diy solar

diy solar

Our NEW GO-TO LiFePO4 vendor...09.20.21

I think this statement can be very misleading. I believe that cycle life - as described by cell manufacturers - is the number of full charge / recharge cycles until the usable capacity has dropped to 80% of its original capacity. For the purposes of discussions here, I think people refer to "calender life" as that same measure, in time. That is, people only care about that same measure translated to calendar months / years. So if I have an application where I cycle the batteries literally every day, if the cycle life is 2,000 cycles, then the useful calendar life of those cells is 2,000 / 365 = about 5.5 years.

If I may ask, where does your 10 years calendar life come from?
The internet and the internet does not lie it knows everything ? I read a report on lfp and it mentioned average calendar life is 10 years. But to be honest I don't think anyone knows the real world life expectations on either cycle or capacity with the way most of use them.
 
As @Ampster alluded to, my last comment is simply meant to clarify a misuse/misunderstanding of a term. Not to make or disagree with any broader point in whatever broader thing is being debated (I haven't followed this discussion closely enough to know, or have an opinion on that).

That's fine, but abusing the cells in terms of storage at high temperature or high SoC will reduce the number of cycles too
Precisely. Both cycle and calendar aging are factors impacting overall cell life, and both are dependent on a number of variables like SOC and temperature and in the case of cycle life C-rate. But to be clear, calendar aging is not only the result of abuse, it is just accelerated the further from optimal you get, but its a factor no matter what.

I stand by my statement, that most people care about how long their cells will last in their situation if they take care of them.
I'm not arguing with that statement, I agree with it for the most part.
What you would like to focus on (how long cells can be expected to last in real world conditions) is what matters to most of us most in a practical sense, its just not what calendar aging means.

Calendar aging + cycle aging together will determine how long your cells last, and both will depend on a number of variables that are not always aligned (for example, low temperatures are more optimal for calendar aging, but mild or warm temperatures are better for cycle life).

Point being its important to understand the difference between cycle and calendar aging, and understand that real world life of your cells is the product of both. Simply dividing the cycle life figure from the datasheet by 365 will not give you an accurate estimate. Both calendar and cycle aging will affect the real world life of your cells and specifically how and in what conditions you use your system will determine to what degree each factor matters.
 
As @Ampster alluded to, my last comment is simply meant to clarify a misuse/misunderstanding of a term. Not to make or disagree with any broader point in whatever broader thing is being debated (I haven't followed this discussion closely enough to know, or have an opinion on that).


Precisely. Both cycle and calendar aging are factors impacting overall cell life, and both are dependent on a number of variables like SOC and temperature and in the case of cycle life C-rate. But to be clear, calendar aging is not only the result of abuse, it is just accelerated the further from optimal you get, but its a factor no matter what.


I'm not arguing with that statement, I agree with it for the most part.
What you would like to focus on (how long cells can be expected to last in real world conditions) is what matters to most of us most in a practical sense, its just not what calendar aging means.

Calendar aging + cycle aging together will determine how long your cells last, and both will depend on a number of variables that are not always aligned (for example, low temperatures are more optimal for calendar aging, but mild or warm temperatures are better for cycle life).

Point being its important to understand the difference between cycle and calendar aging, and understand that real world life of your cells is the product of both. Simply dividing the cycle life figure from the datasheet by 365 will not give you an accurate estimate. Both calendar and cycle aging will affect the real world life of your cells and specifically how and in what conditions you use your system will determine to what degree each factor matters.

Trying to say the same thing in smaller paragraphs and using numbers:

Battery life = cycle life minus aging factors (calendar life).

Example:
4000 cycle LFP. Cycled daily.

4000/365 = 10.96 years, degrades 1.83%/year

Based on cycle life, the battery should last almost 11 years before capacity drops to 80% of rated.

In reality, the 80% level will be hit BEFORE 11 years because cells deteriorate over time whether used or not. Assuming a calendar degradation of 0.5%/year:

YearCycle LossAge lossCap
1​
-1.83%​
-0.50%​
-2.33%​
2​
-4.15%​
-0.50%​
-4.65%​
3​
-6.48%​
-0.50%​
-6.98%​
4​
-8.80%​
-0.50%​
-9.30%​
5​
-11.13%​
-0.50%​
-11.63%​
6​
-13.45%​
-0.50%​
-13.95%​
7​
-15.78%​
-0.50%​
-16.28%​
8​
-18.10%​
-0.50%​
-18.60%
9​
-20.43%​
-0.50%​
-20.93%
10​
-22.75%​
-0.50%​
-23.25%​
11​
-25.08%​
-0.50%​
-25.58%​

80% level would occur in the 9th year.
 
