diy solar

diy solar

Do my new blue cells need to be packed together ?

The interesting thing about their charts, if I'm reading it right, is that they lose capacity faster with the fixtures and the high cycles exists only in their extrapolation.

It suggests that taking a uncompressed cell and compressing it will decrease capacity, and the opposite will increase capacity, but will increase wear as well.

Seemingly like how higher temperature increases capacity but increases wear.

It looks like the lines cross about 1500 cycles.

Based on this strapping cells is absolutely the way to go.
 
Last edited:
The interesting thing about their charts, if I'm reading it right, is that they lose capacity faster with the fixtures and the high cycles exists only in their extrapolation.

It suggests that taking a uncompressed cell and compressing it will decrease capacity, and the opposite will increase capacity, but will increase wear as well.

Seemingly like how higher temperature increases capacity but increases wear.

It looks like the lines cross about 1500 cycles.

Based on this strapping cells is absolutely the way to go.
Sorry, I am rephrasing to make sure I understand.

A compressed cell cannot hold as much power as an uncompressed cell.
A compressed cell will deliver it's capacity for more cycles.

Doesn't this beg the question as to what causes what? Maybe the compression makes the voltage rise earlier in the SOC, thus stopping the charge at a lower SOC. Now that we are using a lower SOC, we get more cycles. Which is to say that if they tested the 90% SOC cycle count uncompressed, they would get the same cycle count as the compressed. Of course none of this gives us any feel for the low C currents that any solar system will experience. Maybe all this is irrelevant, or rather completely unknown, if we never use more than .2C of current.

I am curious how they kept the pack cool while clamped and cycling up and down at 1C. This is asking whether you run the risk of failing to get the heat out if you clamp 8 in a row.

(FWIW, I find this interesting, but there are so little facts that I find no basis for doing extra work. I find it a bit scary to clamp the crap out of some cells in the hopes of getting longer life, when I know I will never know whether that worked or not. If I clamp, and the battery's longevity disappoints me, I will blame it on the clamping. If the longevity impresses me (10 years from now), I will blame it on the clamping.

I am going to snug my batteries so they don't rattle, and will never charge/drain them faster than .2C and mount them in my water bay where they should be cooler in the summer than the interior of the RV and not below freezing in the winter.)
 
If the pressure is caused by gas created at full charge, then heat would increase the pressure, but a slow charge rate would not avoid the problem.

And higher pressure inside the cells would increase the amount of energy required to charge. Hence why the unclamped cells look to have higher initial capacity with their charging profile.

Their charts show battery degredation at twice the rate unclamped after the initial thousand charges when the clamped batteries have more capacity remaining. That is significant, especially so if you intend to use the cells for many decades and past the 80% mark.

Think 10 year vs 20 year life based on their provided numbers
 
but a slow charge rate would not avoid the problem.
Except that if you simply clamp the cells in the obvious dimension, you are dramatically reducing the surface area necessary to get rid of the heat. At higher charge rates, you have to get rid of heat faster. In their tests, I believe they monitored the temperature, and they don't say whether they had to do something interesting to dissipate the heat. In our world, we might slap 8 together in a pack and mindlessly assume this clamping is good without regard to heat. At low C the heat might be irrelevant in relation to the surface area of the battery.
Their charts show battery degredation at twice the rate unclamped after the initial thousand charges when the clamped batteries have more capacity remaining. That is significant, especially so if you intend to use the cells for many decades and past the 80% mark.

Think 10 year vs 20 year life based on their provided numbers
I'm not sure it does show that. I think it speculates that. The red is an extrapolation, right? If we know how the cells/chemistry degrades, then we know how to extrapolate. If we are trying to determine how they degrade, then it is a big assumption to take the first 1000 cycles, fit your favorite simple math curve, and extend it. (I assume you are referring to the two graphs where they have the red dotted lines after the black)

When I look at fig. 1, I conclude that I want my BMS to monitor the resistance and stop as soon as it starts to go up dramatically (at the far left of the chart where the volts rise dramatically). I figure it should apply 3.4V and watch it gulping down current. When the current draw starts to slow, then stop charging. Similarly when the volts hit 3.2, stop loads.

