New cells arrive today! Spend 100 hours top balancing...

[kc130up]

So it worked (I think). I set the BMS overvoltage to 3.61v and it cut of charging at 3.61v.

The problem was the cells are out of balance, so while 1 cell was 3.61v another was 3.46v. The passive balancer isn't fast enough to keep the high cell at bay while the others charge, so the BMS kept triggering overvoltage.

I lowered the balance trigger down to 3.55v but it still couldn't bleed the high cell fast enough.

I lowered the EG4 charge current down to 15a and that got me another 4 hours or so before a cell went overvolt. Then I lowered the current to 10a and that got me another 2 hours or so. I believe 10a was as low as the Eg4 would go.

Then I removed the BMS & balance leads, reconnected in parallel, and got back to charging 4 cells per PSU.

It was a nice feeling to use the sun to "pre-balance" the cells using the inverter. Since I moved the balance lab to the shed, some of the post-balance work is also from the sun since we had a mostly full day of sun today.

 

[robbob2112]

One thing I really like about this board is it's refreshing to watch posters embracing their full nerd. Most in my circle want to talk football 24/7.

That is for the chit-chat forum down at the bottom - but beware - the comments are unmoderated and very harsh and off the cuff... some members that are perfectly capable of giving really good advice and are nice people outside of that are instead crude, rude, and disgusting at times...

I stay out of there except for specific posts that are just general stuff.

 

[Dagoth Ur Does Solar]

I’m in the camp of, if you’re going to compress/fixture the cells, assemble them as they are to be used and charge them. If you need to get an active balancer to make that happen, so be it, even if the balancer is only temporary. I’ve seen cells swell a bit on the first charge. I don’t feel that it’s a good idea to compress once that’s happened.

This is the way I am leaning for my first battery. When my cells arrive, I think it's probably better for me to just let them sit until I am ready to get them into the fixture they are going into and get them under compression before I bother doing anything else.

General idea I am getting from reading here is the same as you- definitely do not want to compress a cell that has expanded.

 

[robbob2112]

Interestingly enough I have charged 4 - 24v packs in a row and they haven't expanded at all... but I was charging using a bench power supply with a 20amp max current. So, that might have made a difference... I will be putting them in a frame when installed.

 

[kc130up]

I had a scare this morning! Though I may have just pissed away $1100 in LFP cells....

I left the chargers running overnight in the shed. The JBD will stop charging the cells if the temp gets too low. These 10a chargers have no such protection. The temp last night got down to 28F.

I ran outside at 6am and flipped the breakers off to stop the charging!

From what I have read, cells can be charged at low temps "near freezing" as long as the current is low. Some sites say stay under .5c if its below 32F. .5c for these cells would be 140a, so a charge of 7a should not have damaged the cells.

The cheap plastic thermometer showed near 35F. The anker F2000 in the shed said it was 45F. The BMS on the other battery says it was 35F. Not sure why the anker was 10 degrees higher.

I suspect the freezer running and the 4 PSU's may have added heat to the shed. Even though it was ~28 outside, temps inside the shed were higher. The

Did I do damage?

Got me thinking about how I will keep the cells warm when winter really gets here. I can either heat the shed with a baseboard hydronic heater like I do in the renal property basements to prevent pipes from freezing or look into a heater of some sort just for the LFP cells.

Ideas anyone?

 

[robbob2112]

I seriously doubt there was any harm at all from that low a current

 

[42OhmsPA]

I'd build an insulated enclosure around the cells and heat the small area. Seed starter mats, heat pads, etc.

You could use something like the attached to control the heaters or the power supplies.

 

[kc130up]

I assume the cells wouldn't be too heavy to sit on top of the seed heater pads?

 

[sunshine_eggo]

From what I have read, cells can be charged at low temps "near freezing" as long as the current is low. Some sites say stay under .5c if its below 32F. .5c for these cells would be 140a, so a charge of 7a should not have damaged the cells.

No site says this. 0.5C is often the maximum charge current allowed. period.

Some sites say 0.05C, and even that is questionable. Unless you have manufacturer data permitting this, avoid it.

I'm using EV grade NMC with Yttrium stabilized cells good down to -20°C. I still avoid any charging below freezing except for what little the system allows to bleed through with transient load changes.

.05C would be 14A, so you're still below that, but it sounds like you likely were above freezing.

Did I do damage?

Almost certainly not.

 

[kc130up]

learning here... Above you said .5C max.

The datasheet for my first set of cells say nominal charging is 1C, peak is 2C.



The data sheet for the cells for the battery I am currently says .5P which i assume is the same as .5C.



Am I reading the sheets wrong?

 

[42OhmsPA]

I assume the cells wouldn't be too heavy to sit on top of the seed heater pads?

I wouldn't sit them on them directly. I'd have a small air space between them and the mat.

 

[sunshine_eggo]

learning here... Above you said .5C max.

The datasheet for the my first set of cells say nominal charging is 1C, peak is 2C.

View attachment 291026

No site says this. 0.5C is often the maximum charge current allowed. period.

I didn't say always...

The data for the cells for the battery I am currently says .5P which i assume is the same as .5C.

View attachment 291027

Am I reading the sheets wrong?

EVE datasheets went from Current to Power ratings in the recent past. 1P is constant POWER, so 896W for an hour. This means the voltage and current change to maintain constant power through the discharge. Most equipment isn't setup for this. This probably happened because these are EV cells, and EVs care more about power than they do about current.

Prior datasheets/models were constant current. So 1P is = 1C when V * 280A = 896W or when V = 3.2V Above 3.2V, XP is < XC and below 3.2V, XP > XC. Yes. It's confusing.

