Be cautious ordering generic blue wrap LFP cells

[hour]

I got these stupid things yesterday and sat them side by side. Two sit flush together, two do not... so I guess there's some swelling. I'm sure not all blue wraps are the same, but they're the same in the sense that you probably have no idea ****** they actually are.

Question for guys that know: I was going to clamp these together tight. Is that still acceptable if there's slight curvature to some cells? Or will that squish their contents and wreck them? I need to get them tight together or I'll end up stressing the connections at the bus bars as things shift.

The ones I bought use studs in the terminals, so those studs won't have to be removed really ever - I just pop the nut off and can slide ring terminals over. Sadly one of the terminals had some additional metal at the top preventing the stud from threading in straight. I did some light filing and then just said screw it..and screwed it in gently. It found its way, but yeah - another inconvenience that they could have easily prevented with better QC.

All cells arrived at 3.06-3.07v which I was cool with. I just got the thing assembled and on the balance charger, and after a minute or two the charger read out their internal resistance. Doesn't seem too well matched. Seller says "The internal resistance is very close, we have tested it before shipment, about 0.28 Ohms". Yeah I don't know about that. Wonder what'll happen once they harmonize when the charging and balancing is done. Could take a very long time given the difference I'm seeing now.

In summary, you should probably buy trusted cells if you're building your own pack. If I can figure out how to get them mounted to ensure no tension is placed on the studs / bus bars (accommodating for warp, somehow) then they may be better than my ancient-cell DIY 90ah pack (that I use at 60ah, because one goes wildly out of balance on the high end). Since these blue cells are advertised as 190ah (capacity test to come..) even if I have to baby the pack because of cell variation, I should be able to get a whole hell of a lot more than 60ah.

Fingers crossed it wasn't a complete waste, but I could have done better.









after forcing them together... obviously not going to move the pack like this to avoid strain on things.


after being on the charger a few minutes... all started at 3.06 - 3.07v

 

[john61ct]

Get the bars with a flex bow in the middle, or just use crimped ring terminals on marine wire.

I'd use 1/4" aluminum end plates with threaded rods allow you to torque the nuts down to a spec.

Don't think trying to squeeze them back to flat would be a good idea.

Make sure to capacity-test each separately, more important than resistance matching.

What's the use case? max discharge C-rate?

Shoulda gone for the (apparently) Sinopoly from the same seller right?

Or even better, official channels, lots of Grade B/C factory QA rejects sold out the back door, probably counterfeits too.

If they turn out to have a way lower Ah than claimed, clear case for a refund dispute.

Did you use PayPal?

 

[hour]

Get the bars with a flex bow in the middle, or just use crimped ring terminals on marine wire.

I'd use 1/4" aluminum end plates with threaded rods allow you to torque the nuts down to a spec.

Don't think trying to squeeze them back to flat would be a good idea.

Make sure to capacity-test each separately, more important than resistance matching.

What's the use case? max discharge C-rate?

Shoulda gone for the (apparently) Sinopoly from the same seller right?

Or even better, official channels, lots of Grade B/C factory QA rejects sold out the back door, probably counterfeits too.

If they turn out to have a way lower Ah than claimed, clear case for a refund dispute.

Did you use PayPal?

Yeah, used paypal. Pretty sure I'd have to pay shipping back if I'm dissatisfied, no clue what that costs, but that also feels like shipping something in to the void. So I'm hoping for at least 150ah - and I'd like to think that's possible if these are anywhere near legit. Ideally they'll clock in at somewhere in the neighborhood of 190ah and I'll treat them like a 150ah pack for longevity.

Seller is sending the curved bus bars for free so I can stretch them to fit the cells. I'll clamp them but not attempt to torque them hard trying to flatten the cells. This will be going in an ammo can, so I'm going to use a single end plate inside the can and threaded rod on each side of the pack, then through the back of the ammo can (back wall of the ammo can serving as the second mounting plate.) With any luck this will accomplish securing the cells in the can, and to one-another. I'll run something down the middle, perhaps attached to the lid of the can, to prevent vertical movement.