If I do that, I will never get 280ah from the battery, but what will the fixture and without fixture charts look like?

I am not an expert on this stuff. I am just thinking out loud, so don't take this as information.
 
I wish we could edit our previous posting. My previous was rather poor.

My first paragraph really had nothing to do with the low C issue that was quoted by me.
I'm not sure it does show that. I think it speculates that.
I mean that at 750 cycles the no fixture version seems to have 96% capacity and the fixture version seems to have 94%, and the "red line" theory is that because of the way that the clamped arrived at this point, it will therefore slow down its degradation relative to the unclamped version. That looks to me like an extrapolation that might not be justified.
 
It's very interesting, the whole idea that managing the expansion in some way could meaningful improve life span.

OTOH, we have a lot of other makers of modern cells that are explicitly saying they should not be compressed, and in some cases may not even be in any contact with each other all. These cells have lab (and extrapolated) storage lifecycle curves that look similar -- not half as many. I suppose it could be possible that the effect is both real and substantial, AND that some other vendor cells have more rigidity or a different internal geometry that does the "compressing" for us.

I'm inclined to wait for more data before getting too excited about the finding, one way or the other.
 
Today i tested and fully charged my batteries, with 28.2V at 50A (They are 280A so little lower than 2C).

They indeed inflate a little bit. They used to rest flat between them, and now some of them a little more chubby so they not longer sit flat:

I stack them together vertically to better apreciate this, on the right (uncharged and flat), on the left, charged and fatty:

1588721822914.png1588721834612.png


Not sure if this is normal. I charged with a BMS and they never passed 3.5V, so i dont think i overcharge them. If this change in volume is expected, i think it could be good idea to clamp the cells, as this continuous change in volume its probably not good for the materials. Im not overly concerned about the heat, since after some testing found out that with my loads, they rarely go above ambient temperature.

So in the clamping vs no clamping debate, maybe you should consider if you are more worried about the temperature, or the possible delamination for the continuos expansion / contraption?
 
280ah charged at 50a is .17c

1C would be 280ah, 2C would be 560a.

If they swelled at such a low C charge the aluminum casing would seem weak and maybe binding them is needed. Just guessing though, no swelling with my plastic prismatics but they are bound together.
 
How long did the charger hold them at 3.5Vpc? What was the current into the cells at the end?

It's totally possible to overcharge LFP at 3.5Vpc, or even a little lower. It just takes a while. (Not saying you did, just stating what's possible.) Without charge current and time data, saying "I charged to N volts" doesn't really tell us much.
 
How long did the charger hold them at 3.5Vpc? What was the current into the cells at the end?

It's totally possible to overcharge LFP at 3.5Vpc, or even a little lower. It just takes a while. (Not saying you did, just stating what's possible.) Without charge current and time data, saying "I charged to N volts" doesn't really tell us much.

They were new from the factory and should been between 40 - 60% state of charge originally (I think it was around 3.2v/cell).

The LV2424 put 50A for 2 - 3 hours, and then while the inverter screen was about to hit top voltage (28.2V), the cells started to creep to values above >3.4V and increased a little bit for half and hour or less... i dont think i ever saw 3.5V but i might not remember right. But im pretty sure it wasnt that long time... 45m tops.

Now, the weird thing is... i charged the second bank, with similar times and they are perfectly flat!. So i might overcharged the first bank... but not sure how!.... The inverter configs were the same, and the algorithm seems to be pretty decent at not overcharging the batteries, there is even some people in the forum complaining about not being able to fully charge the batteries due to the inherent charging method of the MPP Solar inverters.

EVE LF280-72174 Version E has more cycle life "with fixture" per the EVE spec sheet.

