Samsung 30Q INR18650-30Q 3000mah 15a cell
- Thread starter magudaman
- Start date
okashira
10 kW
I am just dividing discharge A-h by charge A-h on each cycle.Teh Stork said:okashira said:
Very interesting information. Is the CE-test rig self built or do you use finished equipment?
Are there any results of the Model S cells out?
Again, very cool
okashira
10 kW
Teh Stork said:
It's close, depending on the current. It's around 0.1% at ~2A and above. Voltage accuracy is around 0.1%. is around that.
But it doesn't need to be accurate. It just needs to be repeatable, which is much easier to achieve.
That and you can take an average of many samples to increase accuracy, which is what I am doing.
I did calibrate it with equipment good to 0.03% and 0.01%. But the ADC's and DAC's of the tester limit it's accuracy.
ScooterMan101
1 MW
What about these batteries ? : http://www.keeppower.com.cn/products_detail.php?id=520
KP IMR 26650 battery 5200mAh max 30A discharge 3.7v lithium battery UH2652
* Quality High Drain Li-ion Rechargeable Battery
* Brand: KeepPower
* Size: 26650
* Part No.: UH2652
* Nominal Voltage: 3.7V
* Nominal Capacity: 5200mAh
Details
Product Description
1. High Drain IMR26650 Li-ion Rechargeable Battery UH2652
2. Typical Capacity: 5200mAh
3. Normal Voltage: 3.7V
4. Charge Method: constant voltage with limited current
5. Standard Charge Current: 2000mA
6. Fast Charge Current: 4000mA
7. Max. Plus Discharge Current: 30.0A
8. Max. Constant Discharge Current: 15.0A
9. 4.2 volts full charged voltage and 2.5 volts discharge cut off voltage
10. High charge and discharge currents reduce cycle life.
>
>
KP IMR 26650 battery 5200mAh max 30A discharge 3.7v lithium battery UH2652
* Quality High Drain Li-ion Rechargeable Battery
* Brand: KeepPower
* Size: 26650
* Part No.: UH2652
* Nominal Voltage: 3.7V
* Nominal Capacity: 5200mAh
Details
Product Description
1. High Drain IMR26650 Li-ion Rechargeable Battery UH2652
2. Typical Capacity: 5200mAh
3. Normal Voltage: 3.7V
4. Charge Method: constant voltage with limited current
5. Standard Charge Current: 2000mA
6. Fast Charge Current: 4000mA
7. Max. Plus Discharge Current: 30.0A
8. Max. Constant Discharge Current: 15.0A
9. 4.2 volts full charged voltage and 2.5 volts discharge cut off voltage
10. High charge and discharge currents reduce cycle life.
>
>
magudaman said:
okashira
10 kW
ScooterMan101 said:
Those are likely garbage... at least when compared to a real EV quality cell like a Samsung, LG or Panasonic NCA cell. Pretty sure keeppower does not even manufacture cells.
okashira
10 kW
I did start a cycle test on a Model S cell. I was going to do 4A discharge, but decided to run the same regimen as the 30Q for fun. A bit unfair to the Model S cell, as it will heat up much more then the 30Q obviously. 7A is pretty fast. Posting data soon...
The 30Q is losing a bit of capacity ...
The 30Q is losing a bit of capacity ...
okashira
10 kW
The model S cell is holding its own for a 2 year old used cell. It basically puts out exactly the same Wh as the 30Q at 7A. But I don't think it's getting as much of a charge as the 30Q so it could be SLIGHTLY better.
I randomly selected the Model S cell, no cherries...
Average efficiency Model S cell 87% - this would be better with a slower charge, I am doing a light supercharge... ~3% - ~96% charge in 1:57 hr. (yes 2 hr is a fast charge for a high capacity li-ion)
Average efficiency 30Q 91%
Both show 100% columbic efficiency, so I'm not going to bother reporting that anymore. Just did that to prove that there is no Peukert effect and no "loss of electrons," like lead acid (Someone needs to fix Wikipedia...)
Change the chart to show 66% of initial capacity at the bottom of the graph, so the trend can be seen more easily. I think 66% or 2/3 capacity is a good failure point.
