Lipo Longevity

Aspendell

Aspendell

Member
Joined
Jan 31, 2017
Messages
70
What voltages should we shoot for so that we can maximize both longevity and capacity of our packs?

I want to encourage everyone to check out BatteryUniversity.com. It's pretty much the authority of physics and empirical testing of batteries.

For me longevity is more important than maximizing capacity of an individual pack, as its still cheap to just build more so we can make our window of usage smaller (ie., 3.2 - 3.9v)

Here is just one excerpt from Battery University that is very applicable to what we are doing:
Similar to an EV, Li-ion in satellites must also endure a lifespan of 8 years and more. To achieve this, the cells are charged to only 3.90V/cell and lower. An interesting discovery was made by NASA in that Li-ion dwelling above 4.10V/cell tend to decompose due to electrolyte oxidation on the cathode, while those charged to lower voltages lose capacity due to the SEI buildup on the anode.

NASA reports that once Li-ion passes the 8 year mark after having delivered about 40,000 cycles in a satellite, cell deterioration caused by this phenomenon progresses quickly. Charging to 3.92V/cell appears to provide the best compromise in term of maximum longevity, but this reduces the capacity to only about 60 percent.


image_fhijha.jpg


~Aspendell
*I wasn't sure where to put this post so please feel free to move it.
 
Another argument for bigger packs so that we can have a lower discharge rate C:

Discharge characteristics of NCR18650B Energy Cell by Panasonic.


image_fhijha.jpg


So as we can see, at least with a new cell there is nearly a 1000 mah difference in delivery at 0.2 of max capacity discharge rate compared to at double the mah capacity rate of the cell. And that is just the current. But actual Power delivery is P = I * E. So when you add the extra 1 volt difference (33% higher) and the fact that at 0.2C the cell can stay above the 3.0v cutoff 33% longer, and you have nearly 200% wattage delivered from a cell at 0.2C compared to one discharged at 2C.

so BUILD BIGGER PACKS! like ol HBPW :cool:
 
Yes you should always go as low C-rating as you can. That saves alot. I have set max C-rating to 0.5 in my setup and avg will be far far less. Im talking about well below 0.05C :)

And as seen what you posteed. Going above 4.1 does not gain much at all. Not even above 4.0v.
 
Aspendell said:
I want to encourage everyone to check out BatteryUniversity.com. It's pretty much the authority of physics and empirical testing of batteries.
Click to expand...

Bullshit!

Battery University is simply a large collection of other people's research, articles and random blog and forum posts etc. No real checking is done to ensure the information is correct. Expect to find one article that disagrees with another.

Lets see what wikipedia has to say:
https://en.wikipedia.org/wiki/Talk:Lithium-ion_battery#batteryuniversity.com
Yeah :rolleyes:
 
Right.
Just make sure you saturate charge at your chosen voltage or you are sacrificing another 5-15% of max capacity as seen in this table:

image_qdkmla.jpg



This is why some new players who are just charging to 4.2v or even 4.22v and then doing their capacity check right away are coming up with some lower numbers. I always charge fully then put the cells in my to-be-capacity-checked queue. But then I recheck them 1-40 jrs later right before they are ready to enter the check, to be sure they are at least 4.19v, or they get further saturation charge.

~Aspendell
 
But you are never going to utilize the cell like that in a power wall. Hence those numbers are quite irrelevant. Ie in pack you charge and will most likely discharge rather soon again for most of us. Also note that most chargers stop charging when reaching 100mA or some % of the max.

Also the diff you could gain is less than 1% in most cases. Note that the equipment used by most people have an error rate >5-10% hence once again irrelevant to this part i would say. I also never pack my packs at a 1mAh rate since if you retest you will see that it diffs alot and if you use more than 1 charger you also see that it easily diff 5-10% between them (The cheaper ones)

Lastly i have never seen a cell charged properly to 4.2v stay above 4.16 after 24 hours. That is very rare to this chemistry i would say :)
Note that self discharge on the cells is quite high on most of the ones we use. You can find LiIon that have down to 1% each year but they are damn rare! Most cells is around 5% per month! And 5% over 1 month is quite alot if you look at the voltage curve.


