Growatt Inverter + Pylontech Batteries : BMS Comms Guide

[djacobs]

Hi All. Installed new growatt and 1 x Pylontech 3000. When following the instructions mine is still flashing 04 and 20. The Li does come on. Other setting also display percentage. Do I need the firmware upgrade. If so could someone please email me. [email protected]. 

 

Thanks All. 

Daniel

[djacobs]

On 2020/08/07 at 1:04 PM, djacobs said:

Hi All. Installed new growatt and 1 x Pylontech 3000. When following the instructions mine is still flashing 04 and 20. The Li does come on. Other setting also display percentage. Do I need the firmware upgrade. If so could someone please email me. [email protected]. 

 

Thanks All. 

Daniel

Sorted Thanks. But why does the PV disconnect when battery is full. I want to use Panels during the day, as first priority. What I see is household will use battery to around 50v and then only switch on Pv again to charge to 53v. 

[djacobs]

[Elroy]

On ‎2020‎/‎03‎/‎04 at 8:21 AM, SiliconKid said:

NOTE TO ALL

We have now established that BOTH of the firmware updates that were provided to me by Growatt in China ARE required to get the BMS comms working correctly between the Growatt SPF5000 inverter and Pylontech batteries.

The two firmware updates they sent me are for two different things inside the inverter (Thanks to @Coulombfor taking a look at the firmware and confirming this):

• The ARM chip.
• The DSP.

Feedback from @Coulomb:   "So there are two firmwares in there, a TMS320F2809 (same DSP as most Axperts), and an ARM chip (Axpert Kings have one of these in the removable display)."

The firmware pack that I am providing to people who PM me asking for help contains both of those firmware updates, and they are both necessary to sort out all the BMS issues properly.

If you don't have both of those firmware updates installed, one of the symptoms is that BMS comms will appear to be working and there will be no errors BUT the SOC level will be stuck at 100% and the batteries will never get charged, which means that if you have solar connected it will never kick in and be used to charge your batteries.

And we have also confirmed that the firmware resolves the issues for a single standalone inverter and more than one inverter connected in parallel.

 

Hi SiliconKid, is it possible to send me the firmware upgrades as well pls. el[email protected] Thanks

[djacobs]

They are not responding. 

[SiliconKid]

@Elroy, @djacobs: I've PM'd both of you the firmware link.

Edited August 11, 2020 by SiliconKid

[Treschen]

On 2020/07/28 at 11:29 AM, Tsa said:

Yes, exactly this. This is also in Li mode where I have seen 1) My datalogger stop collecting data points for a random period of time and 2) Had the power trips. 

In USE/US2 mode none of this has happened. 

So yeah growatt confirmed the operation of turn off solar is to protect the battery. Normal operation according to them 

[j5_immortal]

On 2020/06/10 at 4:22 PM, j5_immortal said:

Do you have setting 14 set to CSO, if on OSO, it will not switch to grid charging. The master battery dip switch must be 1000, the rest you can leave on 0000

 

On 2020/08/09 at 2:07 PM, djacobs said:

Sorted Thanks. But why does the PV disconnect when battery is full. I want to use Panels during the day, as first priority. What I see is household will use battery to around 50v and then only switch on Pv again to charge to 53v. 

Hi djacobs,  it has to do with the the correct setup of priorities as well as the voltage levels.  Your batteries takes over too soon if the voltage levels for switchover to battery supplying the power is too high.  I have 4 x US2000 Pylontech batteries with 12 x 360W PV.  I do not use the Pylontech BMS to control the setup as it limits the charging current for the 4 x batteries to 25A for the stack which is way too low to get battery stack charged up during the day.  The following settings are what I have setup and it allows the PV to supply power the whole day and charge the batteries without being switched off:

01  - SBU
02  - 80A
04 - SDS
05 - US2
11 - 20A
12 - 47.7v
13 - 48.3v
14 - OSO
19 - 53.3v
20 - 53.3v
21 - 47.5v

As you will see, when no loadshedding schedules are implemented, I use setting 14 on OSO (Solar charging only, no ghrid charging), Setting 1 is on SBU: Solar supplies power to load, after sunlight hours, batteries will take over the loads until battery voltage falls down to setting to 47.7 V (setting 13), then grid takes over the load until sunlight and solar takes over again.  During the day if loads are too high or aclouds block solar power, batteries takes over load until it falls to 47.7v, then grid takes over if insufficient sunlight, as soon as battries charges up to 48.3V (setting 13) battries takes over the load again, and so forth....  from my graph below you will see the effect of this setup:   Batteries works minimal during sunlight hours and I get most out of the PV's, grid only kicks in when battery voltage reaches 47.7v under load.

