Triton data on LiIon cells - updated ...

Candle Power Flashlight Forum

Help Support CPF:

wquiles

wquiles

Flashaholic
CPF Supporter
Joined
Jan 10, 2005
Messages
8,516
Location
Texas, USA, Earth
Thanks to Tom and others I realized that my initial data was tinted/tainted due to using the "high-tech" home made battery holder so I ended up doing all of the data again. Since the initial data was bad, I decided to edit this post removing all of that data and replacing it with good, reliable data instead.

For all of these I am using a new Triton charger with the latest software (green sticker on the bottom). For battery holder, I am using a modified DSD battery charger which still works for charging cells.

Note that I am just trying to find the cells capacity under what I consider are realistic loads - I am not trying to verify the cell's "rated" capacity.

Triton_Data_1.jpg


Some things to note:
- All 18650 cells (LG, Dell, Old) are un-protected cells. Use these at your own risk.
- The Pila 168S cells are doing OK, but not expectacular. They clearly are no match for even used 2400mAH cells /ubbthreads/images/graemlins/wink.gif
- Note how little the initial voltage under load sags when 1A, 1.5A, or 2A loads are put on the cell.
- The Pila charger is not bad. Even on a slow charge with the DSD, I did not get much better capacity. Might as well continue to use the Pila charger with these cells. The 1000? means that it must be a charge of at least 1000mA - maybe higher - I have no idea.
- The "cheap" DSD charger so far gives me the highest capacity, even if it is very slow while doing so. I have tried a few charging levels on the Triton, but I have not found a "sweet" spot yet between charge vs. time. Part of this is also due to the final charge voltage from the DSD charger being a little bit higher (between 0.05 and 0.10 volts) than cells charged on the Triton.


EDIT on 7/22/05 - added more data
EDIT on 7/24/05 - corrected voltage under load
EDIT on 7/27/05 - noted that holder used affected readings
EDIT on 8/6/05 - complete re-write of post to reflect new/correct data

Will
 
Re: Triton data on 18650 cells ...

RESERVED - I will try to put this data in a table so that it is easier to read.

Charger setup. Note the fan and the bleading-edge, high-tech holder I am using for now!:

EDIT: This "holder" caused many errors in the measurements above. I am now using the same DSD charger as a plain holder and getting much better results!

IMG_0863.JPG


The grey cels are the LG's, the red ones are the old laptop batteries from a Dell pack. I have another Dell pack of newer, Sony cells (not shown here yet, that seem to have even higher capacity - 2500mA ???) which I will post soon:
IMG_0864.JPG


here is the DSD charger in action:
IMG_0865.JPG


Here is a couple of pictures so that you guys get a better idea of the size of these cells compared to some common lights:
(from left: M4, U2, VB-16, 3 different 18650 cells, Pila 168S, G2, 2xAA MiniM*g)
IMG_0866S.JPG

Link to 1024 x 768 ...

Here is a close-up of the smaller lights shown above:
IMG_0867S.JPG

Link to 1024 x 768 ...

Will
 
Re: Triton data on 18650 cells ...

Will, try charging the RCR123 cells at a lower rate, and see if they load up more. Your success with cell #1 (18650)was achieved charging at about 1/4C, and charge with RCR123 was about 1C. I was wondering if charging 18650's with Triton, then finishing off in DSD would make a difference in capacity. It might not, and if so then we really do not need to take our 18650's up to 4.2 volts, and life of cells would be inhanced?

Bill
 
Re: Triton data on 18650 cells ...

Bill,

I followed your advice and charged the RC123 cells on the DSD charger. Unfortunately, at least so far for RC123 cell #1, the slow charge and slightly higher charge voltage made no difference. I am currently discharging RC123 cell #2 and will post later tonight. I am still discharging at 0.5A for the RC123 since I feel that is a very reasonable load they would see in real lights (same reasoning of why I picked 1.0A for the 18650 cells).

As to the 18650 cells, yes, so far it makes a significant difference to charge them slowly, but at least so far also to charge them at the slightly higher voltage with the DSD charger - it seems to give most of the cells the best charge so far.

