LiFeYPO4 cells from GWL Power, new or used?

[fox87]

Hello folks,

I recently bought 4 300Ah cells from GWL Power. They arrived in good order but the terminals are pretty scratched and dirty for what are supposed to be brand new cells. Or at least that is the way that it seems to me. As I have no experience with these things I figured I would ask here. It it normal for the cells to look like this or are they possible preowned/used cells?

Probably a bit hard to tell but I'm curious what people with some more experience think about it.

The cell voltages are:

3.298
3.298
3.299
3.288

I attached some pictures of the terminals. Looking forward to you input. If they are selling used cells people should be aware...

Thanks for any input!

[BoloMKXXVIII]

Is there any information on the cells (in English) that you can use to search online to try to identify when the cells were made? The scratches all seem to be very similar. Not sure if that is sloppy manufacturing, done in testing, or just used cells being sold as new.

[fox87]

Hi Bolo,

Thanks for your reply. I got these serial numbers:

TSWB-LYP300AHA 181207-Y08871
TSWB-LYP300AHA 181207-Y08269
TSWB-LYP300AHA 181207-Y08270
TSWB-LYP300AHA 181207-Y08271

I sent an email to the manufacturer hopefully the can supply the production dates. Not sure how else I could find out.

[fox87]

Although I suppose 181207 could actually be the date (07 december 2018) in which case the cells are around 5 months old.
Or 18th of december 2007 in which case they would be very old..

[Ampster]

I would just clean off the terminals with a non metallic abrasive cloth and not worry about it. I bought some Winston cells that looked like that when they arrived but they worked fine. The copper does get oxidized easily and the cells do get tested at the factory and maybe at the supplier so that might explain the scratches. Parallel them for a day or two to get their voltages equalized before assembling them in your pack.

[PNjunction]

You'll have to test them. First clean and then get some "noalaox" or "penetrox" or similar. A *light* coating of this should go betwen the hardware and your nickel straps.

PLEASE don't just bolt this stuff together with hardware-store stuff, which is very high-resistance. When you get stuff like this, be sure to get the matching straps (usually nickel, which also need a light cleaning) and bolts.

And voltage means nothing - due to the very flat charge-discharge curve - especially flat with lifepo4 - a resting voltage on one cell may indicate about 80% charge, and another very close could be 20% charged. Blindly paralelling them can cause a LOT of current to flow into the weakling - enough to startle you and cause a knee/wrist-jerk injury. So just beware.

Do yourself a favor - get a large "single cell" lifepo4 (3.2v nominal!) charger and charge each before bolting together. Later you can decide if you want to top or bottom balance them. But first, see how each cell reacts indivudually to a normal charge, otherwise you can be chasing demons in the 4S bank without knowing.

We're talking about 15A / 3.2v charger, which isn't cheap. Or a programmable high-power supply. Wall-warts wont' charge properly. Or of course the bms / bleeder thing --- all covered in the forums here...

[fox87]

Thanks for the replies everyone!

I got the matching straps and noalox stuff for the final assembly. I also have a 3.65V/20A single cell charger. So I guess I will charge them individually first before connecting them in parallel to allow them to level out. Then after a couple of days connect them in series and hook up the BMS.

[inetdog]

Thanks for the replies everyone!

I got the matching straps and noalox stuff for the final assembly. I also have a 3.65V/20A single cell charger. So I guess I will charge them individually first before connecting them in parallel to allow them to level out. Then after a couple of days connect them in series and hook up the BMS.

The advice given was to charge them individually and then immediately put them in series. Not to let them sit paralleled for any length of time first. The suggestion was to immediately put them in series and start working them, with the plan to complete either top or bottom balancing later. Charging them and immediately putting them in series is one way of getting a top balanced configuration.

[Ampster]

The advice given was to charge them individually and then immediately put them in series. Not to let them sit paralleled for any length of time first. The suggestion was to immediately put them in series and start working them, with the plan to complete either top or bottom balancing later. Charging them and immediately putting them in series is one way of getting a top balanced configuration.