I have no idea how to go about testing the capacity of each cell individually, but I understand that having one cell with reduced capacity, for instance, would result in it charging up quicker than the others. Wouldn't variation in internal resistance do the same? My plan was to charge up to 3.6v this first time and let things top balance, which again may take an entire day if two hit 3.6v while the other two are like 3.45v. Then I'll discharge the pack to a little below 3.0v per cell. If one cell begins to plummet I'll quit the initial discharge test there and note all the cell voltages when that occurred.

THEN i'll take them back up to 3.6v and hope that they reach that point in closer harmony, resulting in a much shorter balancing cycle. THEN discharge again to as low as I can before one cell falls off with the BMV hooked up to get an initial rough capacity report. THEN charge up to 3.4v, discharge monitoring capacity once again til the weakest cell begins to fall off. That should give me a rough idea of what I can expect out of the pack in normal operation, where I won't be charging the hell out of them.

I'll then set the BatteryProtect a little higher than whatever voltage I noticed one cell start to drift. Charge up to 3.4v, discharge until BatteryProtect kicks in, and program that final capacity number in to the BMV. With any luck this will be in the neighborhood of 150ah. I may even lower the max charge to 3.35 / 13.4v and do another capacity test. If it's adequate for my needs I'll call it good.

Use case? Running a fridge (~29ah/24 hours peak summer, 110*F ambient). That'll be the constant. It might power my diesel parking heater for winter camping, but that's pretty low draw once it gets rolling. No microwaves, induction cooktops, heating, or cooling. I have a 300w inverter that I never use, but I'll keep carrying it around for air mattresses of friends, or charging my e-bike/battery chainsaw.

Max solar it'll ever be fed is 190w. If I had to find an additional charge source, it'd probably be a 40a DC-DC charger. So in short, very light use, and light charging.

Edit/additional info FWIW: Been on charge for 2.5 hours exactly, and pumped in 445 watt hours thus far. The resistance reported by the hobby charger has gone up a lot. The two lower cells (see photo in first post) are practically the same, 5-6 mΩ. The two that were much higher initially are now 16.5mΩ and 18.9mΩ. I don't know what to make of that. I have two cells that are 3.51 - 3.53 and two that are 3.35. Can a balance even remedy this? Seems like it'll always plague me. I'm charging at 182w right now which is as high as this particular hobby charger will do, and that's OK since it's nearly identical to my solar charging abilities which will almost always be the sole source of power input for the unit.

 

[deeznutz206]

Sorry for all your troubles. Thanks for the updates so far. Waiting to see what their capacities are.

 

[john61ct]

I have no idea how to go about testing the capacity of each cell individually

Charge to mfg spec Full, e.g. 3.6V

Do not rely on BMV for accuracy in load testing.

Use resistance or light bulbs if you don't want to buy a dummy load tester

Rate can be anywhere from 0.05C up to say 0.3C or higher to match your use case.

Hold amps as precisely constant as possible.

Precisely time how long to get to 3Vpc. Don't go lower, and don't sit there, immediately recharge at a 0.3C rate or lower.

Say CC rate is 0.1C, so should take 10 hours,

For new cells, if the LVC cuts out anywhere below 9.5 hours you got ripped off and I'd start negotiating for a refund.


_____
but I understand that having one cell with reduced capacity, for instance, would result in it charging up quicker than the others. Wouldn't variation in internal resistance do the same?

Yes, but that's why you use a good balancing BMS, or at least monitor cell voltages and use a dedicated balancer when needed.

Initial balancing, in parallel and up to 100% and keep hooked up together for as you say a day or two.

3.45Vpc is the lowest I'd charge to. Make sure to charge faster than 0.10C rate, so you stop before trailing amps falls to 0.05C. Or just stop at CC to CV transition.

3.35V is lower than **resting** Full, need 1-2V above that for current to flow.

 

[hour]

Charge to mfg spec Full, e.g. 3.6V

Do not rely on BMV for accuracy in load testing.

Use resistance or light bulbs if you don't want to buy a dummy load tester

Rate can be anywhere from 0.05C up to say 0.3C or higher to match your use case.

Hold amps as precisely constant as possible.

Precisely time how long to get to 3Vpc. Don't go lower, and don't sit there, immediately recharge at a 0.3C rate or lower.

Say CC rate is 0.1C, so should take 10 hours,

For new cells, if the LVC cuts out anywhere below 9.5 hours you got ripped off and I'd start negotiating for a refund.


