like the Crown Jewels.

These inverters also don't seem to honor the RCV/RFV timer or it's not sent over Pylontech porotocl, or it's not a value they expect to read. I've set it to 0.1/0.1 to test -- no luck.

I'm going to continue to run this way, hope maybe there's another update. If the BMS was causing my system to not charge at all or create other problems, I'd go back to open loop, but I want to see if this can be made to work.

 

[prosier]

Wow. Yes I know the percentages are off but I am amazed at how tight the voltages are on the batteries! And the cell differences are equally impressive! Nicely done!

At 3.2v this is not really something special I think? I see this a lot here also on my packs.

 

[agtcovert]

Wow. Yes I know the percentages are off but I am amazed at how tight the voltages are on the batteries! And the cell differences are equally impressive! Nicely done!

EVE cells seem to do really well. I have 64, and 16 REPT. First battery in the screenshot is the REPT pack. All 280Ah, charging finished about 4 hours ago.

Current SmartShunt reading. We'll see how that looks in 12 hours.

 

[JJJJ]

I suspect when I get done I might see a little more total voltage variance. The cells differences per battery are very good, however my 8s system will have two CATL and two EVE batteries.

Always appreciate seeing a successful build, reason for my post earlier. DIY is alive and well.

 

[Steve_S]

I suspect when I get done I might see a little more total voltage variance. The cells differences per battery are very good, however my 8s system will have two CATL and two EVE batteries.

Always appreciate seeing a successful build, reason for my post earlier. DIY is alive and well.

All my cells are EVE 280AH cells. I have Originals (2018) through V1, through V3 with N & K series cells. The oldest are Tapped Cells, then cells with Posts to Double Holed V3's. I have 1 Operating Frankenpack with a Mix of V1 & V2 cells (different spacing between the terminals, makes it interesting). And the numbers provided previously are "typical", in fact deviation is less (on the couple) at higher voltage but they never deviate more than what is shown, down to 2.900Vpc (tested this).

The Trick in in the prep & assembly of the packs.
All Cells get bench Top-Charged to 3.650 allowing Amps to drop to 1% of AH Capacity. in Sets of 4
They are then re-topped as a full "set (entire packs cells) and allowed to balance between themselves.
They are all assembled using Flexible Busbars (Meshed Copper for those with Studs or Tapped Holes) and Laminated Copper bars for the V3 Dual Holed cells.
Each pack is Gently Compressed - NOT Stressed -.
Each pack has an identical JK Inverter BMS (200A) and use exactly the same length Silicone Wiring to BMS & Breaker etc.
Each Packs wire harness (sense leads) are exactly the same length (factory), with double folded wire into a properly dry crimped terminal.
Each BMS is Calibrated to 3 decimal places & set properly.
---- the bank is also a bit stringent in assembly.
Each High Grade Kalas Flexiwhip Cable is precisely 111.12mm (43-3/4") long with "Nickel Plated" lugs more $$ (NOT TIN) (All Hex Crimped & Sealed)
Each (+) cable goes to an identical 200A MRBF Fuse at the Common DC Busbar (AMOND 8 post heavy units)
Everything is resistance tested & all match !
* NOTE, Most people do not realize, but every fuse has resistance and it varies by type of fuse & the amperage of the same, it all adds up and can create inconsistencies when configuring a system. Remember, resistance is Additive and does not take much.

The consistent commonality helps to notice even the slightest variances & issues should they pop up because it becomes quite apparent if/when something gets odd.

Anyways, back to Observer mode yet again... Have a Good One.

 

[xylo]

I think this is on topic...
I got a new phone (Pixel 9) today with Android 15. The JK BMS app now has a problem. You cannot see the top and bottom of the app screen.
This means that on the main screen, you cannot use the menu icon in the top left, or the '...' icon in the top right (to get to the System Log).
Obviously, this is a problem. I am curious if anyone else has this issue in Android 15, and perhaps do not update your system if you have not yet.

 

[Hedges]

I had similar problem with some apps on iPhone.
I had made screen image larger for better readability, but that pushed things off the bottom.
Try a global setting to make everything smaller.

 

[HarveyDent]

I think this is on topic...
I got a new phone (Pixel 9) today with Android 15. The JK BMS app now has a problem. You cannot see the top and bottom of the app screen.
This means that on the main screen, you cannot use the menu icon in the top left, or the '...' icon in the top right (to get to the System Log).
Obviously, this is a problem. I am curious if anyone else has this issue in Android 15, and perhaps do not update your system if you have not yet.

