123 Battery Shoot Out

SilverFox

Flashaholic
Joined
Jan 19, 2003
Messages
12,449
Location
Bellingham WA
123 Battery Shoot Out

Here is another series of tests, this time looking at Lithium CR123 primary cells. These are not rechargeable.

I picked a variety of batteries. My list is far from exhaustive, however I think this is a good representation. If anyone has a brand that I have missed and wants to donate them to the cause, get in touch with me. I'll test them and post the results.

Battery Capacities:
I am not sure at what load Lithium battery capacities are rated at. I know that many of our lights draw 1 to 2 amps, so I decided to test at those rates. Energizer rates their 123 cells at 1.5 Ah for reference.

Graph Nomenclature:
Ah = Amp Hours
Wh = Watt Hours
M = Minutes

The battery line up:
Duracell Ultra
Energizer
Sanyo
SureFire
Streamlight
BatteryStation
12/07 BatteryStation now Made in USA
Maxell Gold
GI
RayOVac
Titanium
Marathon
Golston
4/22/05 New Titanium Cells
8/6/05 New batch of Golston Cells
6/08 Panasonic 1550 mAh Made in USA
3/09 Panasonic 2017 from Europe
3/09 Fuji 2012 from Europe

A big thanks to JohnN and Dano for providing additional batteries for me to test - Thanks.
Another big thanks to Wayne at AmondoTech for providing additional batteries for me to include in the testing - Thanks.
Another big thanks to Chevrofreak for providing additional batteries for me to include in the testing - Thanks.
Another big thanks to Dae for providing some additional Golston cells for testing - Thanks.
Thanks to VWTim for picking up some Everlast Go Power cells for testing.
Thanks to Cdosrun for sending the Panasonic and Fuji cells from Europe.


The graphs:

Edit: 0.5 Amp Rate
123Comparison05ATable.gif


123 at 1.0 Amp Rate
123Comparison1ATable.gif


123 at 2.0 Amp Rate
123Comparison2ATable.gif


Test conditions:
All tests were conducted at room temperature of about 68 degrees Fahrenheit utilizing a West Mountain Radio CBA.
All batteries are fresh stock.

Test Notes:
The tests at the 1.0 Amp Rate revealed a moderate battery temperature increase. The batteries ended up in the 85-88 degree F. range. The battery temperatures shot up during the 2.0 Amp Rate testing. I have my CBA set to terminate the test when the battery temperature exceeds 115 degrees F. The SureFire battery test was terminated early because the cell temperature exceeded 115 degrees. I watched the cell temperature continue to rise, after the test had been terminated, until it leveled out at 122 degrees. The SureFire cells were the only ones that exceeded 115 degrees, but all of the others were very close.

Conclusions:
These tests revealed some differences at higher current draws, but at lower levels, the differences seem minimal.

The Energizer cells were at the bottom of the pack. These were fresh cells and I am not sure what is going on with them. I would normally do additional tests on them, but I need to find a more reasonable source for them than the local store.

The eye opener for me was the temperature increase of the batteries when testing at 2 amps. I would venture an opinion that if a light draws 2 amps or more, thermal management is going to be difficult during long runs. The lamps and LED's are producing heat, and at these rates we have the batteries producing heat as well. Lithium cells have a thermal shut down protection system, and this may shed some light on why sudden shutdown occurs with some lights.

4/22/05 Wayne at AmondoTech was disappointed in the performance of his Titanium 123 cells and decided to do something about it. He met with the manufacturer and had them change the formulation to provide improved performance. He was successful. These new cells may not be the best at 0.5 amps, but lead the pack at 2.0 amps without heating problems.

Here is another graph of these cells. I am very impressed that Wayne made it a point to come up with a better product and delivered on this.
TitaniumNew123atVariousRates.jpg


Dae sent me some Golston cells for testing. These cells performed better than the ones that Chevrofreak sent me, but there is still some inconsistencies. Here is the test results on these cells. The erratic results from the 1.0 amp test was caused by the cell, not the test equipment.
GolstonNewatVariousRates.gif


People have been asking about CR2 cells. I don't use them, but happened to come across some Titanium CR2's, thanks to Wayne at AmondoTech. I believe these are rated at 800 mAh.

Here is the data.