_____
but I understand that having one cell with reduced capacity, for instance, would result in it charging up quicker than the others. Wouldn't variation in internal resistance do the same?

Yes, but that's why you use a good balancing BMS, or at least monitor cell voltages and use a dedicated balancer when needed.

Initial balancing, in parallel and up to 100% and keep hooked up together for as you say a day or two.

3.45Vpc is the lowest I'd charge to. Make sure to charge faster than 0.10C rate, so you stop before trailing amps falls to 0.05C. Or just stop at CC to CV transition.

3.35V is lower than **resting** Full, need 1-2V above that for current to flow.

Hmm, I have a halogen 12v headlight that I wired for discharging individual cells on my last pack, but doing that on each cell would be a many-day endeavor, and having to watch to catch it at 3v... yikes. Not sure if I'm that motivated to figure out the near-exact capacity of each cell. I'll consider going down that route if I'm completely displeased with the results of the test as planned. If I can't rely on the BMV for capacity approximation then I'll have to figure something else out, but my hope is that I could at least determine if the pack is capable of giving 150ah. I've seen the timed discharge capacity tests on youtube using inverters, but they usually set it to 1000w and run the entire load through an inline meter, heat guns seem to be typical for conducting such tests. I have the meter, and a heat gun, but I don't think I'll buy a 1000+w inverter since it'll never be used again.

10-4 on 3.45v per cell. I can't achieve that with my hobby charger (3.6v lowest for charge, 3.4v lowest for storage charge, no free selection) so I'll have to rely on the SmartSolar to hit 13.8v. I do have a 6 amp active equalizer that several others have, but I found on my other pack that it still can't contain the runaway voltage rise of my crappiest cell when charging around ~10 amps. That's why I set that pack to charge up to about 13.6v.

Kind of surprised to hear you say 3.45 per cell minimum, isn't that 100% full? I thought you specifically were always advocating for charging less, and I wasn't opposed to charging to 80% for longevity.. especially if I have more capacity than I could use on an outing even with awful weather.

 

[john61ct]

Voltage on its own does not spec SoC.

How long CV is held, or C-rate if CC-only is just as critical.

If CC-only, at 0.2C stopping at 3.45V is maybe 92-94% compared to maker-spec max as 100%

I just don't think there's any longevity advantage to giving up more than that.

Avoiding high DoD is much greater a factor, so leaving Ah unused at the bottom end rather than the top has a much bigger payoff.

 

[hour]

Thanks for that clarification, do you still stand by 3.4v/13.6 for fluctuating and underpowered solar setups? Saw you post that in another thread recently.

Pack is still charging up at 184w and not behaving like I was expecting. My current house battery isn't a good reference point since the cells are 10 years old and it has...issues, but from charging up my 26650 packs it seems like after 3.5v is hit on a cell, it peaks shortly after. That always made sense to me, hearing that most of the energy will be had below that (3.2-3.4v). Right now I'm at Cell 1: 3.40v, Cell 2: 3.52, Cell 3: 3.55, Cell 4: 3.41. It's been like this for 2+? hours now. I was honestly expecting those 3.5X cells to zip up to 3.6v and then cause the longest balance ever.

5h22m in: 69.2ah, 970wh. Really wish I had my existing pack inside so I could scavenge the active equalizer off of it and watch that process happen with severely imbalanced cells. Also really wish I had bought a bigger hobby charger and power supply last week. Guess I'll buy that now since I have a lot more testing to do on this thing.

 

[john61ct]

Thanks for that clarification, do you still stand by 3.4v/13.6 for fluctuating and underpowered solar setups?

Only if the current is under 0.05C or so for long periods of time.

This is dangerous when the bank is near Full since you can't use trailing amps to determine the stop-charge, easy to overcharge and lose cycle lifetime.

Isolating and measuring OCV at rest is a workaround but not convenient.

Much better to ensure input stays at 0.10C or higher.

 

[john61ct]

The difficulty in getting to 3.6V may well be an indicator of poor quality, impurities in the chemistry.

Can pick any spec, say 3.55V @ .05C as your 100% Full benchmark definition

Since you don't care about a super precise capacity test.

