Be cautious ordering generic blue wrap LFP cells

[john61ct]

If you haven't rigged up a connector to actually **do** the balancing, you might as well be using the higher rate PSU

at under 10A charging you **are really** risking harmful overcharging, even at 3.45Vpc

At this point I'll recommend parallel 1S charging, setpoint 3.55V, and if under .05C, just do one hour at a time, maybe rig a timer,

then let them settle before checking resting V.

3.45Vpc **resting** really is Full-full, and the sitting paralleling is getting them well balanced.

After done, break them up each isolated, and see if any drop more / faster than others. Not saying anything can be done, but possible heads' up as to the weak link.

 

[hour]

If you haven't rigged up a connector to actually **do** the balancing, you might as well be using the higher rate PSU

at under 10A charging you **are really** risking harmful overcharging, even at 3.45Vpc

At this point I'll recommend parallel 1S charging, setpoint 3.55V, and if under .05C, just do one hour at a time, maybe rig a timer,

then let them settle before checking resting V.

3.45Vpc **resting** really is Full-full, and the sitting paralleling is getting them well balanced.

After done, break them up each isolated, and see if any drop more / faster than others. Not saying anything can be done, but possible heads' up as to the weak link.

The cells have been in parallel for over an hour (since my last post at the end of page 2). I initially set the parallel charge to 10 amps (which was showing 3.5v out @ 35w) but was increased to 14 amps (the max charger can do) 30 minutes ago.

The entire time I've had the pack connected in series (the whole thread until my last post), balance leads were connected to the cells and to the smart charger. It doesn't balance actively, it only begins to do it when a cell hits the target voltage. It will reduce current and bleed off as much power as it can from the high cell until all cells eventually make it to 3.6v. Then it dings and enters sustained balancing mode. I am yet to hit this mode with this battery setup.

I don't want to do the PSU dumb charging because that would almost certainly send one or two cells to 3.7 before the other two hit 3.5. I'll do that when charging up after I get the things top balanced and perform the first discharge.

Sorry for the miscommunication, but I assure you this has been smart charged with individual cell monitoring the whole time until I connected them in parallel.

 

[john61ct]

Aha OK good stuff.

So to reinforce now we're (in theory) at almost 800Ah capacity, so now the stop-charge endAmps spec is 40A, latest 20A.

So below that in other words, no way to tell when to stop.

Do the pause-rest-measure-resume routine in one-hour increments, seeing if the group resting OCV is climbing.

But be aware I'm much more cautious than most

Most industry sources would say since it's just one time, no harm in allowing current to trail down to **zero** acceptance.

So that is an alternative, at 3.55V still below the spec maximum, likely really is OK, if the above is too burdensome.

 

[hour]

Thanks, not sure how I'd do that with a max charge of 14 amps - BUT the cells, after a brief rest in parallel from their puny parallel charging, are back in series.

All cells have kissed 3.6v at this point, but not simultaneously. A strange thing happened - cells 2 and 3 were the ones that were rising in voltage more rapidly than 1 and 4 (seen in pictures of first post). They were the cells with double to triple the resistance readings on the charger (though I suppose we've determined that's not very meaningful).

So 20 minutes ago, cells 2 and 3 were at 3.59v and 1 and 4 were at 3.52. I just checked and they literally did a flip. 1 and 4 are at 3.6v and causing the charger to taper down current. Because there's less current, 2 and 3 are now sitting at like 3.4v. How irritating when they were all previously so close to one another and I thought I'd be getting close to the balance cycle.

I'm just going to let it run and take care of itself. It'll eventually conclude that charging is complete and enter the lasting balance routine. I probably should have left the thing running this morning, and I probably should have plugged in the active equalizer to speed things along. I'd do that now if it weren't dark, but that would get everything closer to happy - quicker - and then the charger could do its balance thing.

Between the first series charge cycle, the brief parallel cycle, and this current series cycle, the little charger has pumped in somewhere around 205 amp hours. While this isn't by any means conclusive of battery capacity, it does mean it's a hell of a lot bigger than 100ah.

 

[john61ct]

Bleeding balancing pauses charging, so they do settle during that time.

 

[hour]

Bleeding balancing pauses charging, so they do settle during that time.

Yeah quite a bit apparently. The cell sitting at 3.6 that was previously one of the low cells, was sitting at 3.56v several minutes after I unplugged the charger. So I'm guessing it's full to the brim, and I would have expected the other two to be full too from sitting at 3.6v for countless hours, but they're at 3.4v now with 1.8a charge current (charger's choice).

