andrew96a

so after all this time later I have finally worked out why it wouldn't communicate!!it is a SHbms so the SHbms software did indeed connect to it on rs232 and communicate giving all the battery voltages etc! so that port worked!!
the anenji battery was bought BRAND NEW with the anenji 11kw inverter and never ever would they communicate on the rs485 port when the inverter was set to do so! always error 4b, no communication on the inverter!


how difficult can it be to just connect the rs485 cable supplied with the battery, select lib2 in the battery settings, flip dip1 on the battery on and away ya go!!!!

well not that easy as it turns out! this has never worked for me! always error '4b' communication error so used it with manual battery settings.

we all now know the 'support' Anenji give once purchased, the standard for them...... NON EXISTANT!!! so just beware whatever you buy has no warranty at all once you buy it! even AVG antivirus now flags there website as a scam!!

so now i bought a proper Victron inverter for the house this I decided to move THIS Anenji inverter to the workshop with some pylontech batteries I already have recently. but again the Anenji inverter would never connect and error 4b continued when Lib2 was selected. I know what your all going to say.... wiring and dip switches!!!! but nope that's all correctly set and done! the manual shows it should be that easy!!!


this has continued until recently when I tried again to connect it on the RS485 bus, nope still no joy! so thought lets dig deeper, so this may help someone in the future with RS485 troubles!!

Asking Ai which communicates first it came back with inverter polls with data and the battery reply's when it sees the correct data with its own.... ok somewhere to start!!!

I next got Ai to write a python script at 9600 baud rate to listen on the rs485 but using my usb to rs485 waveshare adapter. plugged it into the Anenji inverter, selected Lib2 in the menu....... and..... nothing... absolute silence from the bus and of course the '4b' error!

So broke open the anenji inverter and pulled from it the communication board.


it has a SP485, a standard RS485 transceiver with a smaller marked 'U6' ic above it, so looked with my cheap handheld oscilloscope and the input to the Rs485 pin on the SP485 as this is what should be being transmitter out the other side of the transceiver ic.


ok that's good there is data there! so why is it not passing through the transceiver? well pins 2 and 3 are connected together on the SP485 and are direction control, level high (+5v) for the chip to allow transmitting and low to enable receiving. these pins were stuck at 0v so never ever is the chip being told to transmit the data!

The direction control pins 2 and 3 are connected to a small chip marked U6 and have the markings on it as C08j


search and search the internet i did but nothing on the chip it actually is. BUT it did come back to being a 7408 AND gate which clearly it is not! Ai to said it was a AND gate to from the markings as that is also used on a Texas ic to identify a AND gate in exactly the same package! removed it to confirm the top right pin was not connected to anything and this turned out to be the case, so definitely not a AND gate


so to get data to pass through pin 2 and 3 need to be high, well luckily U6 power in pin was next to the output pin, so decided to put a zero ohm link so pins 2 and 3 on the sp485 would be high, the handy place to do this was from the space U6 had been removed!

powered the inverter up and out popped the strings on the rs485 sniffer program

7E 32 30 30 32 34 36 36 33 30 30 30 30 46 44 41 39 0D
7E 32 30 31 32 34 36 36 33 30 30 30 30 46 44 41 38 0D
7E 32 30 32 32 34 36 36 33 30 30 30 30 46 44 41 37 0D
7E 32 30 33 32 34 36 36 33 30 30 30 30 46 44 41 36 0D
7E 32 30 34 32 34 36 36 33 30 30 30 30 46 44 41 35 0D


and that just kept repeating! so that's good! proves the firmware is actually trying to talk and correctly and AI decoded it to it looking for any battery 1 to 4 with just the checksum changing at the end to keep it right.

more searching and came up with a blank on what that small chip U6 was!!
then with more searching came across a mention of a chip that has automatic direction control built in! found a Texas instruments ic that was a drop in replacement that just needed pins 2 and 3 held high for it to be fully automatic directing control... well I already did that so half way there!!
could not find that chip anywhere in the UK so did some more searching and found a max13488 which was easily available here in the UK, and would be a drop in replacement just fine also only needed pins 2 and 3 held at high level to work which I had already done!


so now I connected the inverter to a pylontech battery (dip 1 on of course) connected my sniffer and switched on.
to my relief I could now see the rs485 data from the inverter AND pylontech battery reply repeating!


(highlighted area is one send and reply)

the inverter was happy!!! no error 4b and even 'Li' was illuminated on the battery path!

so eventually cracked it! communication is now fine and working after 15 months of never working! so it was nothing to do with me at all, it was faulty all along!

this may help someone in the future, my advice is if there are any problems suspected on the rs485 is to monitor it and see if any data is flowing first.


So I would NEVER NEVER EVER by Anenji or one of the other cheap Chinese inverters again. as the saying goes, buy cheap buy twice! good luck to you all with Anenji stuff!!

I am using the cable recommended by solar assistant and connecting through the third connector on my JK inverter bms 
https://solar-assistant.io/help/battery/jk-inverter-bms 

PBMS is for Pace BMS - for Pylontech use BatteryView. (available in the download section)