Its just my 5 cents to it. There is plenty of strange articles on BU and as said you need to take many of them with gloves since there is many strange scenarios.
Also dont forget that all cells differ and it also differs alot between those used 200 cycles and those used 500.

And if we take all above into 1 and the same advice as i tend to say to everyone: Crosscheck your chargers/testers among them each other to make sure they are pretty close. Then only use the number you get as a health factor kind of. Because a tester for 20-30USD will not be even close to exact....

:)

There is a reason why manufacturer also gives a % when it comes to typical capacity.
 
You make good points Daromar, but the good chargers take into account that a Lipo isn't fully charged until the top end charge reaches 3% of capacity. For a 2000ma cell this would be the cell accepting 60ma or less. For some cells this can be up to 2 hrs after the cell initially reaches 4.2v as seen in the previous graphs, and this can represent up to 15% capacity.
But you are right as this is only really relevant for initial capacity checking for most of us, but not for those that are using them in RC.

As an empirical test, I quick charged some new dark grey cells I got yesterday to 4.2v then immediately threw them on the capacity checker and they came in at around 2850mah. Then I held the rest at 4.2v for 1hr+ and ran them overnight and they came in at 3300mah+ each!
 
But at what current did you stop charging? And what charger?
I would like to see the charge curve between those 2 tests. And also the discharge curves if you can get that.

Yes the proper chargers to top-charge as they should. Opus stop when they reach some where below 100mA. (Not sure right now) And iCharger that i use you can set this predefined and yes that changes the cap test alot! Along side with the C rating used of course.

Its not until then (At least I can say) that the numbers are proof enough for the "general" testing done in the forum.
Note that I do trust your numbers i just want to make sure that I understand the diff and if it may have been due to a premature charge on the first test or not.
Have seen chargers stopping charging at CC instead of actual CV-CC.

Above numbers to me looks like that the charge was done at 1A but it stopped when it reached 4.2 directtly instead of going into CV :)

I like good discussions like this :)
 
By my research it appears that the Opus is the best of the sub $300 chargers out there. It was designed by a US company then sold to the Chinese kinda like IBM, so good engineering. Some of the competitors (not to mention names) use really funky algorithms to actually compute the mah rather than just pure measurement. They figure there are X number of mah left after we reach our cutoff discharge, so we will just calculate (add) that into the total!

I plan to get an Opus once I can afford it. For now I'm using a module that measures pure mah as they pass through a purely resistive load 2.38 ohms. I spent a whole day calibrating this thing so the mah readings are within less that 0.5% accuracy. So since my initial draw is around 1.68 amps (half that at cutoff) and my OC cutoff voltage is around 3.4v, my actual totals are low if anything. I figure I could nurse at least 15% more out of my totals if I was willing to take 5X as long on each cell.

But I totally agree that people don't give enough credence to difference in measurement systems when comparing numbers.
 
What module? What voltmeter did you use to measure as accurate as 2.38Ohms?

What cells are dark grey? Because not many cells are as high as 3300mah and if you say you could gain 15% more thats 3795mAh. Thats even more than most manufactures can make today....
 
The module is a chinese model that measures pure current through a micro shunt. I compared the current readings at all voltages from 0.5 - 5.0V with 2 separate multimeters inline and current matched my 2 meters within 0.01A at all voltages, up to 3.0A, the max for the shunt. Like I said, it took me a full day to calibrate it to that level of accuracy, as it was precalibrated off by about 30% when I received it. It measure OC voltage occasionally, so I set 3.4v as the cutoff. This is typically 2.6-3.0v under the ~1.1 amp load right before cutoff.

The resistance is really irrelevant as we are measuring pure current discharged from the battery. I decided on the 2.38 ohms purely as a compromise between speed of processing and loss of max capacity based on discharge rate. The important thing, as you said, is that we perform all the tests exactly the same for comparative reasons when building packs. And of course I didn't want to exceed 1C at full voltage.

I have a CC module on the way that can do up to 9.99 amps sustained, and I plan to retest several cells to see how they compare. But I have retested 3 cells from full charge on this module and they came up within 2 mah of each other on each run. Out of like 2800mah, difference in room temperature would account for a bigger deviation than that on the electronics :)

Even on $40 commercial models they only calibrate the shunt manually within 3% (compared to mine <0.5%) and you can see a 5-10% difference discharging at 500ma compared to 1000ma.

