• 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

New Sunny Island and JK PB BMS set up - problem and solution.

Calvin98

Solar Enthusiast
Joined
Jun 29, 2022
Messages
430
New Sunny Island and JK PB BMS set up - problem and solution.

We’ve had a US 6048 Sunny Island 12kw split phase set up for many years now. It was set up as back up for when grid power was lost with a lead acid forklift battery. That set up worked well for years. We recently upgraded to the JK PB BMS because it’s supposedly has closed loop communication that works with the Sunny Island. Thanks to some forum members here, we got some really nice CATL 12 V battery packs that we wired together for 48 V packs. Built in heating pads, welded, busbars, compressed metal case, etc. Total of 93 kWh and 8 JK PB 2A BMS boards.

The short version.
We set everything up and kept getting LOBAT1 warnings W755, W756, and external BMS communication loss warnings. The warnings were due to a can bus communication failure on the Sunny Island part. Updating to the latest SMA firmware for the Sunny Island fixed the problem. There is a new firmware version available as of May 2024, 7.304 that among other things addresses can communication error that we had. This completely fixed the problem. I can say with high confidence that the Sunny Island works and communicates well with the JK PB BMS. There are others issues that I’ve heard such as the JK factory making firmware that creates problems as well as solves them so pick firmware you like and try to stick with it if possible.

The longer version.
We kept seeing W755 and W756 and some other warning messages. External BMS communication loss was another error we saw. Depending on the circumstance, when you get these messages, the Sunny Island thinks the battery is at 0% SOC and potentially shuts the whole system down. One shutdown was so serious that we had to disconnect everything from the inverter for half hour before it would boot up again. The 755 is LOWBAT1 warning. Normally, for lead acid these are set at say 40% for number one, 20% for number two and say 5% for low battery level three warning. Warnings are, of course, a bit less severe than failures. Warnings just tell you what’s happening failures actually would lead to a shut down. Anyway, this is supposed to be the short version. We installed a raspberry pie and some monitoring software so that we could monitor all the Sunny Island parameters on Home Assistant. We were able to see that we kept getting these low battery warnings sporadically in the evenings and at night. Generally. First, we lowered those parameters to 20% 10% and 3%. That did not solve the problem. Got the same error messages. We ended up setting all the battery protection levels to 0%. But that did not solve the problem either.

At one point, we swapped the master and slave Sunny Island to see if the problem was in the master only. That did not solve the problem.

There’s also a thing where the Sunny Island uses 100 ohm terminating resistors and the JK uses 120 ohm terminating resistors. We made new resistors to match what the Sunny Island had. That did not solve the problem.

We used a normal ethernet cable with an RJ 45 for communication to Sunny Island from the JK master in BMS. It was running in the same path as the AC and DC cables and was just fractions of an inch away from those. Thinking that the cable may be too close to those AC and DC wires and getting interference, we move the communication cable away from those AC lines. That did not solve the problem.

Thinking it still may be the cable, we replaced it with a shielded ethernet cable and ran it far away from the AC and DC lines. That did not solve the problem.

Thinking maybe the master JKBMS might be the problem we swapped the JK master and the slave. That did not solve the problem.

Thinking that also the BMS to BMS communication cables might be the problem. We also swap those out for short shielded cables. That did not change anything either.

Another option was that we were using firmware on the JK that was slightly newer than other users. We downgraded the firmware to that older version on the JK. That did not solve the problem. You need to have a code generator to install older firmware. Here’s a link below for Google that will generate those codes that are valid for one hour. With that code generator, you can install any firmware old or new.

Info below on how to set up Sunny Island communication with the Home Assistant.
https://diysolarforum.com/threads/s...-6048us-monitoring-with-home-assistant.91280/

With the raspberry pie, we can get the following data from Sunny Island. Home Assistant: see below.
LINK COMING

JK bms code generator to downgrade/force update firmware.

All shown codes change at every full hour. They are valid for

Current JK-BMS-Monitor Force Updating code

https://mirofromdiro.github.io/JK-firmware-code/

In the end, one of the forum members here, Hedges, pointed out that there was another new Sunny Island firmware available as of May. We checked and ours that we were running was actually one version prior to that (7.302). Since Sunny Island was updated with what we thought was the latest lithium battery firmware, we didn’t really look to see if there was another firmware that was available to us. We did install it and immediately saw the results. No more 0% SOC warnings on the battery anywhere. In the picture below, you can see clearly on the left side with all the vertical lines the error messages that Home Assistant recorded before with the old firmware from SMA, and then on the right side, no more vertical lines with misleading SOC presents using the new SMA firmware. By the way. HA was recording every 5 sec, so it might not have recorded every 0% SOC event. Over 2 weeks later and everything is still running well. No glitches or hiccups. Now to deal with some JK issues - cell 01 over charge and abnormal co-processor errors. Fine tuning battery voltage settings, etc.

1727365951012.png



Hope this helps someone in the future. We spent two weeks trying to troubleshoot this. We always figured it was some sort of software issue and it turns out it was. Once we knew what the problem was, the solution was obvious. Of course hindsight is always 20/20. I can say it with very high confidence that the Sunny Island and the JK PB BMS models do work with closed loop communication. If you have Sunny Island’s and want a good BMS that works the JK PB BMS models are good ones to look at.
Our fallback position had this not worked would’ve been just to use the open loop with voltage. As some users here also do.

