Selfbuilt's CR123A Battery Comparison 2013

[photonhoer]

selfbuilt

I really appreciate and often rely on the many reports you have shared with us here! Thank you very much.

As I read your initial post, I found myself wondering how many individual batteries are represented in each of your curves?

Clearly, the higher the number (N) in any sample, the more reliable the information will be. An N=1 can be useful, and we all often rely on such small samples. I certainly understand that you operate under budget constraints [time and $$].

As a user of your data, I would simply like to know — if I could — what your sample sizes are in order to inform the interpretations I make of your data. Can you share any information?

Thanks, John

 

[selfbuilt]

Clearly, the higher the number (N) in any sample, the more reliable the information will be. An N=1 can be useful, and we all often rely on such small samples. I certainly understand that you operate under budget constraints [time and $$].

Absolutely right - this is always the key question. It's hard to know what to make of n=1 (which is often all I have of lights).

For the batteries, all the samples tested above were at least n=2, some n=3. I choose a "representative" runtime to depict in the graphs (since summing the curves alters their typical appearance).

There was surprisingly little variability between samples of a given batch, at least for the USA cells (see the analysis of two batteries tested yeas apart above). I saw a lot more variability on the China cells. But since I didn't have sufficiently high numbers to do a proper statistical comparison, I didn't pursue this further.

 

[Neo9710]

Absolutely right - this is always the key question. It's hard to know what to make of n=1 (which is often all I have of lights).

For the batteries, all the samples tested above were at least n=2, some n=3. I choose a "representative" runtime to depict in the graphs (since summing the curves alters their typical appearance).

There was surprisingly little variability between samples of a given batch, at least for the USA cells (see the analysis of two batteries tested yeas apart above). I saw a lot more variability on the China cells. But since I didn't have sufficiently high numbers to do a proper statistical comparison, I didn't pursue this further.

Just wanted to say THANKS for this thread...My new light uses CR123's and I use my light QUITE OFTEN! So, HELPFUL! Looks like Im placing an order for the Titanium Innovations...

 

[Grijon]

Stellar work, indeed.

Thank you, selfbuilt!!

 

[RI Chevy]

I have a question regarding the useful voltage range of a Panasonic CR123A cell. I had the cell in an Elzetta Alpha. I used the cell until the light output eventually went out. I measured the cell on a DMM and it was 2.49v's. A day later, I put that same cell into a Malkoff CR123 host with a M31WL drop in (.8 to 3.3v's), figuring that the drop in would take the cell down to somewhere around 1.2v's or so based on the voltage range. I based this on my usage of a single AA cell with the Malkoff AA host and the M31WL drop in. The drop in lights up with a single AA cell that is generally 1.5v's, and runs down to about 1.2v's. But to my amazement, the CR123A cell that read 2.49v's on my DMM, would not even light up the M31WL drop in. ??? I am a little confused. Is the useful voltage of a CR123A cell only down to 2.5v's? So basically you only get about .5 volts out of a 3v cell? Any insight into this would be very appreciative.

 

[selfbuilt]

Any insight into this would be very appreciative.

I will have to leave to the experts on lithium battery chemistry to explain, but what you observe is consistent with my experience. The resting voltage of a primary CR123A never drops by very much. For that matter, neither does a typical alkaline cell. This is why you need to use pulse-current tester if you want to gauge the remaining capacity of a partially depleted primary AA/AAA or CR123A. They are not like lithium ions, which show a clear reduction in resting voltage.

 

[RI Chevy]

Thanks for the response. It is just odd, that a battery with 2.49v's would not fire up the Malkoff M31WL drop in.

 

[RI Chevy]

Anyone else have first hand experience?

 

[selfbuilt]

Thanks for the response. It is just odd, that a battery with 2.49v's would not fire up the Malkoff M31WL drop in.

I can't speak to that light, but I have seen plenty of others where a partially-drained CR123A would not activate. It seems like many lights (when already on) will happily drain a CR123A until it is completely dead, but have a threshold for activation that is considerably higher. :shrug: This is why only so many lights are suitable as "battery vampires" (i.e., able to activate and drain a cell no matter how little juice is left).

 

[GrainOfLight]

Hi @selfbuilt

I am in a bit pf a pickle and after reading this forum post I thought this could be the place to get some assistance

Currently i am getting a unique shaped flashlight made for me in China. The hardware I have selected is a CREE XM-L2 LED powered by 1 x 3.7V 650mAh CR123A rechargeable battery (RCR123A).

The manufacturer states that in order to achieve 220 LUMEN output, they will only be able to provide me with 30 minutes of runtime using their battery.

I believe I should be able to achieve longer than this based on a couple of reasons:

1) The data sheet of the CREE XM-L2 states that the Luminous Flux (lm) @ 700mA is approximately 260lm (using Neutral White lamp). If the capacity of the battery is 650mA, shouldn't I be able to achieve closer to one hour of runtime?

2) Looking at existing flashlights on the market using XM-L2, they seem to achieve a longer runtime with their flashlights: for example 'Nitecore MH1C' claims to have 220 Lumens for 1 hour. You can see the device here: http://www.nitecore.com/productDetail.aspx?id=68#.VTliGCGeDGc

Nitecore states to use their own RCR123A batteries for these results.

Do you think that Nitecores claims are reasonable and I should look for another manufacturer to complete my light with better runtime? Or do you think that the claim from my manufacturer i reasonable?

Do you think it has something to do with the batteries that Nitecore are using or is it more related to the driver circuit within the board?

All-in-all I want to know if I can achieve a runtime of 1 hour, using a CREE XM-L2 producing 220 Lumens and if so is there a CR123A battery that would be best suited for this?