During regular load shedding schedules you would want to keep your batteries from draining during the day if loads are to high or cloudy skies. During these times would need to change the change Setting 01 - Sol, 12 - 50, 13 - 53.3.  

Hope this helps
See an example of my graph over 24hrs attached


 

Edited August 15, 2020 by j5_immortal

[j5_immortal]

On 2020/08/09 at 2:07 PM, djacobs said:

Sorted Thanks. But why does the PV disconnect when battery is full. I want to use Panels during the day, as first priority. What I see is household will use battery to around 50v and then only switch on Pv again to charge to 53v. 

Also, even with my settings as above the power from the PV will be shut off or be limited when the batteries are fully charged and there is not enough demand from the load.  If this happens on a daily basis, it means you have abundant solar energy available that are going to waste, it is a call to you to expand your battery storage capacity so you can utilize the additional solar energy and convert it into storage for night use.  On my graph above the PV output is not limited because I have expanded my batteries to a stack of 8 x US2000 recently and there is not enough power from the 12 x PV to fully charge the battery stack and therefore supplying full capacity.  I am currently expanding with additional 12 x PV...

In addition to the settings mentioned above during regular load-shedding you will need to set program 14 to OSO to keep batteries charged even with grid power if required, so you can have power from batteries for when load-shedding occurs after sunlight hours.

 

Edited August 15, 2020 by j5_immortal

[djacobs]

6 hours ago, j5_immortal said:

 

Hi djacobs,  it has to do with the the correct setup of priorities as well as the voltage levels.  Your batteries takes over too soon if the voltage levels for switchover to battery supplying the power is too high.  I have 4 x US2000 Pylontech batteries with 12 x 360W PV.  I do not use the Pylontech BMS to control the setup as it limits the charging current for the 4 x batteries to 25A for the stack which is way too low to get battery stack charged up during the day.  The following settings are what I have setup and it allows the PV to supply power the whole day and charge the batteries without being switched off:

01  - SBU
02  - 80A
04 - SDS
05 - US2
11 - 20A
12 - 47.7v
13 - 48.3v
14 - OSO
19 - 53.3v
20 - 53.3v
21 - 47.5v

As you will see, when no loadshedding schedules are implemented, I use setting 14 on OSO (Solar charging only, no ghrid charging), Setting 1 is on SBU: Solar supplies power to load, after sunlight hours, batteries will take over the loads until battery voltage falls down to setting to 47.7 V (setting 13), then grid takes over the load until sunlight and solar takes over again.  During the day if loads are too high or aclouds block solar power, batteries takes over load until it falls to 47.7v, then grid takes over if insufficient sunlight, as soon as battries charges up to 48.3V (setting 13) battries takes over the load again, and so forth....  from my graph below you will see the effect of this setup:   Batteries works minimal during sunlight hours and I get most out of the PV's, grid only kicks in when battery voltage reaches 47.7v under load.

During regular load shedding schedules you would want to keep your batteries from draining during the day if loads are to high or cloudy skies. During these times would need to change the change Setting 01 - Sol, 12 - 50, 13 - 53.3.  

Hope this helps
See an example of my graph over 24hrs attached


 

Thanks for the feedback. I am not connecting to grid. Will the settings stay the same. 

[j5_immortal]

4 hours ago, djacobs said:

Thanks for the feedback. I am not connecting to grid. Will the settings stay the same. 

Yes, I believe it should work for you as my aim is to function totally off the grid as well with these settings.  Over weekends when my load footprint is lower I am already of grid.  

 

Edited August 15, 2020 by j5_immortal

[djacobs]

I see the L1 light is flashing, is this nornal. 

[djacobs]

2 hours ago, djacobs said:

I see the L1 light is flashing, is this nornal. 

Also since setting to US2 I have lost coms. 

[j5_immortal]

This is correct, the Pylontech BMS will not communicate and the Inverter is monitoring the battery voltages and the charging under the user mode US2 setting.  The Li light will be flashing, that is normal.  The US2 setting works well for me.  

