Why are CR123s so great? (And other battery questions)

[haplo]

Hello,

I did some brief searching, so I'm sorry if this topic has come up before.

What makes CR123s (and RCRs) so great? If I look at the specs on the batteries themselves, I see:

RCR123 (3v) as being 3v @ 680mAh. = 2040mW

However, I see a standard NiMH AA as:

1.2V @ 2600mAh. = 3120mW

So, a AA NiMH has more mW than a RCR123 battery.

The same goes for a AA Primary vs a CR123 Primary. The AA has more potential mWs than the CR123.

So why the craze to use CR123s for lights, and not AAs?

-----

This question came up, when I noticed that my Pila GL3 light does NOT like RCR123 batteries. If I put primaries in, it'll work fine. If I use my RCR 3Vs in it, it'll run full brightness for about 2 minutes before it starts to ramp down in brightness. I would guess the load of the light is too great a demand for the current requested from the cells. Is there a formula one can use to determine the current draw necessary to power specific bulbs?

If I'm mainly concerned about runtime, and I really want to use rechargeables, should I be looking at 18650 based lights, or something else entirely?

-----

And I had a question about protected rechargable Li-Ion batteries. If The flashlight is regulated, is it still necessary to run protected cells? None of the RCR123s I have are protected cells, so I try to make sure I never run the batteries too long in between charges.

Thanks a lot,
H.

 

[Dude Dudeson]

So why the craze to use CR123s for lights, and not AAs?

3 volts versus 1.5 is a big one.

Small form factor is another.

My Malkoff M-60 would need four AA's to run like with 2 123's. That's a really significant size/weight penalty.

 

[Benson]

Several reasons. But keep in mind while reading this, I don't have any (R)CR123 cells or lights that need them -- I'm not real sold on the cell myself. But these are the points, as I see them.

• They're really about the same energy, for apples-to-apples comparisons. Your comparison is a high-capacity AA that will be much lower after sitting for a month -- LSD AAs are what most flashaholics use, and are more comparable in charge retention to Li-ions. (And for hotwires that draw too much current for Eneloops, they use special high-current AAs with even less capacity -- the high-capacity cells are really only suited to moderate-to-heavily used lights that run through batteries every couple weeks or so.) There's also the question of whether to compare Lithium AAs, which have higher capacity than alkalines, and the charge that if you consider the cheap AA instead of the "exotic" Lithiums, you should compare it to Chinese CR123s from ebay. There's a lot of discussion that goes in circles on this stuff, and it wouldn't go on this long if either side was a clear winner.
  
• Higher voltage.
  • The typical LED has Vf of 3.0-3.5V -- if you have less than this, you need a relatively inefficient and expensive boost circuit; if more, you can use a very efficient buck circuit, or cheap resistor or linear circuits. You need 3 NiMH cells to reach the threshold, and 3xAA lights are too bulky, while 3xAAA suffer from the inherent loss of energy density as you move to smaller cells. But 2 Li cells or 1 Li-ion are sufficient.
    
  • Alternatively, if you go for incans, higher voltage means lower current for the same power, and also means more efficient light bulbs due to less heat loss by conduction. In this case there's no magic threshold, higher voltage is always better, and 9V out of 3 CR123s is well ahead of the 2.4V out of 2 AAs, making the same length of battery stack, or even the 3.6V from 3 AAs, with similar total volume and energy capacity.
• Compatibility with 18650. 18650 Li-ions are one of the most advanced cell types in existence because of their heavy use in laptop batteries. They're only rivaled, for some applications, by the 25650 and larger cells used in power tools and electric vehicles. Now not all CR123 lights will accept 18650s, because the 18650's larger diameter, but no AA lights can use 18650s. So if you're looking to swap out components, a P60-compatible light with an 18650-compatible bore running on 2 or more CR123s is a very flexible starting point, much more so than any AA platform I'm aware of.
  
• Pyrotechnics capability. CR123s are much more likely to :devil:

That's my take, anyway. Note that a AA compatible light may take advantage of several of those by using 14500 or 14670 Li-ion cells. (Of course, lights not designed to handle the 4.2V of a freshly charged Li-ion are likely to burn out the driver. Frequently they continue to work as DD on 14500s, but no longer light up on AAs.) While 14500s can't compete with 18650s for density, they're quite comparable with RCR123s.

As for form factor, some people want a short light at the expense of thickness, some want a slim light at the expense of length. Which battery is "best" mechanically depends on your holster, mode of carry, and taste, so I don't see this as an advantage to either type.