You can see the Model S cell is very sensitive to temperature... more then the 30Q. Better efficiency and better Ah and Wh output when it's warmer (during the day)
I randomly selected the Model S cell, no cherries...
Average efficiency Model S cell 87% - this would be better with a slower charge, I am doing a light supercharge... ~3% - ~96% charge in 1:57 hr. (yes 2 hr is a fast charge for a high capacity li-ion)
Average efficiency 30Q 91%
Both show 100% columbic efficiency, so I'm not going to bother reporting that anymore. Just did that to prove that there is no Peukert effect and no "loss of electrons," like lead acid (Someone needs to fix Wikipedia...)
Change the chart to show 66% of initial capacity at the bottom of the graph, so the trend can be seen more easily. I think 66% or 2/3 capacity is a good failure point.
You can see the Model S cell is very sensitive to temperature... more then the 30Q. Better efficiency and better Ah and Wh output when it's warmer (during the day)
Attachments
okashira
10 kW
okashira
10 kW
temperature ... the magnitude surprised me too.
it affects the 30q too ,but the model s cell is affected like 150% more.
model s cell is clearly designed for higher temp. makes sense given the car can heat the cells.
it affects the 30q too ,but the model s cell is affected like 150% more.
model s cell is clearly designed for higher temp. makes sense given the car can heat the cells.
okashira
10 kW
So the 7A discharge on the Model S cell 93% DOD is beating it down...
So I setup a new one with a 4.12V 200ma cut charge and 4A discharge until 2.8V. (fast 1.5 hr charge)
I also changed the 7A Model S cell reginim a little bit. Now, I stop the 7A at 3.0V, and reduce current to 2A until 2.8V. This should keep it from getting as hot and I want to see if it helps.
The new model s one is the green and yellow. This one I did not break in first. So, it's not surprising its picking up capacity on the first few.
Here is what the fast charge on the Model S cell that I am using looks like:
So I setup a new one with a 4.12V 200ma cut charge and 4A discharge until 2.8V. (fast 1.5 hr charge)
I also changed the 7A Model S cell reginim a little bit. Now, I stop the 7A at 3.0V, and reduce current to 2A until 2.8V. This should keep it from getting as hot and I want to see if it helps.
The new model s one is the green and yellow. This one I did not break in first. So, it's not surprising its picking up capacity on the first few.
Here is what the fast charge on the Model S cell that I am using looks like:
Attachments
okashira
10 kW
Ok, so it turns out that there is a little more to supercharging then originally thought.
The Model S cell was still taking a beating on the 7A run after adding in a drop to 2A at 3.0V.
So I set another one up and used a new slower charge routine to ensure it wasn't being charged to fast:
charge 0.3A for 1 minute
charge 1.8A until 3.77V with 0.8A cut
charge 0.8A until 4.15V with 0.12A cut
wait
Discharge 7A until 3.0V
Discharge 2A until 2.8V
wait
cycle
The Model S cell is doing MUCH better this time ... so.... I was charging it too fast (or the low current precharge helps)
It is certaintly the fast charge regimin that I was using in the prior post that was killing it, NOT the discharge rate.
I also reduced the charge rate of the 30Q at cycle 114, didn't seen to have much effect.
The Model S cell was still taking a beating on the 7A run after adding in a drop to 2A at 3.0V.
So I set another one up and used a new slower charge routine to ensure it wasn't being charged to fast:
charge 0.3A for 1 minute
charge 1.8A until 3.77V with 0.8A cut
charge 0.8A until 4.15V with 0.12A cut
wait
Discharge 7A until 3.0V
Discharge 2A until 2.8V
wait
cycle
The Model S cell is doing MUCH better this time ... so.... I was charging it too fast (or the low current precharge helps)
It is certaintly the fast charge regimin that I was using in the prior post that was killing it, NOT the discharge rate.
I also reduced the charge rate of the 30Q at cycle 114, didn't seen to have much effect.
Attachments
Extremely interesting data! Thnx allot for doing these tests!
Just goes to show how much the charge&discharge routine can change the life cycle of cells, would be very cool to continue these tests for more cells to get a indication of a optimal charge routine for these cells.
