EVE 304 Cells from 18650batterystore & zke-40 testing.

[Sam Cho TX]

Well after seeing a slew of videos it seems to me that there is not much difference between grade A and B, so why not try some from 18650batterystore. Anyway I previously tested some of the EV grade from Sunfunktis and was pretty happy with the results, but they have raised the price a bit since then, here is that thread: https://diysolarforum.com/threads/new-toys-arrived-zke-40-tester-sun-fun-kits-eve-lf304-cells.56015/

I could use a few batts so when the 304 came back in stock I ordered 4.

Well... they are packaged Ok I would say, not as nice as Sun Fun Kits but no damage, however, I can already see some issues. I see 2 of my cells are 302AH and 303AH so this is already below the 304 18650store claims on there page... guaranteeing to pull 304 AH.

Also the welds on the terminals look pretty bad looks like they even burned the terminal plastic a bit.

Anyway my total cost was $149.33/cell shipped so lets see how they test. The Sun Fun Kits cells where $202/cell shipped so we are about $53.33 more per cell.

 

[Alkaline]

You mean to tell me after all that hoopla you went on about on the vendor forum you then go out get some grade Bs?

 

[Sam Cho TX]

You mean to tell me after all that hoopla you went on about on the vendor forum you then go out get some grade Bs?

I'm in no mood for your nonsense, why do you all feel the need to abuse new commers? If you don't like what I have to say, plenty of other threads for you to enjoy. Leave me in peace ok??

 

[Alkaline]

Ok nevermind, sorry. I am looking forward to your results.

Thanks

 

[Sam Cho TX]

And now the wait begins...

 

[Bud Martin]

And now the wait begins...

What is that graph showing?
And what is that big dip of the Red line?

 

[Sam Cho TX]

Not sure I think I might have knocked off the v-sense wire, I'm currently charging the cell at 40 amps... assuming it arrived at 50% charge it will be a few hours to full charge.

Once done I'll discharge at 30 amps to 2.5v

 

[sunshine_eggo]

I'm seeing 3.65V and 4.4A, so it looks essentially full to me.

The current oscillation is likely just the control loop over-adjusting.

 

[Sam Cho TX]

Ok well this is very strange it only took me like an hour or to fully charge this cell. The sfk cells took almost 4 hours?

 

[Alkaline]

They maybe sending the grade B at a higher state of charge to compensate for the potentially higher self-discharge rate.

 

[Sam Cho TX]

Well I don't think I charged this cell up properly it only took like one and a half hours to charge. The test results are pretty terrible so I will recharge this cell and then do a full capacity test again this time with another cell as well.

But initial results are pretty bad 281.7 AH and 913.18 WH, however, I'm thinking its because of my error (I hope) and not the cell.

By going with the WH/3.2v method it comes out to about 285.3 AH. Still well short of the claimed guarantee to pull 304 AH. I'm attaching the zke report csv in zip format.

 

[Alkaline]

Hrm, looking at your graph it seems ok you started at 3.54v / cell which is basically full. I would perform a charge test. See how many AH you put back into the cell. The cell said it tested to 302 AH?

 

[sunshine_eggo]

Well I don't think I charged this cell up properly it only took like one and a half hours to charge. The test results are pretty terrible so I will recharge this cell and then do a full capacity test again this time with another cell as well.

But initial results are pretty bad 281.7 AH and 913.18 WH, however, I'm thinking its because of my error (I hope) and not the cell.

By going with the WH/3.2v method it comes out to about 285.3 AH. Still well short of the claimed guarantee to pull 304 AH. I'm attaching the zke report csv in zip format.

You didn't do anything wrong.

These cells are FULL at 3.65V and 15A per the datasheet. You more than met that requirement.

The datasheets detail the discharge test.

 

[Alkaline]

Well if you go by the 285 figure (yes I know controversial but whatever) you have 285 / 304 = 93.75% of rated capacity
If you go by 281.75 / 304 = 92.68% of rated.

I would test all of the cells you have.

