SOLVED: Batrium charging issue, seeking advice on debugging

[callmeburton]

Problem:
Charging > 20 amps on my 280Ah battery sometimes causes cell1 (negative) and often cell16 (positive) to spike over voltage.


Patterns Noticed:

• When first charging every cell ramped up in voltage at the same time. It wasn't until voltages neared 97% SoC that cell 1 would be over active. The SMA SI was pushing 5.2kW into the batteries and everything seemed fine.
• (need to confirm) Removing AC IN and discharging to 50% then allowing for AC IN is when the problem presents itself last.
• (need to confirm) Restarting whole system with AC IN enabled near start up removes problem
• Cell1 / 16 have highest bypass .. all other cells 0
• When I noticed the problem, and adjusted the amps down, I had to go down to about 10 to have the problem stop, then I ramped back up but couldn't go beyond 20

Things I am thinking of trying

• swap out k9 or another one I have lying around (I have a couple waiting for future packs)
• move cells1 and 16 to 2 and 15 to see if problem follows them
• Confirm patterns above which I have noticed at least twice so far
• Check voltage at cell one to confirm reading in batrium and adjust batrium as required

Current settings: (see attached for screen captures) These are mainly stock

The system is currently in this state and I have not tried to restart it yet.

 

[rhino]

Swapping cells will tell you the most of where the issue is.

 

[callmeburton]

Swapping cells will tell you the most of where the issue is.

That is what I thought as well, seemed like a pretty good way to rule out the cells being a problem.

 

[OffGridInTheCity]

What's the awg of the charge wire to the battery? I know it's unlikely you undersized wiring but I've noticed my BattGo showed significant voltage drops on the + / - (end cells) of a 7s when I used the sense wires to discharge a 7s battery - e.g. overloading the wires a bit - to bring the overall battery voltage down. I presume (in my case) it's a resistance thing related to too-small-wires and you're situation makes me think of that.

 

[callmeburton]

What's the awg of the charge wire to the battery? I know it's unlikely but a way too small wire carrying the >20a might play havoc with voltage readings.

2/0 welding wire
Current wire diagram for this setup is here https://diysolarforum.com/threads/i...bank-and-sma-si-load-center.52698/post-707395

 

[callmeburton]

Just thought ... I should have it ramp up a bit and look at the voltages manually to make sure I don't need to adjust the bms reading on the cells in question as well. Adding it to the list.

 

[callmeburton]

UPDATE:
Removed AC last night to let the cells discharge a bit. Noticed cell 1 (neg) was bouncing between 3.02 and 3.27 ... so I went to swap out the K9 unit and double checked the torque on the battery nuts ... the negative was super loose ... no clue how that happen ...

Put the original k9 I just took out back in and testing stuff now.

Going to keep an eye on that nut ... my cables might be setup in such a way that the cycles cause that nut to loosen up ... might need some Loctite lol

Keeping AC Loads off for now ... just going to watch it charge without any draw to see what it does. So far it is accepting the 112amps without issue.

 

[OffGridInTheCity]

Similar to 'too small wire', a loose connection can cause havoc on voltage readings. Will be interested to see if tightening this fixes the hi voltage on each end of the series of cells. Part of my interest is because Batrium is involved - I'm a long time (5yrs and counting) Batrium user on my powewall.

 

[callmeburton]

Similar to 'too small wire', a loose connection can cause havoc on voltage readings. Will be interested to see if tightening this fixes the hi voltage on each end of the series of cells. Part of my interest is because Batrium is involved - I'm a long time (5yrs and counting) Batrium user on my powewall.

So far it is working better than before ...
Key thing I see now is voltages ranging from 335-3.5 as it is top balancing with default ramped settings ... the requested Ah is only 6.0 as it thinks it is at 94% SOC ... when the voltage starts to bypass I think it drops to 0.6Ah ... first time having the cells this high in voltage as last time I only let them get to 3.35 ... oh and one of the 3.35 cells is #1

 

[callmeburton]

Slowly but surely every cell is getting to the bypass voltage. I suspect tomorrow morning it will be really close to "done" if you will and I can set the max SoC after letting the battery rest (I assume)

 

[callmeburton]

UPDATE: All resolved, will continue to keep an eye on the terminal bolts.

sometime around or before 02:50 last night the cells all bypassed and the Batrium marked that as 100% SoC (ignore the timestamps as the unit doesn't know its time zone)

Looks like some cells bypassed a good 70watts over the time frame it took to get everything even again ... but from the original posting time to two/three days later being pretty balanced this is a pretty good result to me.