I have seen calculations that Tesla charges these cells around 4A max around the 50% charged point and lower after and before that. As far as we can see here they are destroying there cells with that. Or would ramping up the current during the charge really make such a difference?
Both cells seem to have some massive drops from time to time, maybe error in the data or some type of massive voltage drops during discharge that makes your machines think they are empty?
a bit of topic question but would a high drain cell have better round trip efficiency then a high cap cell because of the lower internal resistance?
I ask because at my school we are doing a solar race boat challenge and we are mainly focusing on wh/kg but maybe the increased cycle efficiency is more important as we can only use power generated by the solar cells. Therefore should lose less energy in total and make up for the increased weight that way.
Just goes to show how much the charge&discharge routine can change the life cycle of cells, would be very cool to continue these tests for more cells to get a indication of a optimal charge routine for these cells.
I have seen calculations that Tesla charges these cells around 4A max around the 50% charged point and lower after and before that. As far as we can see here they are destroying there cells with that. Or would ramping up the current during the charge really make such a difference?
Both cells seem to have some massive drops from time to time, maybe error in the data or some type of massive voltage drops during discharge that makes your machines think they are empty?
a bit of topic question but would a high drain cell have better round trip efficiency then a high cap cell because of the lower internal resistance?
I ask because at my school we are doing a solar race boat challenge and we are mainly focusing on wh/kg but maybe the increased cycle efficiency is more important as we can only use power generated by the solar cells. Therefore should lose less energy in total and make up for the increased weight that way.
Malloot said:
Only if you use them with same current levels. But you should use lower C rate cells with lower currents, in that case you are not loosing much energy. So, pick your batteries properly according to your needs, and you will gain in energy density (lower C rate cells = larger energy density).
okashira
10 kW
Malloot said:
Optimal charge would be anything below the charge rate that may damage the cells. :-D
Slower is always safer.
The massive drops are just glitches with the tester. Sometimes it stops a discharge early, or stops a charge early. Power outages, messing with settings, etc. I will delete those so no one is confused.
Yes in the same conditions, the power dense cells will offer more efficiency. But you need to look at total system efficiency. If using power dense cells causes you to have less Wh, or to add weight, you will likely negate that benefit.
Tesla uses about 4.4A max at 3% SOC, down to 4A and tapers quickly after 35% SOC. At 50% SOC maybe 2.5A or so.
Btw, I finally dead shorted out two Model S cells last night. Took video.
Brake
100 W
Me too!
okashira
10 kW
Updated.. yes the Model S is doing pretty damn good with the slower charge. Good considering the relativity heavy 7A discharge.
Yes I am still lowing the current to 2A at 3V, but it's at 7A for most of the run. Typically about 2.4Ah at 7A then 0.4Ah at 2A.
Ill also point out it's about ~92% DOD for the Model S cell and ~94% DOD for the 30Q.
ill try to get the model s video up tomorrow, hint, there is a burst disc under the positive terminal that pops, spews out hot electrolyte, but no explosion or fire of any kind.
Attachments
okashira
10 kW
https://www.youtube.com/watch?v=zwpkK8KdQIQ
Model S cell shorted
Model S cell shorted
Ohbse
10 kW
Thanks for the awesome data!
Nice video too, about what I expected. Must have been pulling some pretty decent amps to get the nickel glowing.
Nice video too, about what I expected. Must have been pulling some pretty decent amps to get the nickel glowing.
okashira said:
Updated.. yes the Model S is doing pretty damn good with the slower charge. Good considering the relativity heavy 7A discharge.
Yes I am still lowing the current to 2A at 3V, but it's at 7A for most of the run. Typically about 2.4Ah at 7A then 0.4Ah at 2A.
Ill also point out it's about ~92% DOD for the Model S cell and ~94% DOD for the 30Q.
ill try to get the model s video up tomorrow, hint, there is a burst disc under the positive terminal that pops, spews out hot electrolyte, but no explosion or fire of any kind.
You are doing holy work here, okashira!
Where did you get your cell tester from, if you don't mind me asking. Do they also make full pack testers, something variable voltage like 24-48V or even 72V?