TitaniumCR2atVariousRates.gif


You may wonder why I did a second run on the cell that had previously been tested at 0.75 amps. HarryN is developing a CR2 light as are some others. These lights can operate at different levels. It just happens that Harry is thinking of a turbo mode at 700 mA, and after running the test at 0.75 amps, I noticed that the voltage had rebounded back to 2.8 volts. It struck me that it would be interesting to see what happens if you do another run at a lower draw. This means that you can run the light in turbo mode until the batteries die, wait 20 minutes or so, then still have usable light at a lower level. In this case the turbo mode lasted a little over 55 minutes, after which you still had over 96 minutes of run time at 0.10 amps.

My combined graphs are getting cluttered and are difficult to read. I will add future cell results with individual graphs and you can compare the Ah Wh and M values to compare.

Here are some Everlast Go Power cells that VWTim picked up for me to check out. At $0.99 each, they seem to be solid performers.

EverlastGoPower123atVariousRates.gif



It seems that Sanyo has some CR123 cells that are being made in China. I have been sent some of these cells for testing and it appears that they do not perform as well as the ones made in Japan.

Here are some graphs to compare...

Sanyo Made in Japan

Sanyo123atVariousRates.gif



Sanyo Made in China

SanyoChina123atVariousRates.gif



Vew sent me some Tenergy CR123 cells to test. Thanks.

I checked them on the ZTS tester and got mixed results. Out of 10 cells, 8 showed 100%, 1 was at 80% and 1 was at 60%. These results are from cells that tested at 100% on the ZTS tester. The cells seemed to perform quite well.

Tenergy123atVariousRates.gif


Kevin at BatteryStation sent me a box of his new Made in USA CR123 cells. Not only are then very strong performers, but they are consistent from cell to cell. The cells tested at 2 and 2.5 amps got over 140 F right at the end of the test, but I figure most people will not use a cell down to 1.0 volts in a continuous run.

Here is the graph.

BatteryStation2007CR123atVariousRat.gif


David at the Fenix-Store sent me some of the Panasonic CR123 cells that he sells on his site. These are rated at 1550 mAh which is higher than most other brands of cells. These cells did very well at 0.5 and 1.0 amps, however they started to drop off at 2.0 amps.

Here is the graph.

PanasonicCR123atVariousRates.gif


Cdosrun sent me some Panasonic and Fuji cells from Europe to check out. I have been trying to get this testing done and the data put up here for some time now, and things just have been very busy. Finally, the testing is done and here is the data.

The Panasonic cells had a 2017 date on them and were strong performers. They even performed well at a 2.5 amp draw, but since I only had one cell to test at that rate, I didn't add a graph for it.

Here is the Panasonic 2017 graph.

PanasonicCR1232017atVariousRates.gif


The Fuji cells had a 2012 date on them. They did well at lower discharge rates, but exceeded the maximum temperature (140 F) during the 2 and 2.5 amp discharges. The discharge curve at 2 amps also has a little dip in it suggesting a little higher internal resistance.

Here is the Fuji 2012 graph.

Fuji1232012atVariousRates.gif


12/2006 This round of testing is to check out the ability of the CR123 cell to handle a constant 2.5 amp draw. There are several high output lamp assemblies that draw close to that. If you do a run time test on a SureFire M6, the cells will see this load.

LED61 sent me a bunch of cells for this testing. Thanks Alberto. I also added in some other brands to add to the test data.

Four cells of each brand were tested. Each cell was checked on the ZTS tester and found to be at 100%. You will notice that the initial voltage is close to the same for each cell. In multi cell lights, we have advocated using a load test to match the cells to avoid cell mismatch that can cause problems.

Keep in mind that none of the brands recommend a constant current draw this high. Some are OK with a heavy pulse load, but none of them are rated for a constant heavy current draw.

There is some lag in my CBA temperature probe, so I usually set it at 120 F. This test required raising that to 140 F, and even then some of the cells terminated on high temperature. All of the cells got hot during the test. This current draw is pushing the limits of the CR123 cells. Any weakness of the cell is exaggerated at this high current draw.

Keep in mind that the testing is done on a single cell that is open to the air. Inside a flashlight with other cells, things are expected to get hotter than what I am seeing. None of the shrink wrap distorted during these tests, but that has been reported by others who have done extended runs with these lights.

I was thinking that the best cells for this high current rate would be those that were consistent from cell to cell. The voltage retention under load is also important since many of these high current draw lights are direct drive. Higher voltage means a brighter light.

I will add graphs as the data gets sorted out. The list of cells, so far, includes Titanium, BatteryStation, Sanyo (made in Japan), Panasonic (made in USA), SureFire, Energizer, Duracell Procell, and Duracell Ultra.

I will start by just showing the results for each of the 4 cells of each brand. I will save any conclusions for later.