  • Like
Reactions: Dzl
Broke out the new power supply last night, installed all of the grub-screw terminals, hooked the cells up parallel, and put the charge on them, initially to 3.4V. They've been on for over 18 hours and have only gained about 2 tenths of a volt. LOL They started out drawing about 9A, now down to 5.

I'm doing the step method - to 3.4V and rest, then 3.5 and rest, then 3.6, call them balanced. I'm guessing the second and third steps will go quicker.

So far so good...

Did the cells expand at all when top balancing without compression? I'm about to start top balancing a set of the same cells and would like to avoid using a fixture during this step.
 
Did the cells expand at all when top balancing without compression? I'm about to start top balancing a set of the same cells and would like to avoid using a fixture during this step.
On the two sets of 8 cells I've gotten (8 x 280Ah and 8 x 230Ah) my charging current was 30A for 8 cells, which works out to be less than 4A per cell. I didn't have a compression fixture and didn't see any swelling. I think if you don't go above 3.65V and go with a reasonable current (per cell) they won't swell. That's my experience anyway.
 
Did the cells expand at all when top balancing without compression? I'm about to start top balancing a set of the same cells and would like to avoid using a fixture during this step.
Zero zilch flat as a pancake.
And they stay balanced within 4mAh.

After top-balancing I did assemble them into a compression jig:



 
Last edited:
Another happy customer here. Purchased 8x 230AHr w/ welded studs from Docan. They arrived at my package delivery place in Point Roberts within just a few days of my order, from the Houston warehouse. Packaging was great, and using a calibrated Fluke multimeter I was able to verify that all cells were 3.2930V +/- 0.0015V or so. Next step is to get them into a fixture that I can mount securely in my 27' sailboat, and integrate with the rest of the system I'm building.66440394156__7E56A567-32BA-4E7C-B77A-1E706325F2FA.jpeg
 
Another happy customer here. Purchased 8x 230AHr w/ welded studs from Docan. They arrived at my package delivery place in Point Roberts within just a few days of my order, from the Houston warehouse. Packaging was great, and using a calibrated Fluke multimeter I was able to verify that all cells were 3.2930V +/- 0.0015V or so. Next step is to get them into a fixture that I can mount securely in my 27' sailboat, and integrate with the rest of the system I'm building.View attachment 80612

What freight company in Point Roberts did you use?

No issue driving down there to get them?
 
What freight company in Point Roberts did you use?

No issue driving down there to get them?
I used https://inoutparcel.com Point Roberts is explicitly exempted from a bunch of the COVID border restrictions, so you're allowed to run down to pick stuff up and/or hit the gas stations, but not allowed to do social things. You still need to fill out the ArriveCAN app before returning to Canada, and it will throw a nasty warning about you not having a COVID test, but that's fine at Point Bob, just make sure you indicate the Boundary Bay border crossing in it.
 
On Jan 21, 2022 I received the 8 160AH Eve cells that I ordered from Docan Power on Dec 30. So 3 weeks. One week in I had inquired about the status and was told that something was on a truck to Houston and it would be 1 or 2 weeks. That proved accurate. I had ordered extra busbars which could have been the hold-up, or maybe the cells were not in stock. I don't know but I am satisfied with getting them in 3 weeks. I did receive 6 extra busbars for a total of 14.

They were shipped by UPS.

Each is labeled at 3.29 volts and 0.26 mOhms IR. and the UPS label (or whatever they are called) was intact. Each has LF160, a serial # and 515.2Wh

Excellent packaging in the same foam crates you see in other's photos. Perfect condition visually. Initially they all measured at 3.295 volts with my Fluke DMM. The next day after they had warmed up inside, I had 3 at 3.297V and the rest at 3.296. I don't see any signs of bulging.

I also labelled and weighed each cell. The lightest was 2.961 kg and the heaviest was 2.967. A difference of 6 grams and only 0.2% which is amazing. Does this matter? Not sure. You tell me.

I don't have the equipment to capacity test these cells, but I will parallel them and charge them toward a top balance. My CV/CI power supply can only do 6 amps, so that will go very slowly. I doubt I'll go above about 3.4 volts. My application does not demand maximum possible capacity.

P1030129.JPG

P1030133.JPG

Some details:

The M6 terminal studs are quite short. There is only 10mm of thread. The contact area also seems quite small at only 10mm diameter of which 6mm is taken up by the stud itself.

On these cells, the busbars will work with the cells end-to-end as well as side-to-side which is nice.

The busbars consist of 2 fairly thin layers of metal attached together with heat-shrink. The total thickness is about 1.5mm. I have enough to double them up when connecting in series and I may do that. Perhaps I will have to re-wrap them in sets of 4 though. I am planning to use Batrium Blockmon cell sensors, so it may depend on how thick those actually are.

