5.2kw installed that can do above that on cold days, hitting <100w at times during bad times, or <300w average where we live.
Got so bad for a 2nd year in a row now, me considering more panels every winter, that I gave up on that idea now. What is fits sunny days, more panels is not going to give me the ROI I’m used to/expecting.
2nd year in a row, more panels, for me, has become a HOBBY level spend.
That bad.
Slightly off topic, that link you just posted in the 4x4 Forum about the JK BMS that’s been upgraded.
Do you mind posting it here somewhere as well? I really like that you can see which cell from from which battery is the highest and which is the lowest.
I don’t suppose a Pace BMS can give that info through to VRM?
Got it here from Mark. 
Andy’s Google Drive: JK-BMS – Google Drive
That is a good thing to note. You shouldn’t stop when the SOC says 100%. The SOC could be wrong, after all. You stop once a proper charged voltage is reached and have been held for some amount of time. Some batteries are also a bit more “honest”, if they have a low cell somewhere that is impacting one of the modules, and it cannot get it to 100% because of balancing issues, it will report the lower SOC (even at the higher voltage), which again is why the voltage is the more important factor here.
I’ve seen batteries continue to charge long after their SOC counter reached 100% (because it suffered drift in the opposite direction), and I’ve also seen batteries continue to discharge for an hour after reaching 0% (although this has to do with the cells being slightly better than the BMS knows).
Some people with BMW i3 Rex vehicles have also suffered the wrath of SOC estimates being wrong. The car turns on the petrol engine when the battery is around 6% SoC. Now imagine the SOC drifted and 6% is really empty. It was fixed with a firmware update, but a couple of people actually got stranded as the battery ran empty without the petrol engine starting.
For me this has been a week of BMS issues. Pylontech mostly. I’m actually a little frustrated with how some of this stuff works, or doesn’t work. They want a minimum of 52.5V over 15 cells. That is 3.5V per cell. Sounds fine, but even a small imbalance (and I mean really so small that many manufacturers would not bother during assembly) can have you with 14 cells at 3.48V and one cell overvoltage (>3.6V). Proper dynamic voltage control on a BMS is really kinda sorta non-negotiable these days.
3.600 - 3.480 = 0.120v Delta?
Over 18 cells I’m gunning for staying below <0.085v delta when charging hard, high amps.
Titbit 1: Reminds me of when we brought those cells in.
Got the serial #', ah and resistance tested per cell.
Overall, batch matched. Definitely.
So I thought about it.
Then box for box, cut open, check the minute serial # printed (in fancy ink), and grouped them in batches to each buyer, 17 cells per batch/battery.
Thought it was no use one gets a cell rated at 294AH with a lower resistance, the next cell 286AH with a higher resistance, in the same bank, over 6000 cycles, my “gut” said.
So far nary a problem that I am aware of, been told.
Titbit 2:
It helps a LOT, for some, when the bank exceeds a pre-set Delta, to instantly drop the charge amps in DVCC immediately, incrementally higher drop in charge amps, if the problem persists.
Gives the balancer chance to do its work easier to “lower” the Delta rather than at the top.
Just some continued observations that still holds true after all this time.
Deltas mean nothing once the highest cell goes over 3.5V. That’s the entire issue with LFP batteries, that once you go above 3.45V per cell, even tiny amounts of differing capacity results in huge deltas.
Pylontech batteries tries to get all their cells to 3.55V (the advertised charge voltage is 53.2V), which is playing exactly in that area of the park where things are very spiky and deltas grow large. And that is exactly my contention… why can’t we play a little closer to the flat ground? The battery is still over 99% full…
Edit: Also, the delta between the highest and lowest cell is also not of much meaning once you get this high up. The far more important number is the average delta between the rest of the cells after you excluded the high cell. There is always one cell that jumps out, always. What you want is for all your cells to have the same voltage right up to the point where that one cell jumps out. Once that one cell jumps out, you’re done for today. Arguably, this practice isn’t very good for that one cell either… and since it is already the cell with the lowest capacity in the pack, this makes it worse over time.
What typically does the SOC calculation? Is this a BMS function??
It’s one of those “it depends” things. You get various levels of BMS functionality. At the most basic level, the task of the BMS is to make sure the cells are not damaged, so it has to prevent overcharging (which also has a fire risk) or over-discharging. Your most basic BMS will do only that.
Some BMSes will also do balancing, but you could also think of this as a separate function: The balancer circuit is in many ways an additional and separate “device”, it might even be a separate board within the same enclosure.
On top of that, more advanced BMSes will also do SOC tracking, and the most advanced ones have some sort of communication with the outside world on top of all that.
A good example of this would be the older Victron SmartLithium batteries combined with a VE.Bus BMS. The batteries are 12.8V modules, and every battery has its own balancer board which deals with the 4 cells inside that battery. The “BMS” is really just a thin layer that prevents overcharging and overdischarging, or charging at low temperatures, and the SOC tracking… that’s done by an external shunt. This is a perfect example where the three functions are in separate places.
The Lynx BMS, is really only a protection device combined with a shunt (and this is the shunt product on its own, sans BMS). The balancing is elsewhere (in each battery module).
To add to your post …
Some BMS’es also have a shunt in the mix to facilitate SOC calculation.
Errata:
Using the Delta, being one value to monitor instead of 15/18 individual cell volts, to trigger the software lowering the charge amps of the bank to facilitate the Balancer.
There’s pretty much just two ways to do it. Hall effect sensor, or shunt.
Some BMSes have two hall effect sensors, one for low current situations, one for higher current situations. Without that second sensor, SOC drift can be severe at lower power levels. OrionBms has a dual range sensor for accuracy low down.
OrionBMS (in the context here) can actually use either a shunt or a hall effect sensor.
I find that using the voltage of the highest cell is all you need. It doesn’t matter how far away it is from the lowest cell.
As an example, consider two situations:
A: Highest cell is at 3.3V, lowest cell is at 3V.
B: Highest cell is at 3.5V, lowest cell is at 3.45V
Situation A has a 300mV delta… but there is no reason to stop charging (yet). Situation B has a 50mV delta… and you should stop charging.
Must say, the longer I watch the bank operating the more I’m leaning towards:
Reasons being:
But it is not for everyone, I get that.
Makes sense.
I was more looking at sorting the problem long before one gets to >3.45v.
I reckon this is the most solid way to check SOC.
The BMV monitors Wh in and out of the battery and compares it with the total capacity of the battery…
Indeed, when the cells are closer together, your delta is also more representative of what is really going on. Although even then, the deviation from the mean is probably more important. But that aside… for what I wanted to do, which is about voltage control rather than current control, it was sufficient to just look at the difference between the highest cell, and 3.5V. If the higest cell gets too close to 3.5V, start penalising the charge voltage until it balances out. Much better for that cell as well.
I wouldn’t - I think this process will take time. My 2.5 year Pylontechs do this every now and again.
No. Give it time as Mark said.
Join the club
This is what one of my US2000 batteries started doing last week, after two years of hassle free operation:
Fortunately Pylontech support is a pleasure to deal with: I send them the logs that shows something is wrong with the battery and that I have not abused it, and they organise a replacement.