[djacobs]

Great thanks

[Roland]

On 2020/02/17 at 10:54 PM, SiliconKid said:

I've recently purchased and installed the following:

1 x Growatt SPF 5000TL HVM 5kVA/5kW Hybrid Inverter.

2 x Pylontech US3000 batteries.

1 x Pylontech cable kit with RJ45 BMS comms cable included.

I've just spent 5 days communicating with Growatt in China to figure out why I could not get the inverter to communicate with the BMS on the master battery properly.

After many, many emails back and forward, and far too many hours of research online, I finally got it right today and I think it is worth sharing this information here, in case anybody else needs it, because I couldn't find the right information anywhere on the internet, all in one place, explained properly.

It turns out that the Growatt SPF5000 has to be connected to the Pylontech batteries differently to the way Axpert inverters are connected to Pylontech batteries.

Everything I was reading online was saying that you need to plug the RJ45 comms cable into the CAN port on the master battery and then into the BMS port on the inverter, and then you need to set the battery type to Li (setting 05 on Axpert and Growatt inverters) and then choose Li profile L02 (also apparently the same for Axpert and Growatt inverters).

It turns out, the Growatt does NOT communicate via the CAN port. It communicates via the RS485 port that is positioned below the CAN port on the master battery.

And in addition to that, it requires a 9600 baud rate, which you can set using the 4 small white dipswitches on the master battery.

For the Growatt you must set those dipswitches to:  1 0 0 0 (ON OFF OFF OFF)

When connected to the CAN port and put into Li (L02) mode, the Growatt inverter faults and gives an error 20 and an error 04 constantly.

When connected to the RS485 port with the dipswitches set to 1000, comms is immediately established and it works as intended.

I also installed 2 firmware updates on the inverter, that were sent to me by Growatt in China (who were very helpful through this entire process despite not actually giving me the information I actually needed to solve this), but in retrospect I actually wonder if those firmware updates even made a difference because I was able to put the inverter into Li (L02) mode before applying those firmware updates anyway, I just didn't know about the RS485 port and the dipswitches.

Recommended configuration process

1. Turn inverter on but run off battery only. Disable A/C input and  A/C output (use isolator switches if you have them).

2. Go to setting 5 and select battery type = Li. Then select profile L02 (This is specifically for Pylontech batteries). Push ESC to return to home screen. Go back to setting 5 and verify that it was saved as Li.

3. Now shutdown the inverter and batteries completely. Turn the batteries off using the on/off flip switch, not just the red button on the master battery.

4. Now make sure that the end of the comms cable on the battery side (it will probably have a little sticker on it that says BAT) is plugged into the RJ45 port labelled RS485 that is below the CAN port. And also make sure that the other end is plugged into the RJ45 port labelled BMS on the inverter and NOT into the other RJ45 port that is labelled RS485 (Yes, I know, it's weird, just go with it).

5. Now set the 4 little white dipswitches to: ON OFF OFF OFF (1000).

6. Now turn the batteries back on and then turn the inverter back on.

You should no longer see any fault codes or the red fault light.

How to tell that the inverter really is communicating with the BMS properly

1. You should now see a little Li symbol next to the battery icon on the home screen of the inverter, and that Li symbol should NOT be flashing. When it's not flashing and is solid it means comms is working.

2. If you go and look at most of the battery values on the status screens or in settings related to battery level (SOC) you should now see percentage values where there were voltage values before. Setting 21 is a good example of that.

I hope this helps somebody else because the information available online with regards to the Growatt routers is of no help in this regard and most of the information available relates to Axpert inverters, which appear to be very similar to the Growatt at face value, but communicate with the Pylontech batteries differently from what I can see.

 

 

 

 

 

Hey All,

 

Does these settings include the SPF5000ES version?

[Zimnismoboy34]

funny so i am investigating this whole pylontech and growatt setting saga ...
Firstly seems the chinese just cant get this setup right .. victron a breeze in comparison .

Secondly when charging off solar i thought i was doing something wrong as well the system would charge to 100% solar charge controller off ... then it sits drawing down on the batteries till it reaches 94.8% soc .. and solar clicks back in again if the load goes away the batteries charge up again to 100% and solar off again .. what a poor implimentation.

Secondly because i dont have a pylontech battery im working backwards and trying to figure out why the max charge rate i have is 30Amps ... on setting 02 .. and if i get my trusted code to work on setting 03 then i get a full charge of 140 amps which fed by solar is only 50amps till about 95% where it ramps down to 60amps by any means useless once again it switches off solar at 100% and then decides to turn on again at 94.x% .... useless once again .