I yanked the old system out of the truck in the dark a few minutes ago and scavenged the active balancer. Already doing good work, so i'll leave that connected and the charger is back running trying to hit 3.6Vpc. Hope it figures itself out by morning

 

[hour]

Charge up and balance finished, all four cells holding 3.46v and 0.002v match. ?

 

[luthj]

Those cells were way out of balance. Either that, or they have way different capacities. That explains the different resistances reported. I will be very curious what a capacity test shows.

 

[hour]

Those cells were way out of balance. Either that, or they have way different capacities. That explains the different resistances reported. I will be very curious what a capacity test shows.

So no significant throttling of current happened on the original two cells that were pegging 3.59-6v (the set charge voltage per cell on hobby charger), but did when the other two cells which had previously lagged behind jumped to 3.6v, and at that point current was greatly reduced (down to like 1.8a) and the two previously highest cells plummeted to 3.4some volts.

I hooked up the active equalizer at that point, but the charger was still running and hell bent on getting everything to 3.6v. So while power was being distributed by the equalizer, it only took 3 amp hours charge input from that point forward to get every cell back at 3.6v. Something tells me that isn't a wild capacity difference. But also interesting is that after the first night of charging them in series, page 1 of thread with all the wildly different resistance readings, subsequent charging attempts showed way less difference.

I won't be able to deliver a proper capacity test, but I will attempt to pull a ton of power out of it at a consistent (though low, and very time consuming) rate and also use the BMV for what it's worth, to give me a rough idea combined with the time I was able to pull X watts. I'll call it quits when the first cell hits 3v, since I doubt they'll be anywhere near balanced towards the bottom. No capacity per cell, but just wanna know what I can pull out of it before things get squirrely. Odds are with a 29ah peak daily draw in the middle of summer with fridge running often - I'll never see anything below 70% whoring my setup out for friends and charging everything they/I can think of. I've been wondering if top balancing was the right thing to do, and I suppose it was given my planned use

 

[john61ct]

Charge up and balance finished, all four cells holding 3.46v and 0.002v match.

Yay!

So ready to use as a 4S bank.

For me next is some "commissioning exercise cycling" which will increase capacity a bit (assuming they are actually uncycled new cells)

One early goal is to decide how you want to keep it in top-balance, test between BMS and hobby charger for now.

At the same time, can now do some at least crude capacity testing.

 

[john61ct]

So no significant throttling of current happened on the original two cells that were pegging 3.59-6v (the set charge voltage per cell on hobby charger), but did when the other two cells which had previously lagged behind jumped to 3.6v,

Yes, even one cell at lower V will accept (demand, draw) a high current.

Now that you are in the breaking in stage, say first few dozen full cycles, try to keep charging to 0.2C and discharge to 0.4C, or lower.

This will help total cycling lifetime longevity, as well as the early-stage increase in capacity.

> and at that point current was greatly reduced (down to like 1.8a) and the two previously highest cells plummeted to 3.4some volts.

Sounds like the bleed-resistance balancing, pauses charging while burning Ah from the high V cells.

> I hooked up the active equalizer at that point, but the charger was still running and hell bent on getting everything to 3.6v.
> So while power was being distributed by the equalizer, it only took 3 amp hours charge input from that point forward to get every cell back at 3.6v

That's because the "dedicated balancer" in effect overrides the other balancing efforts (charger and/or BMS)

What is its balancing current spec?

> Something tells me that isn't a wild capacity difference.

Only CC load testing will reveal that with any accuracy. An Ah totalizer / SoC battery monitor will give a reasonable approximation.

Really at this stage, you want to do this capacity testing on each cell in isolation.

Your hobby charger may be able to automate multiple cycle averaging, maybe even output data that you can use to generate very informative time-series charts.

> I'll call it quits when the first cell hits 3v, since I doubt they'll be anywhere near balanced towards the bottom.

Doing this at the 4S pack level will be **very** approximate, so might as well see if the active balancer helps keep balance all the way down,

and if keeping it on fulltime like that has a net positive or negative impact on pack capacity.

> I've been wondering if top balancing was the right thing to do, and I suppose it was given my planned use

Yes, going down to very high DoD wipes out huge chunks of cycling lifespan.

 

[hour]

Thanks John, will stay within those values. BMV is hooked up and configured, calibrated, and finally managed to reset the history from the old battery (had to do it on the actual device)

If I had a means of measuring the actual balance current with a meter on the high cell lead, I'd tell you, but it's advertised as 6a / 10a (assuming 10 is peak). I'm sure that varies on how out of balance the cells are in the first place, so not guaranteed to be doing 6 amps ever.

The iCharger unit that i'm still waiting on may give me some additional options that you mention, but I don't see anything like that in the ISDT Q6 charger I've been using.