 

[CaseyS]

The voltage on RCR123's is 4.2v fully charged and 3.6v when empty.

 

[Oddjob]

Historically higher end flashlights required the higher voltage of CR123's and RCR123's. With the constant evolution of emitters and drivers as well as the introduction of low self discharge NiMH, we are starting to see comparible performances as CR123's and RCR123's. A newer light that uses one NiMH cell can give the same or better performance as a CR123 light of a few years ago. As already mentioned there is also the preference of shorter and wider over longer and thinner.

With regard to protected vs non protected, protected cells are safer to use. A regulated light is not necessarily safer unless it has overdischarge built in to it. Not letting Li-Ion cells run down too much in between charges is actually better for the cells regardless of whether they are protected or not.

 

[mdocod]

Here's one angle to consider as an argument for raw voltage per cell.

Higher voltage per cell often equates to higher driver efficiency, especially in single cell applications.

Boosting a 1.2V cell to ~3.6V for the LED will require nearly a 25% efficiency sacrifice o achieve. At the same time, when the source voltage is so low, it's very difficult to get much current to flow as there is going to be inherent resistance in the driver components, contacts, cell, etc, that all add up...

For example: If you wanted to make a 1 cell light that runs at ~200 lumen, you would need to achieve something like 2.5W give or take on a really decent emitter (roughly). Calculate for driver losses and the input wattage would have to be over 3W, which means close to 3 amps need to flow from the 1.2V cell. In order for this to happen, the total resistance of the input side of the circuit, the contacts, the cell itself, the switch, would have to be ~0.3 Ohm. Achieving this is probably doable, but not practical as the size of the components on the regulator would have to be large and expensive.

Moving up to a 3V cell, or a 3.7V cell, makes a huge difference in making this goal attainable as the resistance of the input circuit and all things involved can be much higher without a problem (which translates to smaller cheaper component size). Combine that with the higher efficiency of the driver (less voltage differential between driver input and output), and the total resistance on the input side can rise to ~2 ohm and the system would work.

------

Now lets look specifically at the GL3 and understand the likely roots behind your disappointing experience on 3V RCR123 cells:

Historically speaking, 3V rechargeable "RCR123" cells have been very low on the totem pole of energy density for several reasons.

There are 2 types of "3.0V" cells out there.

LiFePO4 RCR123:
This type of cell inherently has a lower energy density than most modern li-ion cells as it is using a cell chemistry that sacrifices energy density for power density and safety (in most applications). This chemistry was originally developed, and continues to be developed in larger format cells with a target market of things like power tools and electric/hybrid vehicles. The market saw that the 3.2V cell would have an advantage for owners of devices who wanted a rechargeable option who did not have 3.7V compatibility. The result is a cell with very low energy density (not a lot of R&D can be afforded for such a small market of 1 cell size).

Voltage Bucked LiCo cell:
This is actually just a 3.7V cell using the same chemistry as other RCR123s out there, but they have made the cell itself a bit smaller, and installed a simple voltage buck circuit "under the hood" that cuts operating voltage down to around 3V depending on load. Since the space dedicated to energy storage has been shrunk, and a percentage of energy output is converted to heat by the buck circuit, these cells tend to perform poorly in overall energy density when tested, but often just slightly better than a LiFePO4 cell depending on load.

In either case, the true capacity of these cell types is going to be in the 200-500mAH range depending on load.

------

The GL3 is designed to take either 3 CR123s or 2 18500 size protected LiCo cells (Pila sells protected 18500s as "300P" model rechargeable cells). The 18500s will result in similar run-time as what you are getting on 3 CR123s (depending on if you are using LED or incan lamp modules). Interesting that you asked if you should be looking into a "18650" powered light, as you already own the next best thing, which is something that runs on 18500s. This is the rechargeable direction you should be headed. Pick up a Pila IBC charger and either a pair of Pila 300P cells or AW protected 3.7V 18500s.

-----

To answer your question about whether you would need protected cells;

It's actually MORE important to have the protected cells in a regulated light, because a regulator will be able to continue discharging cells and producing full output even when cell voltage drops WAY below the recommended minimum (which destroys rechargeable cells, especially certain types of li-ion). In your case, with a 3 cell application, the level to which the regulator can discharge each cell before noticeable dimming occurs is even worse than in a "2 cell" size application.