The same issue in Pixel 8. This is related with Android 15. On Android 14 was everything ok

 

[doox00]

If you parallel many of these together and you want to change a setting, do you just change it on the master or do you need to change it on each individual bms? If you only need to change it on the master than what if you want to have one of the batteries with a different setting in it? What is the benefit to parallel these together? I have a new solar system install and using the new jk bms's, 16 cell 200amp 2a active balancer ones with firmware 15.33 on each of my diy batteries (with 3 eg4 6000xps).

 

[doox00]

The same issue in Pixel 8. This is related with Android 15. On Android 14 was everything ok

Same issue on my pixel 9 pro xl after updating to Android 15.

 

[agtcovert]

If you parallel many of these together and you want to change a setting, do you just change it on the master or do you need to change it on each individual bms? If you only need to change it on the master than what if you want to have one of the batteries with a different setting in it? What is the benefit to parallel these together? I have a new solar system install and using the new jk bms's, 16 cell 200amp 2a active balancer ones with firmware 15.33 on each of my diy batteries (with 3 eg4 6000xps).

The master BMS will send and control charge voltage etc. it won't change settings on the other BMSes though, but they won't be used.

The master does average the SOC across all batteries connected.

 

[doox00]

The master BMS will send and control charge voltage etc. it won't change settings on the other BMSes though, but they won't be used.

The master does average the SOC across all batteries connected.

Wouldn't it be better to just let the bms in each individual battery maintain itself?

 

[rogerheflin]

Wouldn't it be better to just let the bms in each individual battery maintain itself?

The master only reports the calculated average SOC% back to the inverter running in closed loop with it.

From what I can tell the non-master bmses simply seem to broadcast their status/values on the shared bus and the master listens to it and averages the SOC's from itself and all of the other bmses and reports that to the inverter (if closed loop). Otherwise each BMS seems to do its own thing.

 

[agtcovert]

Wouldn't it be better to just let the bms in each individual battery maintain itself?

It will, as far as cell balance. If you're not going closed loop, then leaving them unconnected should be fine. If you are connecting to the inverter for comms, you should have the batteries communicating with each other.

 

[Hedges]

If you parallel many of these together and you want to change a setting, do you just change it on the master or do you need to change it on each individual bms? If you only need to change it on the master than what if you want to have one of the batteries with a different setting in it? What is the benefit to parallel these together? I have a new solar system install and using the new jk bms's, 16 cell 200amp 2a active balancer ones with firmware 15.33 on each of my diy batteries (with 3 eg4 6000xps).

I think we have to set current limits and cell voltage limits individually on each BMS.
I've only programmed two, and individually, so I don't know for sure.
I'm using JK app on iPhone, connected by bluetooth (and JK PB BMS).
I'm told PC interface is more buggy, haven't tried it.
We ought to be able to read and write settings, clone from machine to machine. Don't know if that is possible.

 

[kommando]

Don't know if that is possible.

Not achievable as yet and with the firmware releases drying up again not likely in the future. The best alternative is to screenshoot the settings on battery one and then enter these on each subsequent battery manually.

 

[Greglsh]

Please can someone help, I received my JK inverter BMS 150a 2A active balancer yesterday, but looking at the BMS it does not have a Positive battery cable connected like the pictures it has a screw terminal





As you can see on mine it has a screw terminal for Bat+, what version is this, is this a v14 or v15 BMS, I am unable to power it as at the moment as I need to build the battery. The issue is the power cable they sent is just a tiny ring lug with a soldered end and is very thin.

 

[JJJJ]

That is your power cable for the BMS. There should be a second piece in the kit with a ring terminal that connects to this. That goes to the positive terminal of your last cell in the build.

 

[HarveyDent]

My test and observations after update :

Inwerter: AIO ( Easun SMG II 5.6kW 48V)
JK BMS: v15.33
Cell RCV: 3.5V (56 V)
SOC 100%: 3.499 V
Cell RFV: 3.350 V (53.6 V)
RCV Time: 30 min
RFV Time: 6h

SOC fluctuated between 100-99%. Shouldn't it show 90% at 3.350 V (53.6 V) ???
I'm curious how the SOC readings will behave during discharge. For now, it looks ok.

 

[AshleyL]

My test and observations after update :

Inwerter: AIO ( Easun SMG II 5.6kW 48V)
JK BMS: v15.33
Cell RCV: 3.5V (56 V)
SOC 100%: 3.499 V
Cell RFV: 3.350 V (53.6 V)
RCV Time: 30 min
RFV Time: 6h

SOC fluctuated between 100-99%. Shouldn't it show 90% at 3.350 V (53.6 V) ???
I'm curious how the SOC readings will behave during discharge. For now, it looks ok.