See this link:
https://diysolarforum.com/threads/spacer-or-no-spacer.6826/post-72973

Yep!, we were discussing that a couple of pages before in this thread about how necessary its to meet that spec with our fraccional C rates applications... and the compromise that you have by clamping them by reducing the heat dissipation. So its a matter of what can affect the most the batteries, the no clamping or the heat.

In my case afther this im decided to clamp them but i might put some aluminum bars betwen them to have some airflow.


-------------------------------

Will that swelling affect my batteries lifespan?... now im worried about them.
 
The LV2424 put 50A for 2 - 3 hours, and then while the inverter screen was about to hit top voltage (28.2V), the cells started to creep to values above >3.4V and increased a little bit for half and hour or less... i dont think i ever saw 3.5V but i might not remember right. But im pretty sure it wasnt that long time... 45m tops.

Without knowing the approximate current, it's hard to estimate what the true SOC is. 3.40Vpc with prior current of 0.15C is around 75% SOC. But at 0.05C, it would be close to 95% SOC.
 
EVE LF280-72174 Version E has more cycle life "with fixture" per the EVE spec sheet.

See this link:
https://diysolarforum.com/threads/spacer-or-no-spacer.6826/post-72973
Correct me if I am wrong, but that document shows a projection that, if correct, shows more cycle life. It looks to me like they fitted some curve to the first 750 cycles, then extrapolated that into the future. I'm not saying that prediction is wrong, but I don't see why we would conclude it is correct either. Their tests showed that at ~750 cycles, the non-fixture version had ~93% capacity and the fixture version had ~92%. They did not test more cycles.

If someone with knowledge about the chemistry, and build, told me that they expected longer life from clamped cells, and that the first 750 cycles matched their expectations, then I'd be more willing to draw a conclusion from that red dotted line.
 
They were new from the factory and should been between 40 - 60% state of charge originally (I think it was around 3.2v/cell).

The LV2424 put 50A for 2 - 3 hours, and then while the inverter screen was about to hit top voltage (28.2V), the cells started to creep to values above >3.4V and increased a little bit for half and hour or less... i dont think i ever saw 3.5V but i might not remember right. But im pretty sure it wasnt that long time... 45m tops.

Now, the weird thing is... i charged the second bank, with similar times and they are perfectly flat!. So i might overcharged the first bank... but not sure how!.... The inverter configs were the same, and the algorithm seems to be pretty decent at not overcharging the batteries, there is even some people in the forum complaining about not being able to fully charge the batteries due to the inherent charging method of the MPP Solar inverters.

Yep!, we were discussing that a couple of pages before in this thread about how necessary its to meet that spec with our fraccional C rates applications... and the compromise that you have by clamping them by reducing the heat dissipation. So its a matter of what can affect the most the batteries, the no clamping or the heat.

In my case afther this im decided to clamp them but i might put some aluminum bars betwen them to have some airflow.

-------------------------------

Will that swelling affect my batteries lifespan?... now im worried about them.
NO WORRIES: The specs clearly indicate that a certain amount of swelling occurs and is normal... I believe off the top of my head they allow for 2mm of expansion. I have experienced this myself with the EVE-280LF's from Xuba. It is within spec.

BTW: I have BOUND my cells into bundles of 4 cells (20kg / 44lbs is easier to furtle with as a block versus 8 cells) but I did this i a different way. I used block clamps to "snug" the cells nice & square and then I used Fibre-Mesh filament tape to hold them together tight, I added "pull strings" so that if I need to pull a "block" from the battery box, I can do so without having to flip the case upside down to drop them out.

cell-bundling.jpg
 
Correct me if I am wrong, but that document shows a projection that, if correct, shows more cycle life. It looks to me like they fitted some curve to the first 750 cycles, then extrapolated that into the future. I'm not saying that prediction is wrong, but I don't see why we would conclude it is correct either. Their tests showed that at ~750 cycles, the non-fixture version had ~93% capacity and the fixture version had ~92%. They did not test more cycles.

If someone with knowledge about the chemistry, and build, told me that they expected longer life from clamped cells, and that the first 750 cycles matched their expectations, then I'd be more willing to draw a conclusion from that red dotted line.