So far standby applications this is perfect who wants a lithium battery sitting on float all day long unless you want it to age before its time ..

So in the best words i can use on the forum even growatts support lacks and really lacks definition we are talking black and white tv here with valves ..

if anything possibly seek to get a different brand till they eventually sort out their ins and outs .. 
I have not tried the likes of phocos yet .. but willing to give it a shot if their pricing was a little sharper .
I have the 5000es and tried the canbus and the rs485 and both not working . Manual says nothing about which setting to use .. both for canbus and for rs485 .. tried all .. literally 
i have the latest version of the 5000TL HVM firmware if any one wants it ... it changes the charge symbol to A CC in bold .. heavens knows why and the menu has some changes which i have asked for explainations of and again no response .. i am building into our bms board for our battery and find victron so much easier it works slight tweaks but overall their implimentiation so far is the best ....strangely pylontech use the same code victron does ...lol 

[Zimnismoboy34]

just a quick update got it working on canbus .. after abit of fiddling .. the menu settings needed to be adjusted correctly for canbus to work .. seems to be talking now and soc is correct . charge current also reporting correct . will update what it does at 100% take note i am using the 5000es .. not the 5000tlHVM although i still have access to the hvm .. my thinking is that it cant support float correctly using 485 and will end up microcycling the battery on solar ..noy good .

[Tsa]

Interesting stuff @Zimnismoboy34  I also got the lastest major FW update (500.08) and a BMS update, but in my case it started that solar on and off cycle on 91%!

Was cycling solar on at 90% SOC and off again a few seconds later at 91% SOC...then on again at 90%....crazy!  Usually would happen at 100% then down to 95%. Still also the issue where I am still losing datapoints for hours. Again in the same time where this solar on and off cycle is happing.  THis is on the SPF5000TL HVM

Edited August 21, 2020 by Tsa

[Zimnismoboy34]

i presume that the rs485 and the communication network to the wifi module are some how on the same data path silly but anything is possible . if they are indeed . and they some how had limited legs for comms on the MCu then it would make sense but why .. oh but why .. there is so much better that can be done on these inverter systems . i just wonder if it will ever be done . rant over this would explain why the wifi module drops off and comes back online, it is just a push device that recieves info and pushes it to the growatt server . a api protocol listening device can be made up to check the comms and see if its the wifi module or the bus network internally . i have my money on the bus network .. 

[Zimnismoboy34]

confirmed float works as it should . the stages from cc to  c4 strange  unless it is mean to read cv .. then changes to fc when the battery reaches 100% so confirmed this model does work as intended.

 

[Zimnismoboy34]

ok forget that .. seems float is useless not maintaining just the same with all the others . Though solar does feed loads .. the battery sits with a current draw of 45 watts for some strange reason and is now sitting at 96% ... at 4 pm .. hmmm again another example what its meant to do and what it does are two different things altogether .

[jykenmynie]

2 hours ago, Zimnismoboy34 said:

ok forget that .. seems float is useless not maintaining just the same with all the others . Though solar does feed loads .. the battery sits with a current draw of 45 watts for some strange reason and is now sitting at 96% ... at 4 pm .. hmmm again another example what its meant to do and what it does are two different things altogether .

Is the inverter not drawing its own power from the batteries still?

[SiliconKid]

The inverter will be drawing some power constantly. 

I'm running battery backup only and no solar and I can clearly see how the inverter draws from the battery until it hits around 95% and then charges it back up to 100% again. 

[Johan-Henry]

Good day all. I have the Growatt 5000TL HVM (firmware 500.07/002.03) with wifi and currently running with AGM batteries and a few solar panels. To be honest, it is working quite well. I am looking into upgrading to lithium batteries in the near future and stumbled onto this thread. It seems the Pylontech batteries have a good reputation and it seems that most of the issues regarding the BMS comms with the Growatt is sorted with the firmware I have, however I am unsure of this 95% - 100% solar to battery switching all the time being described here. Is this considered to be normal behaviour for these (cheaper) inverters or is Growatt the only ones doing it? Will it not be detrimental to battery life in the long run? Also the loss of comms issue, is this only related to the Dyness batteries or does it also occur with the Pylontech with BMS comms? Thanks

Edited August 24, 2020 by Johan-Henry
added a question