Thanks again to Hedges, Kommando, 1201, Disturbedfuel, and the others who helped us along the way. That is the most difficult troubleshooting I’ve had in a long time, but also one of the most rewarding since we did actually find the actual problem and were able to fix it.


 

Attachments

  • 1727365908647.png
    1727365908647.png
    60.2 KB · Views: 1
Im so glad it got fixed and that all it took was a firmware update.


So ultimately do you prefer closed loop to open loop? I have a jk inverter bms but i haven't installed it yet because open loop is working for me
 
SMA list that firmware as dated May 2024, but I believe that is just the date the webpage was last updated even if the content is unchanged. The 7.304 was available when I downloaded it in Jan 2024 and the internal docs in the zip are dated Apr 2018. SMA do other odd unhelpful things such as only making the current firmware available so if you have not kept a copy of your old firmware then there is no way back unless you find the file elsewhere.

Changelog for 7.304

Firmware release v7.304/7.300 (only for SI 5048U/SI4548-US/SI6048-US)
=======================================================================
- Includes all features from previous firmware versions
- Bug fixing:
- Improved CAN communication between Sunny Island and lithium-ion
batteries

- Improved data transmission (battery temperature) from external
BMS to Sunny Island
- Improved warning handling. No warning when no battery temperature sensor
is installed with lithium-ion batteries

Also not listed is that 7.304 breaks comms with some battery inverter/chargers that used the SMA docs to develop the comms, Midnite ?
 
Im so glad it got fixed and that all it took was a firmware update.


So ultimately do you prefer closed loop to open loop? I have a jk inverter bms but i haven't installed it yet because open loop is working for me
Yes, me too! What a relief. I have been sitting on the sidelines looking at batteries, prices and compatible BMS for several years now. Did not want to take the plunge until there was a good option that was proven wot work with the SI's.
Prefer? Too early to tell. Being able to get the SI and JK data is great.
 
SMA list that firmware as dated May 2024, but I believe that is just the date the webpage was last updated even if the content is unchanged. The 7.304 was available when I downloaded it in Jan 2024 and the internal docs in the zip are dated Apr 2018. SMA do other odd unhelpful things such as only making the current firmware available so if you have not kept a copy of your old firmware then there is no way back unless you find the file elsewhere.

Changelog for 7.304

Firmware release v7.304/7.300 (only for SI 5048U/SI4548-US/SI6048-US)
=======================================================================
- Includes all features from previous firmware versions
- Bug fixing:
- Improved CAN communication between Sunny Island and lithium-ion
batteries

- Improved data transmission (battery temperature) from external
BMS to Sunny Island
- Improved warning handling. No warning when no battery temperature sensor
is installed with lithium-ion batteries

Also not listed is that 7.304 breaks comms with some battery inverter/chargers that used the SMA docs to develop the comms, Midnite ?
Right, yes.
 
We’ve been 100% off grid for the last six weeks now. 100% problem free. Sunny Islands and JKBMS is doing their thing. Knock on wood. Meanwhile, the utility had a power outage a few days ago. Would not have noticed it until I heard the neighbors generators running.

We do have a issue with the JKBMS reporting Cell 01 overcharge protection, but that’s not keeping the system from running.
 
We’ve been 100% off grid for the last six weeks now. 100% problem free. Sunny Islands and JKBMS is doing their thing. Knock on wood. Meanwhile, the utility had a power outage a few days ago. Would not have noticed it until I heard the neighbors generators running.

We do have a issue with the JKBMS reporting Cell 01 overcharge protection, but that’s not keeping the system from running.
That's great. Glad to hear that.

Overcharge protection- what do you have your max charging amps set to on the bms?
 
Maximum charging current in each BMS is set to 100 A. Discharge is same at 100 A. But since we have 8 BMS set up, the maximum I’ve ever seen on one battery is maybe 30 A. The Sunny Islands maximum charge I’ve seen is also like 250 A but that is again also going into 8 JKBMS. Average charge rate is 10-20 A or so. We get that Cell 01 overcharge protection on almost all eight of the BMS’s for some reason. Have not figured out why yet. Doesn’t impede their functionality. It is just a warning message as far as I can tell. Would be nice to get to the bottom of it though.

At a 100 amp charge rate, we should be able to put 800 A into the BMS but the Sunny Island system can’t do that much.
 
Last edited:
Maximum charging current in the BMS is set to 100 A. Discharge is same at 100 A. But since we have 8 BMS set up, the maximum I’ve ever seen on one battery is maybe 30 A. The Sunny Islands maximum charge I’ve seen is also like 250 A but that is again also going into 8 JKBMS. Average charge rate is 10-20 A or so. We get that Cell 01 overcharge protection on almost all eight of the BMS’s for some reason. Have not figured out why yet. Doesn’t impede their functionality. It is just a warning message as far as I can tell. Would be nice to get to the bottom of it though.