The Titanium CR123 cells have had some problems. There was a bad batch that was supposed to have been removed from the market, however, I keep hearing of people ending up with cells that won't pass a ZTS test at 100%. I ordered 12 cells and they all checked out OK on the ZTS tester. Here is the test results at 2.5 amps.

I must add that the battery manufacturer of the Titanium cells does not recommend using their cells at this high a constant current draw. I believe they also caution against using their cells in higher than 3 cell lights.

25ATitanium123.gif



Here is the results from the Sanyo CR123 (Made in Japan) cells. We have found that the Sanyo cells made in China seem to have around 10 - 15% less capacity than those made in Japan, but at this high rate, I wont speculate a guess as to how well the made in China cells will do. I don't have any of the Sanyo cells from China, so I can not do a direct comparison. We will just have to speculate... :)

If anyone has 4 Sanyo CR123 cells they would like to send me, I would be happy to include them in this test. PM me for shipping information.

Cell 1 terminated early due to high temperature. All four cells hit over 140 F, but cell 1 hit it a few seconds before it hit the low voltage cut off.

Sanyo also does not recommend using their cells at this high a constant current draw.

25ASanyoJ123.gif



BatteryStation has been a long time supporter of CPF, and many use their cells. Kevin has worked with the manufacturer to give us good performance and still offers a discount price for the members of CPF. Kevin has gone the extra step of individually testing each cell with the ZTS tester in an effort to add another layer of quality to the cells he sells.

I asked Kevin about testing the BatteryStation cells at 2.5 amps. He does not recommend his BatteryStation cells in applications that draw 2.5 amps continuously. As a matter of fact, Kevin provides a box of 12 SureFire cells free with every SureFire M6 that he sells. He stated that if you are going to continuously use a light that draws that high a current, you should use the batteries recommended by the manufacturer of the light, and let them worry about any liability issues.

Kevin also mentioned that BatteryStation cells are continuing to change. His next batch of cells have all stainless steel construction and, I believe, a new PTC. These are safety related items and will probably not impact performance in any way. Kevin went on to inform me that in the rare event of a "venting with flame" incident, a stainless steel can will not ignite where the aluminum possibly could get hot enough to ignite.

I have also heard a rumor that BatteryStation cells may no longer be made in China in the future... perhaps they may even be manufactured in the US.

Here are the test results of the BatteryStation cells at 2.5 amps. Note that cells 2, 3, and 4 terminated early due to the temperature exceeding 140 F.

25ABatteryStation123.gif



SureFire recommends using their batteries in their lights. I guess this means that the SureFire batteries are manufactured in a way that evens out the tail end of the discharge curve. Many people have questioned SureFire about running the M6 for a constant 19 minutes, and have been repeatedly told that it is OK to do that. Here is the SureFire CR123 data.

25ASureFire123.gif



Here is the data from some Panasonic CR123 cells that were made in the USA. These were individually packed and it looks like I got two separate batches.

25APanasonicUSA123.gif



Here is the data from the Energizer cells. I don't know what happened to cell 3, but it seemed to behave differently.

25AEnergizer123.gif



There has been some discussion that Duracell Pro Cell and Ultra CR123 cells are the same. Here is how they perform under this 2.5 amp load.

First the Pro Cell test data.

25ADuracellProCell123.gif



Now the Ultra cell test data.

25ADuracellUltra123.gif



From this limited test data, I can see why SureFire used the Duracell Ultra cells when they developed the M6. These seem to be the most consistent from cell to cell.

Bill Waites wondered if a different batch of SureFire cells might behave differently under these high loads. The cells I tested had a 2013 date on them and he had some that had a 2016 date on them. He sent me 4 to test. Thanks Bill.

It appears that these are pretty close, but 2 of them got hotter than the other two.

Here is the data.

25ASureFire2016123.gif



CPF memeber Markcm sells CR123 cells from his web site at e-lectronics.net. All 4 of these cells terminated early due to temperatures exceeding 140 F. In spite of that, they seem to be strong performers.

25AMarkcm123.gif



The new BatteryStation cells did very good during the normal testing, so I thought I would give them a try at 2.5 amps.

While they were consistent from cell to cell, they got hot. 3 of the 4 cells shown exceeded 140 F at the end of the test. I ran another batch of 4 cells, and all 4 exceeded 140 F. I ran a third batch of 4 cells and 2 of them exceeded 140 F.

Performance is good, but if you use these in a SureFire M6, run the light continuously, and run the light past where it just starts to dim down, you can expect to have to deal with hot cells.

