Li-Ion Battery Shoot Out

[TheIlluminator]

I have done some more testing and I think that I have figured out how the DSD charger charges your batteries to such a high voltage. As I have mentioned before the DSD charger is a single stage charger that cuts off charging at a specific voltage. Charging in this manner results in a cell that is 60%-70% charged. To give a more fully charged cell the DSD charger applies a voltage higher than 4.2v rather than regulating it at 4.2v. The cutoff voltage for 2 cells that I charged were 4.280v (mp rcr123) and 4.287v (awrcr123). The recommended charging voltage for a li-ion is 4.2v +or- .05v. The DSD charger does exceed the recommended voltage for a li-ion battery, but not by much. Having such a high cutoff voltage results in a cell that is more completely charged. The voltage of my cells after resting for 24 hours is 4.140v and 4.110v.

 

[John N]

Any plans to run the following cells:

- LG 18650, unprotected, 2600mAH
- MP 17670, unprotected, 1500mAh (AW)
- GMB 18650, unprotected, 2200mAh (AW)

It might also be interesting to be able to compare best of breed unprotected 17670 (1500mAh AFAIK) vs a protected 17650 since these are sized for the same application.

I noticed those 2600 mah 18650s are in stock now....



-john

 

[LumenHound]

It's been announced in this thread that some testing and runtime graphs for the LG 18650 2600mAh cells is in the works.

 

[4sevens]

silverF, The graph of the BS cells... is that the new 900mah protected rcr123's?

 

[SilverFox]

Hello 4sevens,

Codeman sent me the BS cells for testing. You will have to check with him to make sure exactly what cells they are.

Tom

 

[SilverFox]

It has come to my attention (thanks Chevrofreak and Flea Bag) that the graphs of the R-CR123 tests are difficult to compare because of the difference in Y axis ranges on the graphs. Looking at the graphs without paying close attention to the voltage scale, it would appear that the AW cells fall off drastically, compared to the Powerizer cells.

The Y axis of the graphs is selected to zoom in as much as possible on the data, yet still allow the legend to be read without overlapping the graph lines. As more tests are combined in a single graph, the legend size grows. To accommodate the larger legend size, the Y axis is extended.

I apologize for any confusion resulting from this.

I have added a comparison graph of the R-CR123 cells tested utilizing the same Y axis scale. I hope this makes it easier to interpret the results.

Tom

 

[chevrofreak]

SilverFox, thank you for that composite image, it makes it quite a bit easier for direct comparison.

 

[cmaylodm]

Are the AW RCR123s in the test the old blue label or new black version?

 

[SilverFox]

Hello Cmaylodm,

Blue label.

Tom

 

[chevrofreak]

I'd be quite interested to see tests of the new black label "high current" cells that AW is selling. They did really very well in the Fenix P1D CE on high.

 

[mdocod]

would be great to see some tests of the *new* AW high current RCR123s... maybe AW would be willing to send ya one for testing, hehe.. My experience with them isn't anything like the graph of the old cells.

 

[SilverFox]

Update:

Icantsee sent me some of the Molicell Emoli IMR26700 cells to check out. The results are posted in the first post.

These cells are Lithium Manganese Oxide, which means that they are safer than the normal Lithium Cobalt cells. They hold up under load very well.

Tom

 

[mdocod]

wow, those cells are incredible. 20 amps only lost about 20% capacity(WH) compared to 1 amp. seems to hold AH up and loose V instead.

 

[icantsee]

Thanks again for taking time to do the testing.

John

 

[SilverFox]

Hello John,

I didn't try a 15 minute charge... I figured I would let you work on that one.

I do know that if you run the cells hot off the charger, you get even better performance.

I would also advise you to pay attention to your charger. The Schulze had a little bit of a problem charging these cells. They have an impedance of around 0.019 ohms, however, at full charge and a low currents the internal resistance more than doubles. This threw my Schulze off and it was terminating the charge at just over 4.1 volts. The Schulze is very picky about internal resistance.

I ended up parallel charging them with my power supply. The charge time at 0.5C was around 3 hours, and the cells ended up at 4.200 volts.

Tom

 

[jayflash]

Yet again, thank you, Tom, for your unending efforts. You've advanced our knowledge, continuously, with useful info for the CPF rank & file masses. I'll bet that almost all of us have benefited from your help.

 

[mdocod]

I'll bet that almost all of us have benefited from your help.

+1 to that!

I find myself referencing your charts constantly. invaluable. Thank You for all the hard work and words of wisdom!

 

[icantsee]

Tom,
My charger is an FMA Balance-Pro. I've been setting it for a .7c charge rate based on 2900ah. Should I be using .5c? My charger terminates at 4.2v. Some of the guys on the RC forums are talking about routinely using 3c charge rates and supposedly getting hundreds of cycles. Do you feel .7c is reasonable?
John

 

[cy]

very much worth revisiting!!!

 

[LuxLuthor]

Update:

Icantsee sent me some of the Molicell Emoli IMR26700 cells to check out. The results are posted in the first post.

These cells are Lithium Manganese Oxide, which means that they are safer than the normal Lithium Cobalt cells. They hold up under load very well.

Tom

WOW on those cells....and yeah, great thread to revisit. Thank you Tom!