~cheers
 
Still:

What cells did you use that could potentially do above 3700mAh?
What module did you use to measure the load that you can calibrate with several readings as input so you get proper result?
What resolution of the ADC in that module?

I have a hard time to believe the 0.5% accuracy without knowing what hardware you did use to get to that conclusion.
 
image_giwdst.jpg

image_yajorj.jpg

My Phones flash tends to blue the colors a bit. But has anyone else seen these grey cells, or know the manufacturer?
 
daromer said:
http://www.globalspec.com/FeaturedP...logy/Energy_CellsCMICR18650F82600mAh/163081/0
http://www.terapeak.com/worth/lot-o...-7v-rechargeable-li-ion-battery/252243257135/

And the datasheet for them:
http://www.homedepot.com/catalog/pdfImages/19/1938cc70-fc49-4077-88cf-83f700767f9b.pdf

2600mAh and 9.62Wh

So you should most likely crosscheck your numbers once again im afraid. (Unless there is some sort of above with higher but i cannot find any right now)
Click to expand...
Dar, thanks for checking. But they definitely aren't those ones. Those are F6, mine are F8 (last 2 digits before dash). Those are teal, mine are dark grey. I do have many that you linked, and they are checking in the 22-2400 range.
The only reference I've found online so far is from vapping sites and they say they may be from Macbook Pro packs?
 
The datsheet says:
CMICR18650F8

And both links says F8 if you look closer. I see no ref to F6
 
I see that now. I was going by the pictures. On the first one they have a completely different battery in the picture than what they have listed. But that website is obviously defunct, as none of the links go anywhere. The ones on the Terapeak site are the right color (better than my pic for comparison) but the text is completely different. specifically the -81 and the -HENV on the first line of mine.

That being said, I'm sure you are right that these are rated at 2800mah, hence the -28A-. But what Im interested in is what these are testing at for others. Otherwise they would be -30A-. HBPW (Pete) mentioned that he had a batch recently that all tested over 3000mah. I wondering if these grey cells were included amongst them. If you look at the spec sheets for rating these cells you will see things like ">= 2700mah at 560mah discharge to 2.75v under load". So that is merely a minimum for the weakest in the classification.
 
In answer to your earlier question re. charging. Right now I'm primarily using a programmable DC-DC boost converter connected to my 1200 watt server power supplies which are configured for 12.5v @ 96 amps for charging lipos (usually 24v @ 48 amps). So I have a 4S cell holder that I can charge 4 batteries at a time, or 4 pairs, under constant current to the preprogrammed peak voltage. Then I wait til a couple hours before I'm going to throw them into the checker to put a saturation charge on them to top them out at 4.2v. Oh and I monitor the voltage of each of the 4 cells simultaneously.

But none of that matters to the actual checking itself. A fully charged battery is, or it isn't, when it comes to capacity test. Same with the resistance. It has no bearing in the accuracy of a capacity check because the the processor samples the shunt 5X per second, so unless you can change the current draw faster than that, it doesn't matter where the current is going, it only matters that it is measured leaving the cell.
the only other thing that could affect the mah readings of my setup would be if the clock in the processor were changing significantly during a test.
All we are doing is measuring how many mah we can pull out of a fully packed cell. It says nothing about the pressure (voltage) that they were pulled out at. But, at least in my case, the mah were definitely pulled out of the cells and not just computed. And most of the cells normalized to 3.4-3.5v open circuit after the test.

*On that note, I have a bunch of dark green cells that deliver 90% of their current at 2.9v or lower under load. So while they may still have 2600mah, they will only deliver 2/3rds of the power if put into a pack. But that is a different topic...

~On a side note I have done several of the dull red Sanyo cells today that were so hard to charge. I have them all labeled HOT in big letters. But after I figured out to charge them at .2C and topped them off overnight, they have all been in the 26-2800 range in checker today without even getting warm!
 
This is the one I have on order that has Constant Current and a few other features

image_shnefg.jpg

Ebay link

Should be here any day now, then I can start comparing. :shy:
 
You must log in or register to reply here.
Back
Top