Here is the data.

25ABatteryStation2007123.gif


Tom
 
Last edited:

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
Tom, my hat's off to you my friend! /ubbthreads/images/graemlins/bowdown.gif /ubbthreads/images/graemlins/bowdown.gif /ubbthreads/images/graemlins/bowdown.gif

This is fantastic data you've collected, and the graphs themselves show some subtleties. For example, they show that stats alone don't tell the whole story... Ah, Wh, and minutes of runtime do provide some useful information, but they don't demonstrate that one particular brand will remain at a higher voltage during most of its runtime than will another brand. IMHO, the most important thing is that the battery will remain at the highest possible voltage under load, for as long a time as possible. For me, that appears to mean BatteryStation cells, but of course other folks will have other preferences.

You also may have shed some light on the SureFire sudden death mystery. If those cells continued to rise in temperature even after testing on them ceased, the cells could conceivably trip the thermal shutdown protection system AFTER the flashlight has been turned off, which explains why we've had various reports of people using a flashlight just fine, turning it off and setting it on the mantle, only to find the following morning that the light wouldn't even turn on. /ubbthreads/images/graemlins/ooo.gif

I do think for high drain lights, it makes sense to surround the 123 cells with aluminum or perhaps even some sort of copper sleeve, to "wick" away some of the heat the cells generate. Lights that have the cells in plastic or rubber tubes of various sorts would seem to lend themselves to shorter runtime and a higher incidence of the "sudden death" syndrome.

I can think of a couple other brands that might be worth testing...

* Panasonic (seem to be very good performers in general)
* Tekcell (mediocre performers)

Thanks again, Tom! You rule! /ubbthreads/images/graemlins/grin.gif

A little off topic: does anyone know of a link to a good cutaway graphic and perhaps descriptive information about what's inside a 123 cell, and how it works?
 

KevinL

Flashlight Enthusiast
Joined
Jun 10, 2004
Messages
5,866
Location
At World's End
Tom, FANTASTIC job on those test results! It's awesome to see the Batterystation cells coming out and doing a great job of supplying power even at the extremely demanding 2A load. Looks like I'll pick up another 20 of them soon /ubbthreads/images/graemlins/smile.gif
 

Roy

Farewell our Curmudgeon Administrator
Joined
Apr 14, 2002
Messages
4,465
Location
Granbury, Tx USA
/ubbthreads/images/graemlins/wow.gif

Great job!! Funny how your plots look like my runtime plots! /ubbthreads/images/graemlins/grin.gif Allways thought that I should have called them "Battery Run-Time Plots"....would have cut down on the confusion of what to call the vertical axis.
 

SilverFox

Flashaholic
Joined
Jan 19, 2003
Messages
12,449
Location
Bellingham WA
I have been looking at the results for a while now and was wondering:

Who makes Streamlight batteries?

Who makes SureFire batteries?

Who makes BatteryStation batteries?

Tom
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
Upon disassembly of the Surefire and the Engergizer E^2 cells, I cannot tell the difference between the two. Even the manufacturing marks, the green "washer" on the bottom, vent hole in the top, hex cutout washer on top, identical finish on the surface of the washers, cell crimping, cell marking, etc. are identical.

Here is an Energizer and Surefire photo:
surefire.jpg


Though at 1.5A loads I tested at, through ten tests, showed the Energizer slightly on top, maybe it was just the age of the cells in both your tests and mine. The Energizers had the 2014 date, the Surefires were fresh from Surefire. The Duracell Ultras I tested showed better results everywhere, and the regular Duracell 123 was slighly less than the Energizer/Surefire 123.
 

udaman

Banned
Joined
Feb 13, 2004
Messages
381
Very interesting results! I note that Li-Ion rechargeable LG 18650 2400ma cells have less of a problem with maintaining voltage and capacity under 1C discharge; of course they are twice the size, and not a one-shot deal.

Just curious, but are there many single cell LED lights that draw 2a? With exception of the Energizer cells, it would seem that all the rest have equivalent capacity and voltage sag-basically you can buy whatever is convenient and cost effective, for 1a loads.

Since Surefire, batterystation cells are relatively less costly, I wonder how 2 of these tested in series will perform (any significant difference between single cell test results), since most incandescent lights will be 2 or more cells?

testing down to 1v will get indication of useful runtime for LED's only. 2v is pretty much the end of useful runtime/capacity for non-regulated incandescent lights.

Surefire are made in the USA...by?