Less wearing, very little actual usable capacity after 3.45V, lots of the charger output is just dissipating against the increasing resistance.

No harm in stopping "early", put in parallel to self-balance, go to higher SoC at 1S voltage, "Top-balance SoC" should be a little higher than your usual "usage cycling Full".

From the latter POV, resting at 3.34-3.35Vpc is as Full as you ever need to go.

 

[hour]

The difficulty in getting to 3.6V may well be an indicator of poor quality, impurities in the chemistry.

Thanks, I was wondering if my charge rate was too low.

Summary of what happened since I can't quite wrap my head around it and maybe your quote above tells the tale...

Charging went on at a full ~184w (13a) for 9 hours before I went to bed last night. Cell voltages were Cell 1: 3.41, Cell 2: 3.55, Cell 3: 3.59/3.60, Cell 4: 3.42 when I retired.

Woke up this morning and charging had been reduced to 10a. I was still surprised to see it that high, since one cell had been hitting 3.6 briefly and that usually makes charging slow down to the single digits.

Voltages were: Cell 1: 3.43, Cell 2: 3.58, Cell 3: 3.6, Cell 4: 3.43 and 169ah had been delivered total in the entire 13-14 hours of charging.

Not knowing what to make of this, I stopped the charge process when I headed to work an hour ago. Within like 10 seconds all the cells were resting at 3.34v - all of them, identical.

I'm going to connect the active equalizer this afternoon and see if I can get them to charge up together to at least 3.5v. I do wonder where all that energy was going. Nothing was really hot, bus bars just barely above ambient temp.

 

[luthj]

Hmmm, I wonder how many wraps of the cell "sandwhich" those have in them? The internal resistance measurements from the controller are just rough estimates.

The easiest way to calculate internal resistance is by the change in voltage between two known currents. Start with the batteries fully charged. Apply a modest current (say C/10) and measure the terminal voltage. Then switch to a higher current. Battery resistance = delta V / delta I. Apply the measured values to the equation, and you can determine the effective internal resistance for each cell. Note that this will change as the battery discharges.

If the cells are more than 20% apart, you will see significant cell disparities at charge/discharge rates above C/10. The higher the currents, the bigger the changes.

You might get lucky with paypal if you show the cells visibly bulging.

A poster over in the Mercedes truck area ordered 16x100AH cells directly from the Frey factory. They came with test sheets from the MFG, and they were all made on the same day/batch.

 

[john61ct]

Thanks, I was wondering if my charge rate was too low.

What is your C rate?

Is this just bare cell, or is a BMS in the mix?

If bare, you really should be stopping soon as the first cell hits 3.6V

Best to bring the others up separately, or drop the high one down then continue in bulk mode.

> Within like 10 seconds all the cells were resting at 3.34v - all of them, identical.

So you're letting them sit, wired in parallel right?

> I do wonder where all that energy was going.

Hurting longevity

Internal temps just rise slightly unless you're really pouring in high C-rate

 

[hour]

What is your C rate?

Is this just bare cell, or is a BMS in the mix?

If bare, you really should be stopping soon as the first cell hits 3.6V

Best to bring the others up separately, or drop the high one down then continue in bulk mode.

> Within like 10 seconds all the cells were resting at 3.34v - all of them, identical.

So you're letting them sit, wired in parallel right?

> I do wonder where all that energy was going.

Hurting longevity

Internal temps just rise slightly unless you're really pouring in high C-rate

The cells started out at 3.06 - 3.07 out of the shipping box. I forget what they were with a higher precision multimeter, but they were all pretty much equal. So I put them in series and began charging them up as a 12v pack.

No BMS in the mix currently, this was being done with a hobby charger that drops current and begins to siphon power (waste to heat) when one cell reaches the set voltage (which I set to 3.6v).

I'm spacing how to calculate what the C rate was of my charging, but 13 amps @ 14.X volts was coming out of the charger, or 184w. In to a 190ah 12v 3.2x4S battery. So pretty low, but the highest I can charge with my current hobby charger. If this will be insufficient for the size of battery then I'm in hot water, since I'm maxed at 200w of solar on the roof of my truck... and what I was charging at via the hobby charger is about as high as I'd ever see with my solar setup on the best of days.

 

[luthj]

Nothing wrong with charging slowly, it just takes longer.