Like you stated, it looks like a slightly faster charge on the Triton and finish on the DSD might be the best combo. I will run some of these test over the weekend - too bad the end voltage on the Triton is not adjustable /ubbthreads/images/graemlins/frown.gif

Will
 
Re: Triton data on 18650 cells ...

VERY good observations Will. I am going to end up purchasing a Triton sooner or later for some multi-cell packs for some hotwire lights. I have a cheap little charger that i got from 4sevens back when the R123's FIRST hit the market. He moded it to stop charging at 4.2v, It takes a few hours to charge a cell so i imagine it has a slow recharge as well.

Ive got a Dell laptop battery that doesnt hold a charge anymore. Maybe i should disassemble it to see if possably there is just 1 bad or reversed cell in there ruining the whole thing. Or in the very least, i might find some half decent cells in there still.

Shane
 
Re: Triton data on 18650 cells ...

Great stuff Will!

Thanks so much for doing all these measurements.

I'm surprised the Pila168S registered so low. Any idea what was going on with that?

Wilkey
 
Re: Triton data on 18650 cells ...

Hello Will,

Welcome to the wonderful world of battery/charger testing... /ubbthreads/images/graemlins/grin.gif

I agree with Wilkey, your results seem to be a bit low.

I see the LG cells came up when they were "topped off" on the other charger, perhaps you should try charging the Pila's on the DSD charger as well. /ubbthreads/images/graemlins/smile.gif I have noticed that my Pila cells will start charging again if I let them sit in the charger for a little while after the main charge. The green light will go off and on a couple of times, then stay green. This indicates to me that when the green light first comes on, the cells are not quite fully charged. The Pila 168A cells should test out around 2000 mAh, or slightly more. Edit: I can't read very well, we were talking about the Pila 168S cells not the 168A cells. The 168S cells should come in at around 1300 mAh, or slightly more.

LG claims a minimum capacity of 2380 mAh for their cells. To get this, they charge at 1.2 amps to 4.2 volts and terminate the charge when the current drops below 50 mA. Then you discharge at 0.48 amps to 3.0 volts (at a temperature of 23C). You may want to try to duplicate this test for reference.

The other comment that I find a bit revealing refers to the voltage dropping to 3.4 volts. Li-Ion cells are usually rated at a midpoint voltage of 3.6 (or 3.7) volts at their maximum continuous designed draw. The maximum rated current for the LG cells is 4.8 amps, so I find it interesting that they are dropping below 3.5 volts early on. I suppose this could also be an indication of the state of charge...

Warning!!!

Be advised that battery/charger testing is similar to all other projects...

A friend of mine has a sign prominently displayed on his shop wall where we work on our old cars together. It reads:

"PROJECTS AE BORN PREGNANT !!!"

Just when you think you have something figured out, a whole new batch of questions arise... and it seems like you always end up breaking a few other things when you try to repair something.

Tom

Edit: Wrong Pila cell comparison.
 
Re: Triton data on 18650 cells ...

Tom,

As you being one of the (if not the most) experienced with batteries here, I really appretiate your insights /ubbthreads/images/graemlins/bowdown.gif

As you are, I am really surpriced at the voltage under load as well on all of these 18650 cells - I expected higher. I need to investigate further what is going on.

I will definitely try charging the PILA's on the DSD charger - as we know the PILA charger is more-or-less a fast charger, so a slower charge should be better as well.

I should note that my intent is not to duplicate nor verify mAH ratings, but to rather see what cells behave like in likely scenarios. Although the LG's would/should come up to 2380 or higher, a .48A draw is not what I consider realistic. That's is why I am using the more realistic 1Amp draw, which is what I draw from my PILA 168S in my lights. I do expect that of course I will not see the highest values, but the Triton does an excellent job of being able to repeatedly draw a constant current just as a constant regulator in a light would /ubbthreads/images/graemlins/smile.gif

I am also trying to "quantify" if the slow/slightly higher voltage from the "cheap" DSD charger is worth something. So far it apears that to be the case, but as you noted, this small project seems to have a life of its own now /ubbthreads/images/graemlins/wink.gif

Will
 
Re: Triton data on 18650 cells ...