Thanks for any assistance, it would be greatly appreciated, i look forward to hearing back

 

[HKJ]

You can get the actual capacity of some RCR123 on my website, but why not make the battery tube slightly larger in diameter and support 18350 cells, you get considerable more energy for 2mm more in diameter.

 

[selfbuilt]

Do you think that Nitecores claims are reasonable and I should look for another manufacturer to complete my light with better runtime? Or do you think that the claim from my manufacturer i reasonable?
Do you think it has something to do with the batteries that Nitecore are using or is it more related to the driver circuit within the board?

Sorry, but I haven't tested various RCRs to confirm their capacity. I suggest you follow HKJ's link, since he has done the most extensive work here.

But as a general rule, I would treat all manufacturer's claims as suspect until verified independently. My own flashlight testing shows that most runtime specs are not very accurate. Any battery with the same chemistry should give you equivalent output - it is mainly runtime that will vary, directly with actual capacity. For that, you will need find independent tests - manufacturers claim what ever they want.

 

[GrainOfLight]

You can get the actual capacity of some RCR123 on my website, but why not make the battery tube slightly larger in diameter and support 18350 cells, you get considerable more energy for 2mm more in diameter.

Good point and trhanks for the suggestion. I didn't know these batteries existed. They seem to make CR123A batteries obsolete... Are they a new technology, because I can't see a lot of flashlights out there using this battery. What are the differences between 18350 and 16340 (CR123A), besides the capacity and diameter. Are there any downfalls?

Thanks again!

 

[GrainOfLight]

Thanks! Yeah im being very suspicious on these Nitecore claims. I just want to know a general runtime operation for the common XM-L2 using a rechargeable RCR123A at approximately 220 LUMENS. After hearing HKJ's comment I am looking forward to hearing about the 18350 capabilities!

 

[HKJ]

Good point and trhanks for the suggestion. I didn't know these batteries existed. They seem to make CR123A batteries obsolete... Are they a new technology, because I can't see a lot of flashlights out there using this battery. What are the differences between 18350 and 16340 (CR123A), besides the capacity and diameter. Are there any downfalls?

They are just another size of LiIon batteries and the slight increase in diameter leaves space for more capacity.
When working with RCR123 size you also has to be aware of the different voltages. 16340 and 18350 usual has higher voltage than CR123 (A maximum of 4.2 volt), but you can get some that has a maximum of 3.4 volt (Most common in 16340 size).
I have written a bit about it here: http://lygte-info.dk/info/CR123A and rechargeable substitutes UK.html

 

[GrainOfLight]

Thanks, th einformation on your page is amazing, I really appreciate your time and effort gone into this! I really like the look of the Ultrafire XSL18350 batteries. I think Trustfire make these, do you know? The Ultrafire XSL18350 seems to be perfect for my light and according to your graphs and data along with the specifications from the CREE XM-L2 we have as follows:

The CREE XM-L2 delivers approximately 220 LUMENS @ 700mA discharge current. Therefore according to your results, if I request drive the LED @ 700mA, we should get approx 220 LUMENS and see the runtime fall somewhere between 90 mins (discharge @ 0.5A) and 45 minutes (discharge @ 1A): hopefully this falls close to the 1 hour mark

I see the battery doesn't perform too well at currents over 2A, but for my flahslight with high setting of 220 Lumens, I think I will be fine.

Do you have any other comments on this battery since doing the test? or do you have another alternative in mind for me that blows this battery out of the water?

Thanks again, you really have saved me a lot of research time with your data!

Terrence

They are just another size of LiIon batteries and the slight increase in diameter leaves space for more capacity.
When working with RCR123 size you also has to be aware of the different voltages. 16340 and 18350 usual has higher voltage than CR123 (A maximum of 4.2 volt), but you can get some that has a maximum of 3.4 volt (Most common in 16340 size).
I have written a bit about it here: http://lygte-info.dk/info/CR123A and rechargeable substitutes UK.html

 

[HKJ]

... Ultrafire XSL18350 batteries. ...
Do you have any other comments on this battery since doing the test? or do you have another alternative in mind for me that blows this battery out of the water?

Ultrafire is often a bad name on batteries, the reason is that multiple factories uses that name and some of them only print wrappers. The biggest problem is with 18650 size, I do not know how much problem there is with the 18350 size.

 

[GrainOfLight]

Thanks, have you hear of Trustfire? I am looking at these and ordering direct from the Trustfire factory, so they should be legit.

http://trustfire.en.alibaba.com/pro...0mAh_18350_Ecig_MOD_rechargeable_battery.html

Ultrafire is often a bad name on batteries, the reason is that multiple factories uses that name and some of them only print wrappers. The biggest problem is with 18650 size, I do not know how much problem there is with the 18350 size.

 

[HKJ]

Thanks, have you hear of Trustfire? I am looking at these and ordering direct from the Trustfire factory, so they should be legit.

http://trustfire.en.alibaba.com/pro...0mAh_18350_Ecig_MOD_rechargeable_battery.html

Trustfire is not a quality brand, but may work.

 

[GrainOfLight]

Hi HK and everyone else.

After a lot of forum reading and questioning, I get the feeling that there is a strong following for AW batreries including the IMR 18350.

From what I can understand is that IMR allow for a higher discharge current but lower capacity that ICR's.

In my application I will only have a drive current of around 700mA to 1A. So am I correct in saying that the bets battery for me to use will be a ICR 18350 rather tan an IMR?

If so, unfortunately AW does not provide the ICR, this may be because Panasonic does not produce the cells... or something along those lines.

Does anybody know of a manufacturer that sources high quality (potentially Japanese) cells and distributes them as ICR 18350, or am I better suited to selecting an IMR?

Thanks for all the help.

Terrence

Trustfire is not a quality brand, but may work.