I determined that the aluminum chassis of the cells is connected to the positive terminal. I think this is normal. Obviously one needs to be very careful not to damage the insulating cover on all sides.

Looking good. Thank you Jenny Wu!
 
The M6 terminal studs are quite short. There is only 10mm of thread. The contact area also seems quite small at only 10mm diameter of which 6mm is taken up by the stud itself.
If it is like most others purchased by folks here on the forum, the contact area is actually 11mm, but that one mm doesn't make much difference. I think I've settled on a 1/4" aluminum washer as the best way to get the maximum contact area. Put the washer down on the stud first (perhaps with some MG conductive paste under), then put the bus bar down on top of that.
I determined that the aluminum chassis of the cells is connected to the positive terminal. I think this is normal. Obviously one needs to be very careful not to damage the insulating cover on all sides.
It seems to be a point of debate, but I don't think the aluminum case is actually connected to the positive terminal. I think there is a voltage between the case and the negative terminal (or maybe even both terminals on some cells), but it is a surface / static potential, not a real potential. That is, if you put a load between the case and the negative terminal, you would likely find no current flow. I know that is the case on some of the cells I've messed with, but I can't say it is the case on all cells.
 
I don't think the aluminum case is actually connected to the positive terminal. I think there is a voltage between the case and the negative terminal (or maybe even both terminals on some cells), but it is a surface / static potential, not a real potential. That is, if you put a load between the case and the negative terminal, you would likely find no current flow. I know that is the case on some of the cells I've messed with, but I can't say it is the case on all cells.
My guess is the jelly roll of anode and cathode which is stuffed in the aluminum case has some contact with the case and causes that. I agree is is not consistent but to reduce leakage, I cut plastic file folders into sheets to separate my cells.
 
On Jan 21, 2022 I received the 8 160AH Eve cells that I ordered from Docan Power...>snip

I don't have the equipment to capacity test these cells, but I will parallel them and charge them toward a top balance. My CV/CI power supply can only do 6 amps, so that will go very slowly. I doubt I'll go above about 3.4 volts. My application does not demand maximum possible capacity.

Very slowly is right. My four 230A sat at 5-7A for four days before I thought maybe something was wrong. Nope, just a LONG way to go.
3.4 is too low if your current drops off any, and if so set your supply for 3.6 or even 3.65 then connect and turn on. Ring terminals recommended in lieu of alligator clips. In fact, 10G wire is also recommended - make up your own charge leads.

After going through this I now understand the value of the recommendation to wire them in series with your bms and bulk charge them to 3.4/cell with a 12V charger. THEN change them to parallel and finish topping them up. They arrive at about 30% SOC and take a LOT of amps to get nearly full charge. If I was local to you I'd lend you my 10A power supply. It still takes forever.

The good news is my cells stay within 4mAh volts up and down SOC.

 
Very slowly is right. My four 230A sat at 5-7A for four days before I thought maybe something was wrong. Nope, just a LONG way to go.
3.4 is too low if your current drops off any, and if so set your supply for 3.6 or even 3.65 then connect and turn on. Ring terminals recommended in lieu of alligator clips. In fact, 10G wire is also recommended - make up your own charge leads.

Browneye: thanks for the advice. I'm thinking I'll put them in the final compression frame first and then make up some diagonal connections to parallel them. Maybe use some left over 6 or 4 AWG wire I have with lugs of course.

This arrangement would also allow me to charge in series (6A x 25V instead of 6A x 3.xV) for awhile first. That might make sense while I manually monitor the cell voltages. Not clear into the knee of the curve though for sure . But the Eve data sheets say the knee starts at around 3.3V. I have not ordered the BMS yet so can't use that.

I can't find a datasheet specifically for the Eve 160AH cells so I am looking at the 280AH curves.

I'm also not clear how much the V rises just because of the charge current. The low IR implies not much, but I won't know until I try it. I also figure that charged cells need to settle off the charger for awhile to see where they are really at.

Or I could just wait for the BMS and ... but that basically means wiring up everything before I can top balance and I'd like to top balance while I wait for the BMS and some other stuff I still need to order.

Or I could just buy the B&K 1900B that I covet for doing this job ;)

I'm nervous about top balancing without a compression frame or even removing the fully charged cells from the frame to turn them around for series wiring. Maybe no issue, but I'm new to this and want to be careful since I'm learning as I go. Probably 4 or 6 AWG is way overkill for the balancing current I have, but my impression has been that the goal is to get the cells as exactly equal in voltage as possible - beyond what can even be measured accurately because of the flat curve. So low ohms between cells seems good.