[ QUOTE ]
Ginseng said:
Great stuff Will!

Thanks so much for doing all these measurements.

I'm surprised the Pila168S registered so low. Any idea what was going on with that?

Wilkey

[/ QUOTE ]
I had ten Pila 168S cells and eight had a large capacity loss just like what's been posted here with only a few cycles on them. One was down to only 50mAh when tested on a CBA-II!
 
Re: Triton data on 18650 cells ...

Hello Bill,

I just realized that I had the wrong Pila cell comparison... I have edited my post to reflect the Pila 168S cells.

At any rate, did you ever find out what was going on with your Pila cells? 50 mAh of capacity is "less than stellar..."

Tom
 
Re: Triton data on 18650 cells ...

[ QUOTE ]
SilverFox said:
Hello Bill,

I just realized that I had the wrong Pila cell comparison... I have edited my post to reflect the Pila 168S cells.

At any rate, did you ever find out what was going on with your Pila cells? 50 mAh of capacity is "less than stellar..."

Tom

[/ QUOTE ]
Tom:

Nope! I didn't have anymore Pila to punish but I've tried my darnest to kill RCR123 cells and couldn't! /ubbthreads/images/graemlins/confused.gif

I recently got a new Motorola Razr V3 cellphone power by a Li-Ion cell. You should see how misinformed about batteries and charging that community is. I've found websites selling generic and Motorola cigarette lighter adapters, advertizing them as being able to charge Ni-CD, Ni-MH and Li-Ion automatically! Never seen a charger yet where the wasn't a seperate switch as there is no way to tell what type of battery is conected to it. Can you imagine the possible outcome of putting over 4.0V on a Ni-CD or Ni-MH cell.

Well, the good thing is that the connector isn't standard, so you'd have to cut it off and adapt.

I bought a desktop charger(alone), where use use the DC(5V)adapter with it. I just want the holder.
 
Re: Triton data on 18650 cells ...

[ QUOTE ]
SilverFox said:
The other comment that I find a bit revealing refers to the voltage dropping to 3.4 volts. Li-Ion cells are usually rated at a midpoint voltage of 3.6 (or 3.7) volts at their maximum continuous designed draw. The maximum rated current for the LG cells is 4.8 amps, so I find it interesting that they are dropping below 3.5 volts early on. I suppose this could also be an indication of the state of charge...


[/ QUOTE ]

Tom,

I decided not to trust at face value the voltage shown on the Triton while discharging used my Fluke 87 to verify what is going on. I found out that this must be an internal volgate load not the actual cell's voltage. When re-testing the two PILA 168S's (on slow charge) I found out that the actual, initial voltage at the cell was more like 3.7-3.8 volts, which is more in-line with what we expected /ubbthreads/images/graemlins/smile.gif

I also checked the votage at the end of the discharge, and the 3.0 volts the triton claims is also the internal load voltage. When the cells end the charge, the actual loaded voltage is 3.3 volts at the cell.

I updated the first post to correct this error.

Will
 
Re: Triton data on 18650 cells ...

Will - the voltage differences between what the Triton shows and your Fluke 87 (measured at the cell itself, I presume) are probably related to the resistance of the battery holder/wiring circuit. This difference will become more pronounced at high current charging/discharging, and small at low currents.
 
Re: Triton data on 18650 cells ...

wasBlinded,

I have been working on a new post for tonight or tomorrow, but since you asked, YES, I tend to agree with your assesment. In fact, I stopped the testing on Sunday until I can re-wire the holder into something better. And I did /ubbthreads/images/graemlins/grin.gif

Just last night I took the DSD charger and "modify" it to become my new 18650 cell holder. I put a small switch to disconnect the "+" charging wire (inside the unit, which was not too hard to open) so that I have no way of damaging the internal charing circuit when using the holder as a charger. Now my DSD is both an active charger or a passive holder with the flip of a switch /ubbthreads/images/graemlins/wink.gif

At any rate, I have these fairly new/recent Sony 18650 cells that I got from a damaged Dell pack at work, and I am so far getting GREAT capacity from these cells (I would bet that these are the new 2500mA cells from Sony). After an al-night charge on the DSD charger, even at the relatively "high" 1Amp draw, the cell gave me almost 2200mA !!!. Best of all, these cells were very low cost to me (free) - how cool is that /ubbthreads/images/graemlins/happy14.gif

Not only that, but using this setup I checked the voltage at the cell before the discharge cycle (4.18 volts) and when the discharge stated, it only dropped to 4.06 volts !!!. It is early, but I feel MUCH better now with this setup so far.