If you have been using unprotected 3.0V RCR123 cells in your Pila GL3, and you are saying that you are seeing dimming very quickly, this is very revealing about what probably has happened to those cells. All you would have had to do, (depending on what cell chemistry your cells are), is discharge to the point of noticeable dimming a few times, and each cell would have been discharged to ~1V give or take, even more likely, one of the cells has probably been reverse charged a few times in this application, (where 2 cells have held above ~2.5V and one under-performing cell depletes first and gets reverse charged on the circuit, further destroying it). Basically, the first time you used your 3 RCR123s and ran it till dimming could have really "done a number" on those cells, and resulted in poor and getting poorer performance ever since. Remember, with 3x3V cells in there, you have a ~9V source voltage being regulated down to ~3.6V for the LED, which means that the input voltage has to drop below the Vf of the LED plus whatever over-head the regulator requires (could be as little as ~0.1V) before any dimming occurs.

In a direct drive light, like an incan, you would get very noticeable feedback from the light (output would drop) when the cells started to deplete close to their minimum recommended level. I only recommend using unprotected cells if they are safe chemistry (like LiFePO4 or LiMn {IMR}) and I only recommend using them in applications that provide user feedback regarding the state of charge.

Before you use your 3.0V RCR123s any more, I think it would be crucial for you to share with us just exactly which cells you own and what charger you are using to charge them. As there may be more to learn from this and there may also be some serious safety hazards at play depending on which cells you have.

Just reviewing this post, I feel like I have been a bit scattered in my presentation of thoughts, if anything is unclear, let me know, I'll try to re-iterate more clearly....

-Eric

 

[haplo]

In either case, the true capacity of these cell types is going to be in the 200-500mAH range depending on load.

I always thought those cells didn't last very long, even when new. They say 680mAh, but they aren't LiFePO4 cells, as the website shows the "new LifePO4 cells" being available, long after I bought these rechargeables.

------

Pick up a Pila IBC charger and either a pair of Pila 300P cells or AW protected 3.7V 18500s.

Is the Pila charger really the way to go? I searched CPF, and several other people were also complaining on the lack of availability and the high price of that unit.

-----

In a direct drive light, like an incan, you would get very noticeable feedback from the light (output would drop) when the cells started to deplete close to their minimum recommended level. I only recommend using unprotected cells if they are safe chemistry (like LiFePO4 or LiMn {IMR}) and I only recommend using them in applications that provide user feedback regarding the state of charge.

I don't know on my Pila GL3 if its direct drive or not. Its a 5 year old Pila GL3 with an incan main. It doesn't look like the new Pila GL3s at all.

Before you use your 3.0V RCR123s any more, I think it would be crucial for you to share with us just exactly which cells you own and what charger you are using to charge them. As there may be more to learn from this and there may also be some serious safety hazards at play depending on which cells you have.

-Eric

The cells I have been using are:

http://www.batteryspace.com/rcr123a30v680mah204whrechargebleli-ionbattery1pc.aspx

I've had these for a number of years, long before I heard anything about protected cells.

The charger is from them also:

http://www.batteryspace.com/4pcsrec...smartworldwidetravelchargerch-rcr123ac01.aspx

-----

I bought those cells originally to power a couple of single cell Fenix lights. Since then I've been drawn into flashaholism, and now own a few Surefires, the Pila, numerous Fenix lights, and an Eagletac. The Surefires, and the Pila have always had primary CR123s in them, until just recently, when I decided to use my RCR123 3Vs in the Pila because I was out of CR123s. It worked good for the 30 minutes of continual burn, but then started to get dim. It wasn't until after I charged the batteries that I noticed that it won't stay bright anymore for more than 30 seconds.

So, considering what I batteries I have, would you continue to use them in single cell applications, or would you pitch them all and start over? Both the Pila and my Eagletac can take 18xxx series batteries, so I've been thinking about getting a charger and some cells.

Thanks a lot for the thorough reply,
H.

 

[mdocod]

I always thought those cells didn't last very long, even when new. They say 680mAh, but they aren't LiFePO4 cells, as the website shows the "new LifePO4 cells" being available, long after I bought these rechargeables.

Is the Pila charger really the way to go? I searched CPF, and several other people were also complaining on the lack of availability and the high price of that unit.

The Pila IBC is slightly harder to track down and more expensive, but for a general user for 3.7V cells, it's about as good as it gets besides hobby style chargers that would require custom rigging to perform charging operations.