I also have a bunch of Flexibar I got cheap (2 layer, discontinued product) that I could use instead. Avoids lugs and crimping and associated issues. Seems like a waste of good Flexibar though.

I was going to just use the Batrium (I plan on using) to balance the cells, but I see that its "auto leveling" mode is not recommended for LFP, so I'm rethinking that idea.

I have trouble also understanding how 3.29 volts can be 30% SoC. The Eve curves indicate that 80 or 90% of the energy is between 3.1 and 3.3 volts and the knees are very steep so there isn't much energy above 3.3 or below 3.1 Seems like 3.29 volts is closer to the top then that - or probably I'm just not understanding something here.

Eve seems to publish discharge curves, but not charge curves. Maybe that is why I am confused. How different are they?

Heh, after I'm done with this I might have some idea of how to do it!
 
This arrangement would also allow me to charge in series (6A x 25V instead of 6A x 3.xV) for awhile first. That might make sense while I manually monitor the cell voltages. Not clear into the knee of the curve though for sure . But the Eve data sheets say the knee starts at around 3.3V. I have not ordered the BMS yet so can't use that.
Just be very careful charging in series without a BMS. One cell will reach the max of 3.65V before the others, and without a BMS you are likely to go well above 3.65V with that cell. The charge curve is almost vertical when you get close to the end, so there isn't much time to notice.
I have trouble also understanding how 3.29 volts can be 30% SoC. The Eve curves indicate that 80 or 90% of the energy is between 3.1 and 3.3 volts and the knees are very steep so there isn't much energy above 3.3 or below 3.1 Seems like 3.29 volts is closer to the top then that - or probably I'm just not understanding something here.

Eve seems to publish discharge curves, but not charge curves. Maybe that is why I am confused. How different are they?
As you seem to already know, the charge curve is different than the discharge curve, and both are different than the resting voltage. Most tables and such that you see regarding voltage vs. SoC are using the resting voltage, meaning no charge or discharge for some amount of time. At 3.29V resting, you are certainly below the 50% SoC.
I'm nervous about top balancing without a compression frame or even removing the fully charged cells from the frame to turn them around for series wiring. Maybe no issue, but I'm new to this and want to be careful since I'm learning as I go.
Charging for top balance with the cells in parallel is less likely to cause the cells to swell, because the current is split between the cells. It sounds like you are working with 8S (since you mentioned 25V). If you have 24A charging the cells in parallel, each cell is only seeing a nominal 3A (24/8) going through it. If you are putting them in series and running the same 24A, each cell is seeing the full 24A go through it. That may not be a high risk of causing swelling, but it is a higher risk than the 3A case in parallel.
 
Same for me. Arrived today. First impressions (Only opened 1 box): They look great, no visible damage. (But I wasn't expecting anything less - and if it was beaten up it wasn't likely to be Docans fault)

It was supposed to ship by sea, but I personally doubt it. The DPD tracking started at a logistics hub in Poland, near Warsaw. That's quite a strange port ;) (The nearest port with a container terminal is like 300km away)
All orders from other sellers went by train and also arrived in Poland in the EU, and UPS/DPD took care of the final part of its journey.
And next to the DPD hub is a rail container terminal so I think they actually came by train, not by sea.

Also: The Netherlands (where I am) does have a rather large containership terminal. It doesn't make sense to ship them first all the way to Poland. Rotterdam or Antwerpen as port destination would make much more sense.

Doesn't matter to me, the cells are arrived and that's all that matters, in about 2 months, including the holidays so I'm not complaining!

And: I noticed UN3481 labels on the box. This is the first supplier so far which had these boxes properly labelled

Cool!

Do you have to pay taxes or custom when ordering to the EU? I am thinking of ordering to Sweden soon.
 
On the two sets of 8 cells I've gotten (8 x 280Ah and 8 x 230Ah) my charging current was 30A for 8 cells, which works out to be less than 4A per cell. I didn't have a compression fixture and didn't see any swelling. I think if you don't go above 3.65V and go with a reasonable current (per cell) they won't swell. That's my experience anyway.
Horsefly thanks for the feedback and advice.
 
Thanks for the input Horsefly and nhhiker. Yes: 8S, 160AH, 4 kWH. A teensy battery. I've seen warnings about charging sans BMS before, so I think I'll take that idea off the table, even at just 6 amps. Those curves do look awfully steep at the ends to be watching for 1 cell to go critical by hand with a DMM. I think I'll set them up for series wiring with the busbars in a frame and make some sort of crossover connectors so I can charge in parallel. I have nothing but time - and besides I'll be waiting for that shipment from Australia ... Probably over-cautious but I'd rather be safe than sorry as they say.
 
Back
Top