Opinions vary as to whether paralelling them is useful. The distributor actually recomends chaging them in parallel. The OP stated that they were at approximately 3.3 volts and within .011 volts of each other. That is close to the resting voltage of a fully charged Winston LFP cell. They probably wont take much more charge. Parellelling them is simpler than charging them individually. After they are in series you can observe their behavior as you discharge and charge the pack. If your BMS reports cell voltages you can log the delta or just measure with a good VOM. littleharbor is correct that cell voltage will not be a good measure of SOC, but cell voltage deltas can give you a clue if one cell ihas less capacity than the others. To prolong the life of the cells during regular use I dont charge them past 3.5 volts, since there is less than 5% capacity beyond that. Opinions also vary as to the benefits of bottom balancing.

[fox87]

And voltage means nothing - due to the very flat charge-discharge curve - especially flat with lifepo4 - a resting voltage on one cell may indicate about 80% charge, and another very close could be 20% charged. Blindly paralelling them can cause a LOT of current to flow into the weakling - enough to startle you and cause a knee/wrist-jerk injury. So just beware.

Charging them in parallel was my initial plan. Would it be safe enough to just connect them in parallel for initial charge or can it cause a large current flow like PNjunction was saying?

[PNjunction]

You have a single-cell charger, so let's do it right for brand-new purchases which you've never seen charge before. ESPECIALLY if the cells are of unknown pedigree.

1) Use your single cell charger to charge each individually. Watch for problems. Will your single-cell charger dutifully charge 3 of them with no problem, but then seemingly take forever on the 4th one? That cell may have a problem. This is an easy way to spot a problem up front before you put all that time into it, and where an initial series connection may hide an issue from you.

Assuming the cells are healthy, they should be quite close to usable from a "top balanced" standpoint if you now wire them up in series.

If you feel the need to put them in parallel after this and let them sit like that for maybe a month to do magical self-balancing, be my guest.

What I'm saying is that the single-cell charger doing an initial charge individually, just saved you a LOT of time. Now you can go to town in series and do what you will.

Have you ever just banged a bunch of these together in parallel right after purchase, with varying states of initial charge, and hoped the would magically self-balance under charge? Uh, yeah they don't.

[PNjunction]

Just a p.s.

Voltage at rest for LFP is quite different from cell voltage *under charge*. Cells can be quite unbalanced, but rest seemingly close to each other, leading to false conclusions. You can get a general idea, but not good enough to claim that the cells are truly balanced.

What is most important is the voltage under charge conditions. THAT is where you don't want to exceed 0.1v difference max between cells at the end of charge.

[Ampster]

I am not disagreeing with @injunction but have a different perspective. A single cell charger simply means it cuts out at the voltage of a single cell. That is the same voltage as 4 cells in parallel but 1/4 the Amps. When in parallel, physics insures that the voltage of all the cells will be the same. Amperage into each cell could be different depending on the SOC of each cell.

There is no way to perfectly balance them as far as SOC is concerned because each will have slightly different actual capacity in terms of Watt hours. Either approach will get you close enough. I just prefer to let physics match the voltage than trust even the best VOM or charger. I agree they could settle at slightly different voltages after a few hours. What is important is that any cell deltas do not change significantly as they discharge in series. The weakest (the one with lowest capacity) cell will reach a lower voltage near the bottom. Stay away from the knee of the discharge curve because things go bad quickly as that curve gets steep very quickly. I recommend downloading a copy of the charge discharge curves to give you some insight into where you want to set your high and low voltage cutoffs. With Winstons I used 3.45 and 3.0
Here is the discharge curve:

https://66.media.tumblr.com/f43567c6b90989ba253abbf5a93ec2be/tumblr_ncof643gn51qa2swjo1_500.gif

[Ampster]

........
Have you ever just banged a bunch of these together in parallel right after purchase, with varying states of initial charge, and hoped the would magically self-balance under charge? Uh, yeah they don't.

Yes, I have done it a lot. I only have to let them settle for a day or two while I am fabricating buss bars or other tasks. It is probably not much different either way for 4 cells. However it saved me a lot of time in a 36 cell pack for an EV. I also did it for 48 cells for a stationary pack using Thundersky's which were later assembled into a 3P16S pack. I don't charge them in that configuration, I just want them to be at the same voltage when I put them in series. I am doing it now for 10 modules from a Nissan Leaf pack that I am putting in a GEM NEV. Either way the OP has some opinions that he can chose based on his situation.
NB Just to be clear, I am not under the misconception that any of the approaches discussed actually balance the cells as far as SOC is concerned. My goal in this process is simply to get them close in SOC but at the same voltage.