@johntaves
Agree that we all need to read/analysis documents during our build projects. Quality/completeness/update history of documentation helps us to be able to analyze graphs like this. Agree that the "known & tested" cycle life shown is for ~750 cycles. Then it is extrapolated on both graphs. I don't like the extrapolation either, but it is better than a one page spec sheet many of the vendors on Aliexpress and/or Alibata try to use that masks the source of the cell.
Additionally, section 6.1 Electrical performance provides the exact method of testing to determine/project the capacity and cycle life.

My assumption, considering the document completeness, update history, and capacity testing information plus manufacturer reputation is that these projections are relevant and need consideration.

To me -it is clear that EVE expects these cells "with fixture" will cycle longer in a low C situation. If your situation has higher C levels of discharge and charge then "without fixture" might be a suitable.
It is also clear to me that EVE expects the cells to expand and contract from ~30% SOC to 100% SOC. This plus other information in the document about how heat affects capacity retention and overall cycle life leads me to a conclusion that in low C situations "with fixture" is right for my current RV/Fifth wheel project.
 
Quality/completeness/update history of documentation
The graphs leave me with no confidence at all. They have totally different scales, and the red extrapolations are different too. The fixture version seems like they might be real data points. The other does not. I cannot comprehend how an engineering team could produce those two, unless they were created at totally different times by different people and pasted in there. At least that part of the document is not quality.

I too am in an RV. I will snug mine, so I am confident they won't be moving too much. I won't attempt to replicate the pressure they used.

Now, the weird thing is... i charged the second bank, with similar times and they are perfectly flat!. So i might overcharged the first bank... but not sure how!.... The inverter configs were the same, and the algorithm seems to be pretty decent at not overcharging the batteries, there is even some people in the forum complaining about not being able to fully charge the batteries due to the inherent charging method of the MPP Solar inverters.
It would be nice to hear from mrdavvv if this can be reconciled. It would be nice to see similarly charged batteries with similar swelling and get a feel for whether that swelling occurs disproportionally at high or low SOC or whether it is linear.
 
NO WORRIES: The specs clearly indicate that a certain amount of swelling occurs and is normal... I believe off the top of my head they allow for 2mm of expansion. I have experienced this myself with the EVE-280LF's from Xuba. It is within spec.

BTW: I have BOUND my cells into bundles of 4 cells (20kg / 44lbs is easier to furtle with as a block versus 8 cells) but I did this i a different way. I used block clamps to "snug" the cells nice & square and then I used Fibre-Mesh filament tape to hold them together tight, I added "pull strings" so that if I need to pull a "block" from the battery box, I can do so without having to flip the case upside down to drop them out.

View attachment 12710


Hello Steve!.

Thanks for the info, that relaxed me a bit,.... however in your experience the cells ended up at the original size after discharge?. I expected that and let them go for around 30 - 40% SOC, but one of the its still a litle bit swolled, i think im inside the 2mm spec but a little worring still.

The tape looks great!, you think it would be able to hold aging and pressure with multiple cycles?.
 
Hello Steve!.

Thanks for the info, that relaxed me a bit,.... however in your experience the cells ended up at the original size after discharge?. I expected that and let them go for around 30 - 40% SOC, but one of the its still a litle bit swolled, i think im inside the 2mm spec but a little worring still.

The tape looks great!, you think it would be able to hold aging and pressure with multiple cycles?.
I did have cells swell a bit when they were fully charged and they did appear to contract a bit. This happened when I had them in the "free" while testing. I'm not sure about the tape but this is for commercial packing and the glue is nuts, will it last 10 years ? I dunno. Because of how I am "boxing" them, I doubt there will e any issues with expansion / contraction. I am contemplating a thread on the pack assembly, batt box build and such with the XUBA EVE-LF-280's, here is a couple of photo's of how I built my battery boxes for the 280's.

280AH-Battery_Box.jpg


Box Assembled (A).jpg
 
Back
Top