Sanyo made by Sanyo in Japan?

Batterystation...made in China?

Umm, is it just a function of test equipment, that the starting voltage on the 2a test is much lower than the 1a test? I'm assuming no load voltages are about the same between the fresh cells?

Energizers are smaller in diameter than the rest of these cells? At least they are smaller in diameter compared to Duracell.
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
[ QUOTE ]
udaman said:
Just curious, but are there many single cell LED lights that draw 2a? With exception of the Energizer cells, it would seem that all the rest have equivalent capacity and voltage sag-basically you can buy whatever is convenient and cost effective, for 1a loads.

Since Surefire, batterystation cells are relatively less costly, I wonder how 2 of these tested in series will perform (any significant difference between single cell test results), since most incandescent lights will be 2 or more cells?

Surefire are made in the USA...by?

Energizers are smaller in diameter than the rest of these cells? At least they are smaller in diameter compared to Duracell.

[/ QUOTE ]

The differences between the Energizer cells and the Surefire cells are a function of the batch they made in, from what I've tested at 1.5A loads.

With the wrappers off, I can not tell the difference from a Surefire to an Energizer, physically, mechanically, or electrically. Everything is the same.

So, the Surefires must be made by Energizer.
 

Bullzeyebill

Flashaholic
Joined
Feb 21, 2003
Messages
12,164
Location
CA
No expert here, but sag of 123's due to higher amp drain (2A's) woould explain lower starting voltage.

Bill
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
Yes, the higher load does cause the initial voltage to drop significantly.

5mod.jpg


The Surefires and Energizers consistently produced the results in yellow above, with the Surefires consistently slightly lower than the Energizers. Unfortunately, across several boxes of the SureFires, they were all dated the same.

The Duracells always produced curves more like the blue line.

With any mod I did to my circuit, the Surefires and Energizers consistently underperformed the Duracells. I bought various cells from Safeway, Rite-Aid, Fry's, Sportsman's Warehouse. The Surefires were all marked 05-2014.

The circuit I was using there is close to a NextGen 1000, but a little higher efficiency, since the ARC4 provides more room for components, I was able to take advantage of this room for higher efficiency enhancements.

Personally, I feel a person would want to run cells from multiple batches from each manufacturer before drawing any hard conclusions. And if they are like any other manufacturing process, what you get one month may be different six months from now.
 

elgarak

Flashlight Enthusiast
Joined
Jul 30, 2004
Messages
1,045
Location
Florida
/ubbthreads/images/graemlins/happy14.gif

Very nice results! Good to see that batterystation's hold up quite well -- have to send out the next order soon.

Now we just have to find out what goes on with the Energizers, in light of the suspected Energizer/Surefire connection.

Where can you get the Streamlight batteries? How much are they?
 

jayflash

Flashlight Enthusiast
Joined
Oct 4, 2003
Messages
3,909
Location
Two Rivers, Wisconsin
Thank you for your effort and the education, Tom. You represent what's so great about the CPF - the sharing of time and information.

Streamlight did well in another test, some months ago and they performed excellently in my TL-3. According to your test this could be a winner for high drain applications.

Although messy, might spreading thermal grease to the cell sides help to conduct heat for long on time use? Would a water soluble type as is used for ultrasounds be feasible, or plain old KY (NOT Kentucky /ubbthreads/images/graemlins/grin.gif) jelly? It may be that some other problems, such as migration, may prevent the addition of said goo from being a viable option. Just kinda thinkin' again. /ubbthreads/images/graemlins/icon15.gif
 

John N

Flashlight Enthusiast
Joined
Oct 12, 2001
Messages
2,201
Location
Seattle
This is great info!

Thanks very much for doing this!

Any chance of talking you into a very low amp comparison? It would be great to see these same batteries at .25 and .5 amps or something like that.

I am willing to send batteries for this effort. I have Surefire, Sanyo, and I have some Streamlight on order. If you PM me with your address I'll send you whatever would be helpful.

BTW, does anyone know what limitations there are, if any for shipping Li batteries?

Again, thanks!

-john
 

SilverFox

Flashaholic
Joined
Jan 19, 2003
Messages
12,449
Location
Bellingham WA
Hello John,

At this time, my CBA's minimum discharge rate is 0.5 amps. West Mountain Radio has an upgrade in the works to test at lower rates and as soon as that is available, I will send my unit in for an upgrade.

I can do a run at 0.5 amps, but the only way to go below that is to run two cells in parallel. I have done that, but have had some consistancy problems.

PM sent.

Tom
 
Top