At a 100 amp charge rate, we should be able to put 800 A into the BMS but the Sunny Island system can’t do that much.
Hmm that's very strange. I wonder why only cell 01.
 
Right. Never any other cell. And across all eight BMS. Another friend who has the same BMS saw the same error message. Very curious situation. Sounds almost like a software issue. Looking at the actual voltages of the cells, they all look normal.
 
SI Update.

We were running off grid successfully for over two months now. After having solved the BMS communication issue above everything was running perfectly. Then today I hear all the computer UPS’s buzzing and all the computers off. Wondered why so many would fail all at the same time. They all showed 120 V but when I tried to turn them on, they wouldn’t turn on again. Short story is the Sunny Island were putting out the correct voltage 120 per phase 240 total but the frequency was at 70 Hz for some reason. This was at 7 AM in the morning. Sunny boys were just beginning to wake up. I did see on one of the sunny boys walking by and it showed the error message 701 meaning the frequency was out of range. Didn’t pay much attention to it at the time. Since the batteries generally fill out about 1 PM, the Sunny Island is frequency shifting in the afternoon anyway. Then to counteract the increased frequency the Sunny Island runs everything at 59 Hz to compensate. I thought it may have been due to that. Then thinking it may have been slave two or slave three, I turned those both slaves off and just ran with the master and slave one. I did hear some loud louder humming in slave two yesterday. Louder than slave three or slave one or the master. Not sure if it was due to that. Sunny Island was still putting out 70 Hz. I disconnected all the AC and DC sources from the Sunny Island waited a minute and then connected everything again. Turned on the master and slave one and voltage was OK and the frequency was OK at 60 Hz also. So far, so good. But a few minutes later, the frequency was rising to 65 and then 70hz again so I turned everything off again. Put a support ticket in the SMA and hoping to hear something from them. Hopefully it’s something easy to fix. Just when I was getting really confident in the SI system, this happens. All of my concerns about reliability that we had in the beginning with the communication issues, shutting the Sunny Island’s off were 100% gone as of yesterday. I felt the system was reliable and ready to be left on its own for long periods of time. The system was running flawlessly for the last 60 days. Today confidence is gone again. Anyway, hoping for the best. If anybody has any ideas on this, please let me know.
 
SI Update.

We were running off grid successfully for over two months now. After having solved the BMS communication issue above everything was running perfectly. Then today I hear all the computer UPS’s buzzing and all the computers off. Wondered why so many would fail all at the same time. They all showed 120 V but when I tried to turn them on, they wouldn’t turn on again. Short story is the Sunny Island were putting out the correct voltage 120 per phase 240 total but the frequency was at 70 Hz for some reason. This was at 7 AM in the morning. Sunny boys were just beginning to wake up. I did see on one of the sunny boys walking by and it showed the error message 701 meaning the frequency was out of range. Didn’t pay much attention to it at the time. Since the batteries generally fill out about 1 PM, the Sunny Island is frequency shifting in the afternoon anyway. Then to counteract the increased frequency the Sunny Island runs everything at 59 Hz to compensate. I thought it may have been due to that. Then thinking it may have been slave two or slave three, I turned those both slaves off and just ran with the master and slave one. I did hear some loud louder humming in slave two yesterday. Louder than slave three or slave one or the master. Not sure if it was due to that. Sunny Island was still putting out 70 Hz. I disconnected all the AC and DC sources from the Sunny Island waited a minute and then connected everything again. Turned on the master and slave one and voltage was OK and the frequency was OK at 60 Hz also. So far, so good. But a few minutes later, the frequency was rising to 65 and then 70hz again so I turned everything off again. Put a support ticket in the SMA and hoping to hear something from them. Hopefully it’s something easy to fix. Just when I was getting really confident in the SI system, this happens. All of my concerns about reliability that we had in the beginning with the communication issues, shutting the Sunny Island’s off were 100% gone as of yesterday. I felt the system was reliable and ready to be left on its own for long periods of time. The system was running flawlessly for the last 60 days. Today confidence is gone again. Anyway, hoping for the best. If anybody has any ideas on this, please let me know.

Have you made any recent changes to the system? Bms?
 
Nope. Haven’t touched it since we fixed the communication problem by updating the Sunny Island firmware over 60days ago. That’s one of the things that is so surprising.
 
Kommando suggested that I try reversing the master and slave and see what happens. Hedges suggested I try setting it to VRLA battery with the appropriate cell voltages to see what happens. Maybe it is some issue in the BMS itself? Although I’m not sure why the BMS has anything to do with the AC frequency. Waiting to hear from SMA and then I’ll give those options a try if I need to.

Really weird thing. I could not for the life of me figure out why so many UPS power supplies failed at the same time NY I couldn’t get them running again. Then once I saw that the frequency was at 70 Hz that sort of answered. I guess the lesson is if there’s an error message or something is happening there’s a reason behind it. Don’t just brush it off. Once you know the solution, then of course then it all seems obvious. We spent weeks on that communication issue and threw the kitchen sink at it. Once we knew what the problem was everything was so obvious and simple. I really hope this is something like that.
 

diy solar

diy solar
Back
Top