I didn't top balance my packs before installing

[Andy]

Once the 2 bats were balanced then once 3.449V is reached it takes 20 mins to get the delta V back from typically 0.012V to 0.008V. Even with a gap of 3 weeks between reaching 3.449V it only takes 40 mins.

Good to know what's required in the future. Thanks.

[Tinbum]

Well, I've been charging as fast as possible to give more time at the top each day for balancing. It definitely seems to be getting on top of things finally.

Here is a chart of the difference between top cell and bottom cell in the pack. The trend is in the right direction. Thanks for all assistance.

That's great.

[Andy]

It's been getting 4-5 hrs a day balancing so it must have been more out than I thought.

As a side thought, I said the JK is balancing at only 0.5A. I was wrong. It is actually 2A because although it is losing energy transferring from the highest voltage cell to the lowest voltage cell it is still removing 2A at a time from the top cell. It is constantly moving 2A at any time so it is in fact a 2A balance. Just a shame that it is so inefficient.

[Kommando]

When you consider the earlier version of top balancing was to use a lab power supply to take each cell to 3.65V or to use the same lab supply to take all 16 cells in parallel to 3.65V, which with 280AH cells could take weeks depending on the starting V, then the JK is a big improvement. Once the long first balancing is done it has enough power to keep the cells all in check without getting alarms from runner cells or needing to add in an external balancer. To get rid of the watching paint drying feeling its best to walk away and just let it do its thing.

[Ronski]

It is unbalanced at low voltages here because clipping my charger on to the cell bumped the terminal and loosened the nut. It took me a few hours to work out why I was getting random voltage spreads when I didn't normally have any Luckily I don't have a huge charge or discharge current at this time of year.

For me this sets alarm bells ringing, knocking a terminal really shouldn't loosen it, unless you knocked a cable that happened to turn it.

You have correctly torqued all the terminals haven't you?

I've never topped balanced my cells, just let the BMS do it's thing as per Andy's settings pretty much, batteries are cycled everyday.

[Andy]

It is unbalanced at low voltages here because clipping my charger on to the cell bumped the terminal and loosened the nut. It took me a few hours to work out why I was getting random voltage spreads when I didn't normally have any Luckily I don't have a huge charge or discharge current at this time of year.

For me this sets alarm bells ringing, knocking a terminal really shouldn't loosen it, unless you knocked a cable that happened to turn it.

You have correctly torqued all the terminals haven't you?

I've never topped balanced my cells, just let the BMS do it's thing as per Andy's settings pretty much, batteries are cycled everyday.

They are all torqued to 4Nm. I am slightly confused how it got so loose. I did bump the balancing cable as I was trying to attach cables from my charger. Going back through my records I can see the connection was probably not perfect a while ago as the voltage measurements on that cell are a bit jaggy. They got worse over time. My initial checks post tightening where to use a flir with a very high current input. Once I tightened it a few weeks ago the readings became cleaner. I'm going to go over all 160 terminals again to make sure they are secure. I am a bit weary of tightening too much as unfortunately the second batch of cells I had didn't have the correct laser settings welding the studs on to the cells. The laser was accidentally left on the settings for a smaller cell size. Colin's support was very good and he replaced the two that detached with very little pressure.

[Ronski]

I don't think 4Nm is enough for an M6 (standard is about 9Nm), but given how weak some of those terminals are I wouldn't want to go more than 4Nm. I intentionally got the LF280K with the twin threaded terminals for that reason, much better design although it doesn't seem to have caught on quite so much, I still see a lot of single stud versions.

I have my balancing set to start at 3.45v, I charge to 56.5v (3.53125v per cell), absorption time is set to 2hrs, then it floats at 56v (3.5v), been running this way for a long time now.

Cell diff is set to 0.005v, the two packs are currently 0.003v and 0.004v.

PS. I use the old JK BMS not the inverter BMS, no communication to my Victron system - open loop.

[Andy]

I don't think 4Nm is enough for an M6 (standard is about 9Nm), but given how weak some of those terminals are I wouldn't want to go more than 4Nm. I intentionally got the LF280K with the twin threaded terminals for that reason, much better design although it doesn't seem to have caught on quite so much, I still see a lot of single stud versions.

I have my balancing set to start at 3.45v, I charge to 56.5v (3.53125v per cell), absorption time is set to 2hrs, then it floats at 56v (3.5v), been running this way for a long time now.

Cell diff is set to 0.005v, the two packs are currently 0.003v and 0.004v.

PS. I use the old JK BMS not the inverter BMS, no communication to my Victron system - open loop.

They were meant to be good for 6Nm but unfortunately I don't want to tempt fate any more. Most YouTube videos for eve 280K seem to work to 4Nm. I'm just going to regularly monitor the setup.

[Joeboy]

Double nut them giving a locking nut scenarion and your 4nm torque will be intact. Seems low to me too but tbh I nipped mine up by hand.

[Kommando]

If like like Envision cells you have M6 aluminium studs then 4nm is correct as aluminium is weaker than steel.


MAX TIGHTENING TORQUE ALUMINIUN in Nm
Head Type 6mm
HEX Head 7.00
Dome Head 4.00
Countersunk 4.00

https://www.bst-bolt.uk/content/10-tightening-torque

Same goes if double nutting, its the aluminium stud that snaps.

9nm is for a steel stud.