 

[Sam Cho TX]

So I fully charged the cell and it says it took in 295 AH and 986WH. Why such a big difference between discharge and charge? Guess I will run a discharge again, so this is how my weekend will be, cleaning attic and testing cells?

 

[sunshine_eggo]

So I fully charged the cell and it says it took in 295 AH and 966WH. Why such a big difference between discharge and charge? Guess I will run a discharge again, so this is how my weekend will be, cleaning attic and testing cells?

The datasheet indicates the discharge test may be run up to 3 times if it falls short, so that implies cycling may increase tested capacity.

And while near 100%, the charging efficiency isn't 100%.

EDIT: This is especially true when looking at it on a Wh basis. Discharging drops the voltage and results in less Wh drawn while charging increasing the voltage resulting in more Wh input.

Lastly, I would look for some other way to confirm that the current measurement on the tester is correct. if you have a shunt or a DC clamp ammeter, that's good validation of the two instruments read the same current. Otherwise, a very small current measurement error could make a noteworthy difference in the results.

 

[Sam Cho TX]

Ok So the 2nd test produced much better results, 297AH. Still short of the 304 but not too bad. I'm not sure what happened on the first test. Next I will test the 303 cell, I'm not sure why the first test was bad but I think it was because it did not charge fully.

Using the WH/3.2 Method it means I have 301.5 AH, very close to rated, not bad at all.

 

[Alkaline]

Interesting, I have not seen this before, perhaps you have what is called a RUNNER cell something that jumps to full voltage when its not, these you usually have to keep the charge on a good bit longer to fully saturate.

 

[sunshine_eggo]

Ok So the 2nd test produced much better results, 297AH. Still short of the 304 but not too bad. I'm not sure what happened on the first test. Next I will test the 303 cell, I'm not sure why the first test was bad but I think it was because it did not charge fully.

Per my comments above, the datasheet strongly implies that subsequent discharges my be higher than the initial discharge. A test may be run 3 times to establish the capacity of the cell.

Your results assume that your equipment is accurate within 2.3%. Have you validated voltage and current readings with calibrated equipment?

Using the WH/3.2 Method it means I have 301.5 AH, very close to rated, not bad at all.

The Wh/3.2 calc is inaccurate. If a device measures and reports both Ah and Wh, those values are accurate.

Wh is an integration of the instantaneous Amps * Volts over time. The device is taking the average current and average votlage over a short period of time and adding it up over the entire course of the test.

Interesting, I have not seen this before, perhaps you have what is called a RUNNER cell something that jumps to full voltage when its not, these you usually have to keep the charge on a good bit longer to fully saturate.

This is incorrect. A RUNNER is a cell that is at higher state of charge than the other cells in the battery. It is not possible to identify a runner as an individual cell.

 

[RCinFLA]

Did you use the remote voltage sensing directly on cell terminals? This is necessary to avoid high current lines voltage drop.

3.300v initial voltage slump for only 30A load says they are used cells. This assumes you used remote voltage sensing.

It is good to insert a 5 minute, zero discharge current rest period, at least at three points, approximately 75% SoC, 50% SoC, and 25% SoC. You can add more SoC points if you want more info, up to number of cycle sequences allowed by tester. The tester will still compute correct AH results with the no-load 5 minute rest periods.

Looking at the voltage slump after load picks up again gives good info on quality of cells. The less voltage slump the better. For 30A discharge on a 304 AH cell you should only have 50-60 mV max. slump from rested open circuit voltage. The 3.300v shown on first 25% discharge says you have about 100 mV of terminal voltage slump with 30 amp load. That is a high amount of slump voltage.

The voltage slump gives an indication of cell impedance which increases with aging/use of cell.

You should use 40 amp discharge to get a better test on cell quality. 40 amps is max of tester and still marginal on load test current. You would really like to have discharge current between 0.2 to 0.4 C(A), 60-120 amps for a 304 AH cell.

This curve is approximately what you should expect for terminal voltage slump for various amounts of discharge load current.

