IMR16340 vs RCR123A

[ElectronGuru]

mdocod has an excellent perspective about how these two cells compare. Posted with permission:

There are situations where a protected LiCo RCR123 is preferred over the IMR16340. For example: When the device in question has the capability of over-discharging a cell before any noticeable "warning" is given to the user. This applies to most LED lights, where the boost or buck circuits (1 cell vs 2+ cell options) would run with solid bright output until the cell was well below 2V.

In any case where this does not apply, which translates primarily to incandecent applications that deliver very noticeable feedback in the form of dimming to the user, the IMR16340 is going to be better (IMO) than any protected RCR123 when viewed as a whole... Here's some reasons I feel this way:

• no protection circuit means less things to fail.
• lower internal resistance, and no resistance added by PCB to circuit means brighter performance.
• capacity (translating to run-time) is almost always similar or better when compared at common tactical incan loads to LiCo cells in this size.
• cycle life at the loads typical of incandecent applications is going to be far better. A typical pair of LiCo RCR123 driving something like a P90 or SR-9 (standard output lamps), is operating at 2C. Testing has revealed that these cells are usually only good for a few dozen full cycles at these loads, whereas IMR16340s will handle several hundred.
• Freedom to move into "fun but impractical" territory when you just want to play. (like a WA1111 on 2 IMR16340s)

And yes, IMR16340s tend to hold up very close to the same capacity to most RCR123s because the PCB in an RCR123 does consume a fairly substantial portion of the overall cell size. In some cases, we are seeing very over-sized protected RCR123s that do deliver better capacity, but won't fit in all applications.

In the end, the best cell will always depend on application. IMR16340s have IMO made RCR123s obsolete for driving tactical incans, but protected RCR123s still have their place in LED lights.

 

[leukos]

I'll agree with that comparison at the 16340 size. IMR's are preferable for incan applications in the 16340 size. However, when comparing larger sized cells, the disparity in capacity becomes more of a case by case preference.

 

[ElectronGuru]

Yes. A nice way to think of it is as if the protection setup has a fixed size. The smaller the cell, the larger a % of the cells space it takes up. A theoretical cell, a quarter of the size of a 123, it might be 50%!

Bigger cells, smaller %, different story,

 

[csshih]

how about some runtime graphs?

odd. even though the quark is drawing almost 2C from the 750mAh cell, it still holds a pretty good capacity. hmm...

 

[Linger]

thanx for the graph, very pertinent comparison you gave us.
rcr's definitely have a place for pocket led's. That capacity difference cchih just showed, IMR's with %73 capacity of li-po, plays out more dramatically on lower light settings (e.g. @250ma, rcr hypothetically provides 1h additional run-time.)

 

[Henk_Lu]

I made a very primitive non-scientific test this evening. I took my Dereelight C2H (old version, 4 modes, Q4-5A) and ran down first an AW RCR123 (750mAh) on high, then an AW IMR 16340 (500 mAh). Both cells were freshly charged and at about 4,13 Volt

RCR : 16 minutes until the protection kicked in
IMR : 18 minutes until the protection kicked in

I know that these results can vary and that the C2H is a sucker on high, so that IMR have perhaps an advantage. I would like to know however, why I even bought the RCR123? Perhaps I don't see the advantages, I only know that IMR are safe while li-ion are not... :thinking:

 

[ElectronGuru]

I would like to know however, why I even bought the RCR123?

IMR's are better in every way, except capacity. They're cheaper, safer, and can push more sustained power. The only reason to still buy black cells is because they hold more juice. If IMRs had as much (or even nearly as much) total energy, there would be little point in getting anything else.

The question of this thread is, at what point is the cell so small, that the protection circuit of the black cell takes up so much physical space as to eliminate its capacity advantage.

 

[Deputy T.]

...there would be little point in getting anything else.

The protection circuit can come in quite handy also. I've already had one IMR cell die on me when I lost my E2e. Fortunately someone found it, but when it was returned, the tail cap (Z52 twisty) was tightened down all the way. When I got to my Ultrafire 138, one cell wouldn't charge. The Nano charger seemed to charge it, but the it didn't hold. After less than a minute of runtime, my Lumens Factory IMR-E2 dimmed lower than I've ever let it before. After replacing the cell in question, I had no further problems. I've been very satisfied with the IMRs especially in my E1e with an LM EO-E1R. There's no doubt that they out perform the LiCo cells in every way in incan setups. I even have a pair powering a Pelican 3854 bulb and it the brightest, whitest ROP in my collection.

It seems though that even at the high end of the black cell's maximum recommended amps, the capacity differences are virtually negligible. So when it come to LED applications, the protection circuit would become quite necessary. I myself can barely tell the difference in brightness between the medium and high on my Fenix PD10, so I know I'd have difficulty telling if the emitter was getting 3.7 vs 2.8 volts. Incandescents on the other hand not only change in brightness, but to my eyes the yellow tint is easily distinguishable.