If you TREAT P as C, you should be fine, but the datasheet support my assertion that some cells are limited to 0.5C, or at least that's their "standard/preferred" charge rate even if max is higher.

Point is that those are clearly not allowed below freezing.

 

[robbob2112]

Seems like I read a white paper on the topic of low temp charging and they basically said once you get more than a few degrees below 0c there is a knee where the max charge current falls off a cliff.... And there is a similar knee at the other end where once the battery temp is to hot the max charge falls off a cliff... but between those it was pretty much flat.... You would have been in the flat approaching the cliff after which it drops to the 0.015C rating mentioned above...

 

[fnnwizard]

learning here... Above you said .5C max.

The datasheet for the my first set of cells say nominal charging is 1C, peak is 2C.

View attachment 291026

The data for the cells for the battery I am currently says .5P which i assume is the same as .5C.

View attachment 291027

Am I reading the sheets wrong?

Your not reading it wrong but most these specs won't detail out what happens. Adding to what Egg wrote, charging at 0°C is a slippery slope to be avoided if possible. We ran into issues back in 2008, yes back in 2008 even with A123 cells (A123 were 10C with no clear defined charging specs back then).

The consensus is at 0°C you want to be no more than .1C charge rate. Then not charge at all below that. Thats a very binary line.

In laymans tems, the ions are extremely lazy at that temp and don't want to move.
By trying to pushing them to their destination, they'll just revolt and destroy the destination. The more technical term is lithium plating and it's permanent. The damage results in much higher IR, = loss of capacity and varies depending on the actual C rate, temp gradient of the pack and of the cell.

 

[kc130up]

"Pre-Balancing" to 3.5v saved weeks! Already down to less than 1a.




I got a seed heating pad and placed two between the cells. It wasn't that cold last night, but it made me feel better!

 

[kc130up]

Solar assistant shows the gradual decrease in power of the DC chargers as they approach 3.65v...

 

[time2roll]

So it worked (I think). I set the BMS overvoltage to 3.61v and it cut of charging at 3.61v.

The problem was the cells are out of balance, so while 1 cell was 3.61v another was 3.46v. The passive balancer isn't fast enough to keep the high cell at bay while the others charge, so the BMS kept triggering overvoltage.

I lowered the balance trigger down to 3.55v but it still couldn't bleed the high cell fast enough.

Need to reduce the charging voltage to stay under 3.60 on all cells to avoid cell over volt shut down. Leave balancing overnight. Up the voltage the next day and let it continue to balance.

Could be doing this while in service.

 

[robbob2112]

The simplest way since you have dc psu... Just connects to the low cells one at a time with 3.65v and bring them up... walk through all cells... when they are down to only 0.5a or so draw from the psu move to the next cell.... you don't need to disconnect... but i would turn off the balancer and charge on the BMS while you do this.

Once they are all topped up...will take 30 to 120 minutes per cell...you can do multiple at a time.... then put 3.6v x cell count across the battery...

I am going to throw out that instead of doing a battery with 1 bms and 16 cells for 4s it is better to make 4 separate 4s batteries and use 4 BMS... it will help avoid this type issue

 

[kc130up]

That last mile....

 

[42OhmsPA]

That last mile....

View attachment 291529

I'd call them done.

 

[740GLE]

Holding that little current is def fork stuck in them done.

How long have they been sitting at 3.65?

 

[robbob2112]

agreed - done -- I stopped individually charging cells when the current was half an amp... once they settled a bit I went to charging the string through the BMS..

When I went to bed I changed these
balance start voltage 3.43v
balance delta 5mv

Charger off.... just let it bleed all the cells to spot on... then normalized the setting to 3.5 for start and 15mv for delta .... and drained then charged over 50% of the SOC.

 

[sunshine_eggo]

"Pre-Balancing" to 3.5v saved weeks! Already down to less than 1a.

View attachment 291302


I got a seed heating pad and placed two between the cells. It wasn't that cold last night, but it made me feel better!

You were likely done before this.

According to the datasheet of the "P" rated cells, they are "full" at 3.65V when subjected to 448W of constant charge power,.

3.5. Charge Calibration

At the ambient temperature 25±2℃, rest for 5h; Discharge at a constant power of 448W to 2.5V, and rest for 30 min; Charge at a constant power of 448W to 3.65V, and rest for 30min.

448W/3.65V = 122.7A, i.e., when charging at 122.7A, the cell is full at 3.65V. The new specs say constant POWER, not constant voltage.

I really think this is why it's so common for quality EVE 280Ah cells to test closer to 300Ah. It's because they've "enhanced" their cycle life by increasing capacity, but they specify a charge criteria that never actually gets the cells to max capacity, BUT will deliver 280Ah. All of my cells charged to 3.65V @ 0.05C (14A) tail current all tested around 304Ah. Charging to full is the most stressful operation to the cells, and they've set the charge specification to leave some % on the table.

 

[kc130up]

Balance done, battery reconfigured as 16s, BMS connected, and battery is connected to the bus bar with the other battery.

I changed the BMS settings, so the SOC is wonky. It will fix itself tomorrow. Supposed to have full sun all day!



I discharged the new battery to ~53.8v to get it to the same voltage as the other battery before i flipped the switch on. Immediately the new battery discharged into the old battery at ~2.3a.

After a few minutes the voltage(s) had equalized. Both BMS's showed (about) the same voltage.

I connected a heater and noticed only 1 BMS showed a draw. I used the Anker to charge at 1000w and the bigger battery was taking more of the current.

I must read up on how the BMS's will behave in parallel so that I can spot any potential issues.

 

[sunshine_eggo]

They will behave as though they have no idea the other is there.

Many BMS do not recognize small currents and will not register a draw even if there is one.

"FET software locked" may mean you need to turn both chg/dischg off and then back on.