 

[sunshine_eggo]

Looking at the voltage slump after load picks up again give good info on quality of cells. The less voltage slump the better. For 30A discharge on a 304 AH cell you should only have 50-60 mV slump from rested open circuit voltage.

Does this still apply when there is residual surface charge immediately following a charge? 3.30V as an initial voltage sounds about right to me.

 

[RCinFLA]

Does this still apply when there is residual surface charge immediately following a charge? 3.30V as an initial voltage sounds about right to me.

No, but that quickly dissipates out within the first few percent of discharge. I am looking at 90-80% SoC range on the 3.300 vdc terminal voltage. No load in that range should be about 3.39x volts, therefore the comment about 100 mV slump.

A new cell will run about 3.34 - 3.320v with 30A load in the 90% SoC range. That is only 20-30 mV additional slump but that is 0.7 to 1 milliohm of additional cell ionic migration impedance at 30 amps which is significant. When you start talking about 70-100 amps of use load it gets worse.

The greater the average slump the greater the delta between charge and discharge total AH's. For only 0.1 C(A) load, the round trip efficiency should be about 98.7%-99% at only 30 amp load. That high of efficiency is hard to resolve based on charge and discharge end point selection.

When you only have 0.1 C(A) loading you don't have large mV numbers so any errors in readings have a large impact on results. That is why it is better to have greater load currents during tests.

 

[Alkaline]

The Wh/3.2 calc is inaccurate. If a device measures and reports both Ah and Wh, those values are accurate.

Wh is an integration of the instantaneous Amps * Volts over time. The device is taking the average current and average votlage over a short period of time and adding it up over the entire course of the test.

It is perfectly accurate, Watts = volts * amps and watt-hours = nominal voltage * amp-hours. I had some doubt about this but having applied the above equation even my 280K Luyuan meet passing results. The manufactures are rating the amp-hours at 3.2 nominal. I have actually found this to be the most accurate way to measure capacity assuming you have a v-sense capable tester.

Amp-hour is a measure of electric current
Watt-hour is a measure of electric power

Looking at Sam's 2nd test he measured 297.1 AH and 964.88 WH

So doing basic algebra: 964.88 WH = 297.1 x (Nominal voltage)

Nominal voltage = 964.88 / 297.1

Nominal voltage = 3.247 volts.

So you are measuring the capacity at 3.247 volts, but these cells capacity are rated at 3.2volts. So the WH/3.2 makes perfect sense if you want to know capacity at 3.2V

In other worlds you can get 297.1 AH out of this cell if you draw at a constant 3.24 volts for 1 hour

Or

you can get 301.5 AH of of this cell if you draw at a constant 3.2v for 1 hour.


I understand some people have issue with this, but I have applied this above with over 50 or so results and its is the most accurate way of testing cells. Otherwise you will get cells that only average like 3.15 v nominal voltage showing same AH as something that has 3.25 v nominal voltage. Yet 1 has way more energy.

FAA, TSA, FEDEX, UPS, UN3480 all require battery to be listed in WH and not AH, not to mention your utility bill, electric cars, the grid as a whole all are based on Watt-Hours.

Lithium batteries with more than 100 watt hours | Transportation Security Administration

Spare (uninstalled) lithium ion and lithium metal batteries, including power banks and cell phone battery charging cases, must be carried in carry-on baggage only.

www.tsa.gov

PackSafe – Batteries, lithium

www.faa.gov

https://www.ups.com/assets/resources/media/en_GB/pack_ship_batteries.pdf

 

[Alkaline]

I don't want to make a fuss about WH Vs AH, we are at best taking about 4-5 AH difference either way...

 

[sunshine_eggo]

@Alkaline

NOMINAL voltage is a definition. It's always 3.2V.

You are literally just making stuff up.

Taking MEASURED Ah and multiplying by nominal voltage (3.2V) is only an approximation of Wh

Taking MEASURED Wh and dividing by nominal voltage (3.2V) is only an approximation of Ah.

Just because you've done things wrong many many times, it doesn't make them right.