Right now I have 114w of load on the system, including the inverter. My inverter is 300w but I don't have much to connect to it that isn't way above (or below) 300w.. so right now two Vornado fans are giving me the ~100w of draw and the rest is coming from the inverter losses.

I'm timing this but not really to calculate anything with those numbers. I'm mainly just seeing how much I can take out (according to the BMV) before I call it quits. Hell, I might just do this for a couple hours and then charge up again - to see if they charge in harmony better than they did yesterday+day before.

Since I don't have a BMS on the battery yet (and I disconnected the equalizer, too much crap on terminals already until I rig up something better), I'm relying on a hobby cell monitor that live updates... and... a webcam! This updates every second and if cell 2 is no longer the lowest it'll switch to displaying the appropriate cell in this mode. Just for...safety, and curiosity, and to not have to walk over to the battery and check it manually every few minutes.

 

[john61ct]

If I had a means of measuring the actual balance current with a meter on the high cell lead, I'd tell you, but it's advertised as 6a / 10a (assuming 10 is peak).

With some the rate is adjustable, and works regardless of the delta between cells. Others, the actual rate goes down to nothing as the Vs approach balanced.

Please use Amps and Ah for capacity testing, even though voltage may vary a little it is a much better unit for such measures.

So 10A total loads for now we will assume is around 0.05C, aka the "20-hour" rate, happens to be the industry standard for measuring lead bank capacity.

If you can hold that pretty steady, then say it took 21hrs to go from 100% Full to 11.99V

that shows a 210Ah actual capacity, going over at 105% is normal for high-quality brands.

If the LVC cuts at 17hrs, then 170Ah means you got scammed by over 10%, which you've stated is OK.

But remember, actual Ah capacity may increase during the low-rate breaking in period, just as a result of doing full cycling.

Getting an accurate peak Benchmark is what sets your State of Health baseline FFR, see how fast SoH% declines as you rack up the cycles.

Vendors state 2-3000 but that usually assumes harsh high C-rate usage. My guess is, if properly coddled, 5-10,000 is feasible starting with a good product, leave the bank to your grandkids


> webcam

There are apps can use a smartphone to take a timestamped photo every X minutes

 

[hour]

Of course I screwed up my timing because 1) I decided to increase draw a few times as I found abandoned fans in my house, and 2) Had some downtime as I lopped off the ciggy plug on my inverter in favor of a more robust connection (XT60) so I could feel comfortable loading the 300w inverter down more.

We're coming up on 10 hours of draw with maybe 30 minutes downtime, and going from 114w to 141w to 206w to 287w to 262w.

BMV shows 150ah consumed as of typing this. Lowest cell is 2.99Xv @ 262w load, and has been dancing around there for hours (I know not an indication of capacity, but I wanted to know how much I could pull before the lowest cell gets weird at the bottom).

I'm not going to take it to the weird point since I've concluded the battery as a whole is capable of giving me no less than 150ah from full charge and everything is in a safe range.

Min max voltage difference between cells right now is 0.150v (total voltage on BMV @ 12.15v) under 262w of load. This is without the active equalizer connected.

150ah is 4 days of running my large fridge @ 36*F per compartment in 100*F+ ambient, and some other light DC stuff (LTE router, charging phones and speaker). I've never struggled to keep up with demand with 200w of solar on my truck's roof and the 57ah usable existing battery in the truck bed, so I'm good. Also worth noting that I'll never be pulling 262w continuous from the pack at this low of a state of charge like I am now. The only thing that would ever draw power right before sunrise when pack would be at its lowest is 70w from the fridge so the cell imbalance would presumably be less, and overall pack voltage higher. In fact, i'm going to switch most of my load off now and see what she says at 70w:

74w draw after consuming 150ah - pack has rebounded to 12.72V, cell imbalance high-low is 0.050v. ~3.160v is now the lowest cell, and the lowest cell is now switching between Cell 1, Cell 2, and Cell 4 (cell 2 was the lowest during the entire test until reducing to 74w)

This should work great for my purposes. I wouldn't be surprised if the cells are in the neighborhood of the 190ah they were advertised at, and I'll push it a little further on the next discharge test with active balancer connected.

Now to see if she'll charge up without another crazy long bleed/balance/current reduction routine

 

[john61ct]

Yes it may take dozens of at least "pretty full" cycles to get to peak capacity, consider them warmups, and giving you a chance to get familiar with gear and a baseline for just as important tuning in your intuition with the pack.

Looking good so far!

Keep the discharge rate below 500W / 40A, do **not** allow deeper than 11.95V, and see if you can get better precision maintaining a steady CC load rate.

Look forward to learning how the top balance maintains. . .