I have had my bank of 8 US5000 for a couple of years and I just got 8 more to add to the system. I’m off grid in Alaska with limited sun. I have a generator and an EG4 Chargeverter. There seem to be two trains of thoughts on how best to add these new modules.
A: use my generator and EG4 to bench charge each of the new batteries using a small current until the current drops to very low then shut them off. Then wait for a sunny day and use the generator and the sun to get my existing batteries up to 100% SOC. Then connect all the batteries together. The down side of this is it will take a lot of fuel to run the generator to supply a small charge if I do each battery individually. I might instead wire the new 8 batteries up into a dedicated bus bar and provide 8 times the current until I get them close, then individually top each battery off. That might save time and fuel.
B: disconnect my existing batteries from my system, wire them in with the new batteries, turn them on, and let the batteries balance each other out. Once the LEDs match and enough time has gone by, wire them all back into the system. This seems like it would be quicker, wouldn’t require the generator, and is what most people seem to be doing on YouTube to introduce new batteries into the system.
The new Pylontech should have an SOC of at least 50%… which means there won’t be such a big voltage difference with LiFePo batteries.
The communication between 16 batteries should work via the BMS.
I would simply connect the batteries together… switch them ALL off and then switch them back on via the master after connecting the communication.
But I would definitely use one of the new batteries as the master…
That’s good to hear, that sounds pretty simple. I just put the 8 new batteries on top of the shelves with the other batteries in them. Will make new shelves for them next summer. The new batteries are all 49.5V except one is 49.4V.
A German proverb says … “Exceptions prove the rule.”…
take a look at the “MultiSIB Control” program and the necessary RS232 connection cable for the “console” connection from the master …
then you can easily monitor all of the pylontechs cells …
Using a network-to-USB converter, you can then read all values via the network and don’t always have to take your laptop to the shed
If you’re already using Node Red anyway… you can read out all the cell information and then evaluate and display it. Then, of course, the program wouldn’t be necessary…
Have fun building shelves and doubling your pylon tech capacity
What??!! Without all those fancy cables I can get individual cell info in Node Red? I’m going to dig into that. So much gold hidden in Victron’s Node Red modules…. I need to go click on the tree and see if I can find those values, great info!
Don’t be so hasty
The Victron system cannot use the information that Pylontech does not send for NODE RED. The highest and lowest voltages, etc., are processed in a register somewhere.
Everything else can certainly be done using some kind of workaround…
You may be able to retrieve the data from the system via Mqqt and an API interface…
You don’t need a RS232 cable to master and MultiSIB to have all that info…
You can read all necessary info, for each and every battery, over CAN bus, from master, from within Venus OS.
And I mean here: module voltage, SOC, min and max cell voltages, min and max temp, mosfet and bms temp, status of the moduie, etc.
With 16 modules, you will have all 240 cells info, individually.
Exactly… that’s what I mean.
You can’t get ALL the information you get via MultiSIBControl and RS232 via the OS… only the information that the OS already has from the BMS. But this information is usually sufficient… unless you want to convert a 65-inch television into an information centre for the 16 Pylontech and display ALL cell information.
Yes, you can… over CAN.
Without the need of an additional RS232 connection.
Directly at the Venus OS command prompt.
For batteries with firmware newer than the beginning of 2021.
I got them wired in today. I let the existing batteries discharge down to about the same voltage 49.5V as the new batteries. I wired them together with the only load on them being a 48V to 12V Orion. When they came on, all 16 batteries were showing the same number of LED lights. The new batteries were all taking a charge from the existing batteries when they were powered on. I let them be like that for 15 minutes to stabilize a bit, then turned on the Cerbo GX and all the batteries registered and the lowest to highest cell voltage was within .02V or so (although I suppose it’s not accurate at 59% SOC or whatever the batteries were at). I then turned on the all system and the batteries began charging from the sun. I will be interested to see 100% SOC so I can see what the disparity in cell voltages are. I see that section in details in Venus, and I think it should be the lowest and highest cells throughout all 16 batteries and not just the master? I hope that’s the case.
That is correct.
In the picture, a certain cell of battery 6 has the lowest cell voltage (but you don’t know which cell exactly) and a certain cell of battery 15 has the highest cell voltage.
I thought I’d get them all to 100% today so they could do any balancing they needed, but at 1:30 pm they were at 97% SOC with 15KW of solar available or so that was being throttled. It took about two hours to get to 99% at around 3:30 pm with an abundance of solar being throttled, and after two hours of pure sun until the sun went down at 5:30 and they remained at 99% SOC and never got to 100%. Pretty frustrating to have all that power available unused and it still couldn’t get full.
With our short days in Alaska we’ll likely never see 100% batteries so they can actively balance given this enormously slow rate of charging. 30KW of solar panels and yet for four hours it wouldn’t charge the batteries with anything more than 150 watts and yet still couldn’t get to full.
Don’t worry, they will get to 100% in the next days. That happened to me too.
Throttling is because of the charging algo implemented, in order to not get overvoltage alarms.
And more of that, because the balance is passive with about 60mA maximum.
Unfortunately that was our only sunny day forecasted in a week or so. Yesterday the most they got was 90%. I hope the fact that I put them online and it will be weeks before they ever see 100% isn’t an issue for them.
Don’t worry, they will get in sync.
Have fun with them.
PS.
With Pylontechs, as soon as any cell of a battery is above 3.36V and the difference between min and max voltage cells on that battery is greater than 30mV, the balance will start. So, you see, the balance could start as low as 50.4V (3.36 x 15) bank voltage.
Therefore, slowly, they will all get to that “harmony”…
Thanks, that is good to know. I had thought they didn’t start active balancing until they hit 100% SOC which seemed odd to me but I don’t know a lot about batteries. I’ll stop stressing about it then, thanks!