I don't know on my Pila GL3 if its direct drive or not. Its a 5 year old Pila GL3 with an incan main. It doesn't look like the new Pila GL3s at all.

Almost every incan flashlight ever made is going to be "direct drive," and your GL3 is no exception to this 99% rule of thumb. The reason your cells are performing so poorly when driving this light is a combination of factors but the bottom line is that these cells, at that age, are not going to work well for driving that lamp....

The cells I have been using are:

http://www.batteryspace.com/rcr123a30v680mah204whrechargebleli-ionbattery1pc.aspx

I've had these for a number of years, long before I heard anything about protected cells.

The charger is from them also:

http://www.batteryspace.com/4pcsrec...smartworldwidetravelchargerch-rcr123ac01.aspx

-----

Interesting, those are some of the only unprotected 3.0V buck regulated LiCo cells on the market. Most other cells of this chemistry with built in voltage bucking are protected.

I bought those cells originally to power a couple of single cell Fenix lights. Since then I've been drawn into flashaholism, and now own a few Surefires, the Pila, numerous Fenix lights, and an Eagletac. The Surefires, and the Pila have always had primary CR123s in them, until just recently, when I decided to use my RCR123 3Vs in the Pila because I was out of CR123s. It worked good for the 30 minutes of continual burn, but then started to get dim. It wasn't until after I charged the batteries that I noticed that it won't stay bright anymore for more than 30 seconds.

So, considering what I batteries I have, would you continue to use them in single cell applications, or would you pitch them all and start over? Both the Pila and my Eagletac can take 18xxx series batteries, so I've been thinking about getting a charger and some cells.

Thanks a lot for the thorough reply,
H.

My gut tells me that the combination of age, use, and that first 30 minute burn in the GL3 (which would be a very high discharge rate for those cells), has probably pushed the cells to the end of their usable life. It's also possible that the higher drain rate of the GL3 may have over-heated the voltage-buck circuit in the cells which may have caused irreversible damage to both the circuit and the cells.

The fenix single cell lights you were using them in would have been very capable of over-discharging those cells repeatedly. This adds up to major wear and tear and possible danger. (The boost circuit used in the fenix 1xCR123 lights is very similar in some cases almost identical to those used in the 1xAA lights, meaning that you would get good output all the way down to around 1V on the input side, which means that these cells have probably been over-discharged numerous times if you haven't been taking care to prevent this from happening)

Interesting note: If the cells had not been abused in the Fenix lights for a few years beforehand, you probably would have blown the bulb in the GL3. "3.0V RCR123" cells do not behave exactly like the CR123 primary cells that the manufacture tries to tell you that they can replace. Under the load of the stock lamp, CR123s will sag to ~2.5V normally, with peak output fresh from the package still dipping quickly below 3V. Whereas, the voltage bucked LiCo cells, are actually capable, when new, of delivering 3.2V+ per cell into those loads, which would have likely blown the bulb.

The type of cells you have are 3.7V LiCo cells with buck regulation that results in ~3.0V running voltage. Given the history here, If it were me, I would retire those cells and move on to new options as they could be getting to a point in their life where they may not be entirely safe.

I think you should try to move in the direction of 18XXX cells in any way possible for rechargeable stuff. As for your surefires, each one has different compatibility and some are iffy but if you list which lights you wish to "convert" to rechargeable myself or others here may be able to help in getting you on the right track. In some cases there are going to be tradeoffs and in some cases it may be better to just stick with primary cells in some lights. I generally recomend that folks just run their single cell fenix lights on CR123s and buy an alternative that is compatible with a 3.7V protected RCR123 or 3.7V protected 14500 to replace them for daily use. (or just sell the Fenix lights..)

Picking up a Pila IBC will open the door to using every size of 3.7V cell from RCR123 up to 18650s. This will give you a "launch point" to work from for future flashlight purchasing and conversion of existing flashlights. I always urge people to try to avoid standardizing around "3.0V RCR123" cells as they are not going to provide a good performing and widely compatible platform to work from in a flashaholics progression through new trends and performance capabilities. 3.7V cells in a variety of sizes can cover a wide range of possible flashlight types.

I think a pair of protected 18500 cells would really breath some new life into that GL3, and if you want to really go nuts, pick up a new lamp for it from LumensFactory. The EO-9 is a favorite among many. 18500s should run the stock lamp for well over an hour, and something like an EO-9 for about 40 minutes.

-Eric

 

[shark_za]

The facts in your first post are misleading.