What are you talking about? LFP cell voltage doesn't work that way.
LFP cell has a very flat 3.2v - 3.35v range between 10% - 90% of its SOC.
If the charging is stopped after reaching 3.5V, it will quickly settle down to 3.35, its natural rest voltage after full tank.
From your open loop graph.....that way of multiple rebulk charging till 3.5v in a day might be detrimental to your cell health in long term.

Remember, there is a funny bug with v15.33 firmware. If the SOC percentage at the BMS drop to 99%, it will trigger rebulk charge again and ignore the RFV Time.
However, most Voltronic clone or OEM inverters will ignore charge command and only start charging again after the SOC drop below 95%.

 

[HarveyDent]

What are you talking about? LFP cell voltage doesn't work that way.
LFP cell has a very flat 3.2v - 3.35v range between 10% - 90% of its SOC.
If the charging is stopped after reaching 3.5V, it will quickly settle down to 3.35, its natural rest voltage after full tank.

Thx. Clear.

Remember, there is a funny bug with v15.33 firmware. If the SOC percentage at the BMS drop to 99%, it will trigger rebulk charge again and ignore the RFV Time.
However, most Voltronic clone or OEM inverters will ignore charge command and only start charging again after the SOC drop below 95%.

If I interpret it correctly, then the RTV time counter is not reset because the inverter "fixes" this error?

 

[AshleyL]

If I interpret it correctly, then the RTV time counter is not reset because the inverter "fixes" this error?

The RCV/RFV counters or charging algorithm are restarted right after the BMS SOC% falls down to 99% with v15.33 firmware. In my Growatt case, yes, my Growatt inverter "fixes" the error. 100% -- 95% -- 100%. My inverter basically ignore the RCV and not charging unless the SOC% fall below 95% after reaching 100% for the first time in the day.

Interesting data. After overlaying your graphs, it appears that your Easun SMG II seems to respect the charging command sent by the BMS as it restarted the charging the moment at 99%. Which battery protocol do you use? Pylontech CAN/RS485? I am using Growatt CAN protocol. Do you cycle your battery or you just keep it topped up?From the graphs, seem like you are keeping it at full tank all the time.

Since most of us cycle our battery pack, the SOC% tends to drop to 99% whenever there is a sudden load from running equipment. Not ideal because the RCV Timer will be running again and keeping our cell at elevated voltage until the RCV Timer expired.

 

[HarveyDent]

Pylontech CAN/RS485? (..)Do you cycle your battery or you just keep it topped up?From the graphs, seem like you are keeping it at full tank all the time.

I confirm, Pylontech. Only for the needs of this test I used the SUB (SolarUtilityBattery) program in the inverter to charge the battery from the power grid. Now winter is starting in Poland. In everyday use I use the SBU (SolarBatteryUtility) program and use the entire battery capacity. Inverter jumps onto the power grid at 49,6V or 12% SOC.

At this point the closed loop works fime for me. Unfortunately I didn't do any tests on an earlier version of the software, but I assume that it works correctly from V15.33

 

[Steve_S]

V15:33 is ONLY posted into Andy's Google-drive
Nothing at all has been heard from any sort of "Official Source" whatsoever.
Nothing is posted on ANY of the JK Web Sites wether official or not or vendor clone sites.
JK still refuses to interact with ANYONE with the exception of occasionally tickling Andy's Fancy for some reason (maybe to make it appear they are attempting to help while not actually doing squat).

The Spooky part... Stuff is appearing (firmware, Phone & PC Apps) that are supposedly JK Stuff but random places, none official. BEWARE !!!

 

[AshleyL]

I confirm, Pylontech. Only for the needs of this test I used the SUB (SolarUtilityBattery) program in the inverter to charge the battery from the power grid. Now winter is starting in Poland. In everyday use I use the SBU (SolarBatteryUtility) program and use the entire battery capacity. Inverter jumps onto the power grid at 49,6V or 12% SOC

Oh, I am using SBU(solar-battery-utility). Try switching to SBU and see if the inverter charger will charge back the battery if the Master BMS SOC% drops to 99%, assuming if it is possible since you mentioned winter is coming. As far as I know, Easun inverter is often based on Voltronic inverter design, so I would expect it will use the same 100%-95%-100% algorithm.

I will be testing the v15.33 with Solis S6 inverter later using Pylontech CAN protocol.......as soon as possible if my AWG0 battery cables arrive......

I recalled there is someone here with SMA Sunny Island inverter too....I wonder how will the v15.33 react with SMA inverter since the inverter is charging based on voltage and the inverter only use the SOC percentage to control the switching from battery to grid.

At least in Andy's Victron inverter, he confirmed the charging RCV cycle begins again the moment JK Master BMS SOC% drops to 99%.