I will be posting new data as I re-test things, which of course will take some time. Something about projects being born "pregnant", right Tom? /ubbthreads/images/graemlins/tongue.gif

Will
 
Re: Triton data on 18650 cells ...

I've been using my DSD charger as a cell holder too, but did without the switch, simply wiring the ICE charger in parallel with the existing circuit, with no ill effects that I've seen. Since the existing circuit has to be able to withstand the voltage of a fully charged cell sitting in the holder, I figured applying the charging voltage (max of 4.2 volts) to the same contacts would be no different.

I do get different numbers than you do using the LG 2400 mAh cells. All 5 of mine give better than 2200 mAh in a discharge test with the Duratrax ICE, though none actually hit 2400 mAh (at 1.5 amps). Again, maybe the resistance of your cell holder and circuit has something to do with it.
 
Re: Triton data on 18650 cells ...

Well, I am back home, and so far I have much better data - certainly more credible now:

Charged on DSD, discharged on Triton at 1Amp:
- Dell(Sony) cell #1 2180mAH
- Dell(Sony) cell #2 2226mAH

Voltage at rest were 4.18 each, once loaded was 4.05 and 4.06 respectively. Given now much I got at the 1A drain, these "must" be the new 2500mAH cells /ubbthreads/images/graemlins/smile.gif

I just started re-testing the LG#1 and rested was 4.21, while loaded was 4.08. I should know the actual capacity in a couple of hours, but I expect it to be even higher now /ubbthreads/images/graemlins/grin.gif

In addition, I wrote down the voltage loaded shown by the Triton:
Dell #1 3.96
Dell #2 3.97
LG #1 4.00

So now I only have about 6-8mV difference between the triton and the Voltmeter - that is also much better /ubbthreads/images/graemlins/happy14.gif

Will
 
Re: Triton data on 18650 cells ...

Thank you Tom /ubbthreads/images/graemlins/blush.gif

I feel much better now that I have good data to share. In fact, that LG#1 gave me slightly more than 2300mAH /ubbthreads/images/graemlins/smile.gif

I am re-testing LG cell #2 today, and will soon re-test the PILA 168S's next as I want to see what their "real" capacity is now that my setup is working properly /ubbthreads/images/graemlins/wink.gif

Will
 
Re: Triton data on 18650 cells - updated ...

Quick summary of the tested cells so far. Slow charged on the DSD charger, and discharged at 1A on the Triton, using the dissabled DSD charger as battery holder:

These are all unprotected 18650 cells, voltages at the cell with voltmeter:
cell / Vnoload / Vload / mAH
Dell #1 / 4.18 / 4.05 / 2180
Dell #2 / 4.18 / 4/06 / 2226
LG #1 / 4.21 / 4.08 / 2311
LG #2 / 4.21 / 4.08 / 2321
LG #3 / 4/16 / 4.04 / 2212 (this cell was rested a little longer)
LG #4 / 4.20 / 4.08 / 2304
LG #5 / 4.20 / 4.08 / 2311
old 1 / 4.20 / 4.10 / 2126
old 2 / 4.19 / 4.10 / 2084

I also re-tested the two PILA168S after being charged in the PILA charger. I will do a slow charge to compare soon:
PILA 1 / 4.17 / 4.02 / 1240
PILA 2 / 4.21 / 4.06 / 1332

I must say that I am very impressed with these 18650 cells. Not only their capacity at a "high" 1A draw is excellent, but look at the initial voltage under load - true, it drops from there, but this is probably why they can provide such high currents even under load - the voltage holds pretty well /ubbthreads/images/graemlins/wink.gif

Will
 
You must log in or register to reply here.

Latest posts

Back
Top