 

[ElectronGuru]

The protection circuit can come in quite handy also.

Excellent point, this is turning into quite a good discussion.

My issue with LiCo's on LEDs is the sudden cut out. You're cruising along and everything seems fine, then (assuming regulated output), the safety kicks in and you're sitting in the dark. Because of the same circuit that would have kept your cell from dying.

My understanding is that with IMR's, the critical thing is less about how far the cell is discharged, then how soon it is recharged. If your E2e was gone a while, that would explain it.

So there are pros and cons of each cell type. LiCos have a safety net, but that net comes with a few 'costs'. For each person, the question is if its worth it.

 

[Deputy T.]

how far the cell is discharged, then how soon it is recharged.

I wasn't aware of that, and it is very good to know.

 

[ToyTank]

Old thread but good information. IMR's are safer than Li-co but not SAFE. IMR is a compromise between safety and performance. Li-Co unprotected still gives better capacity, while lifepo4 is SAFEST you loose a little in energy density mostly to lower voltage, but gain in safety,shelf life, life cycles, and abuse tolerance. I just found a 18500 lifepo4 in my basement I forgot about. Sitting for over a year after being completely drained voltage was 1.4...well below cut off. Cell was rated 600mah but was really 800 and cycled today at 780mah @1A. Not shabby at all. I think any other chemistry would be useless after 18 months of storage while depleted.

 

[Joe Talmadge]

Actually, this thread just reminded me of something. In my Novatac, which (I think) detects battery chemistry if I do a reset, and then cuts out before overdischarging a Li rechargeable, I have a protected 16340. In my Sunwayman v10r, which is happy sucking the battery dry, I have an IMR16340. I should probably charge both batteries back up and switch them, eh?

 

[ampdude]

Old thread but good information. IMR's are safer than Li-co but not SAFE. IMR is a compromise between safety and performance. Li-Co unprotected still gives better capacity, while lifepo4 is SAFEST you loose a little in energy density mostly to lower voltage, but gain in safety,shelf life, life cycles, and abuse tolerance. I just found a 18500 lifepo4 in my basement I forgot about. Sitting for over a year after being completely drained voltage was 1.4...well below cut off. Cell was rated 600mah but was really 800 and cycled today at 780mah @1A. Not shabby at all. I think any other chemistry would be useless after 18 months of storage while depleted.

Ni-cads would be fine as well. They can be easily nursed back to health.

A couple other advantages the IMR16340's have over the RCR123A's, is 1. the IMR cells are slightly lighter, and 2. the IMR16340 is the exact same size as the CR123A primaries, while the RCR123A is slightly fatter.

 

[Wangstang]

Are there any IMR16340's that are safe for LED use now?

Thanks
Wes

 

[Colonel Sanders]

Are there any IMR16340's that are safe for LED use now?

Thanks
Wes

I use AW IMR16340s in my SC30Ws and Incendio all the time. Nothing wrong with using IMRs but they are best used when your light has overdischarge protection built in. I find the IMR16340s to be better than the RCRs when current draw will be greater than about 1 amp.

 

[Howecollc]

-my Malkoff M30F pulls 1.36 amps from a cell hot off the charger
AWRCR123 runtime was 34 minutes
AWIMR123 runtime was 26 minutes

-my LuxV KL4 pulls 1.84 amps from a cell hot off the charger
AWRCR123 runtime in full regulation was 20 minutes
AWIMR123 runtime in full regulation was 16 minutes

The only reason I can ever see to use IMRs is if your application is pulling more than 2C from your Li-Co RCRs (which I'm aware the KL4 was).

 

[Colonel Sanders]

-my Malkoff M30F pulls 1.36 amps from a cell hot off the charger
AWRCR123 runtime was 34 minutes
AWIMR123 runtime was 26 minutes

-my LuxV KL4 pulls 1.84 amps from a cell hot off the charger
AWRCR123 runtime in full regulation was 20 minutes
AWIMR123 runtime in full regulation was 16 minutes

The only reason I can ever see to use IMRs is if your application is pulling more than 2C from your Li-Co RCRs (which I'm aware the KL4 was).

Strange how two people get different results. At 1a draw (measured using a hobby charger) I get almost identical mah pulled from either with the voltage sag being less with the IMRs. :confused: I get about 550-560mah from either. I have tried a few different cells of both type and they have been consistent.

I wonder if with some lights, the IMR's ability to maintain voltage under load maintains current draw while the ICR cell runs at a lower current due to the lower voltage under load? Just a thought. That wouldn't be the case with a nicely regulated light like a Zebralight, for example, which actually raises the current as voltage sags. In testing (SC30, SC60, SC600) I can easily see where the current is forced to rise quicker on ICRs than on IMRs.

 

[Howecollc]

That could very well be the case; especially with the M30, considering its strange regulation.