RCR123 are 4.2v as pointed out and you will struggle to find a AA NiMh that can actually give 2600mAh.
Realistic values are around 2000-2300 when drained at anything close to a practical 500mA.

 

[haplo]

The Pila IBC is slightly harder to track down and more expensive, but for a general user for 3.7V cells, it's about as good as it gets besides hobby style chargers that would require custom rigging to perform charging operations.

I'm taking your advice and ordering the Pila IBC. I've sent an email to a CPF friendly retailer asking for a final price on the IBC this morning.

My gut tells me that the combination of age, use, and that first 30 minute burn in the GL3 (which would be a very high discharge rate for those cells), has probably pushed the cells to the end of their usable life. It's also possible that the higher drain rate of the GL3 may have over-heated the voltage-buck circuit in the cells which may have caused irreversible damage to both the circuit and the cells.

The fenix single cell lights you were using them in would have been very capable of over-discharging those cells repeatedly. This adds up to major wear and tear and possible danger. (The boost circuit used in the fenix 1xCR123 lights is very similar in some cases almost identical to those used in the 1xAA lights, meaning that you would get good output all the way down to around 1V on the input side, which means that these cells have probably been over-discharged numerous times if you haven't been taking care to prevent this from happening)

I forgot to mention that 2 of those cells were first used in my Eagletac T100C2 Mk2 light for almost an hour before being moved to the Pila after a fresh charge. So, I probably did kill those cells, since they were starting to dim in the ET after an hour. That was the first time I've ever "abused" those RCR cells. I usually used them in my Fenixs for EDC, so they were recharged frequently, and never ran for very long. This time around, I was deep in the woods and had no choice - I had to run my lights to find my way out.

Fortunately, the ET light will take an 18650, so that is going to be ordered sometime this week as well.

I think you should try to move in the direction of 18XXX cells in any way possible for rechargeable stuff. As for your surefires, each one has different compatibility and some are iffy but if you list which lights you wish to "convert" to rechargeable myself or others here may be able to help in getting you on the right track. In some cases there are going to be tradeoffs and in some cases it may be better to just stick with primary cells in some lights. I generally recomend that folks just run their single cell fenix lights on CR123s and buy an alternative that is compatible with a 3.7V protected RCR123 or 3.7V protected 14500 to replace them for daily use. (or just sell the Fenix lights..)

Since I have the Fenixs, and I'm not quite ready to ditch them yet (my next EDC will be a single cell neutral white (c'mon Quark, re-run the 123 in neutral white!)), is there any harm in buying a few protected LiCo 3V RCR123s and using the charger I currently have? I was thinking of these cells here:

http://www.batteryjunction.com/rc390reliba.html

On the site where I purchased my unprotected cells, they sell a different charger for the LiFePO4 cells, so they don't advise using the same charger for different 3V chemistries.

On another note - do you have any experience with the PrincetonTec Apex headlamp? They recently released an 8 AA cell version (I have the 4). I was wondering if the regulator of that headlamp is safe to use 14500s, and if I'd be any better off using 14500s over normal NiMHs that I currently am using. Total mWh are similar between batteries, but I'm not sure what is more advantageous - higher voltage and less mAh, or lower voltage and more mAh for the cells running the unit.

Thanks again for all your input (and thanks everyone else who has responded as well!),
H.

 

[Launch Mini]

I have a question also about the 123's in general. Might be the wrong place or answered already.
I bought some Surefire 123's at $3.50/battery ( Canadian), yet I see Duracell and Engergizer 123 lithiums going for $8.99 to $9.99/battery in the camera section of various stores.
Is this just advertising money they are covering or is there a significant difference in these batteries.
Thanks
John

 

[yellow]

the "importance" of CR123s was at a time, when led were totally unknown/uncapable for lighting.
And/or before all that Li-Ion thing exploded.

1*18650 + a driver that pushes the led, no matter what voltage the cell has, thats the king atm

 

[mdocod]

I'm taking your advice and ordering the Pila IBC. I've sent an email to a CPF friendly retailer asking for a final price on the IBC this morning.



I forgot to mention that 2 of those cells were first used in my Eagletac T100C2 Mk2 light for almost an hour before being moved to the Pila after a fresh charge. So, I probably did kill those cells, since they were starting to dim in the ET after an hour. That was the first time I've ever "abused" those RCR cells. I usually used them in my Fenixs for EDC, so they were recharged frequently, and never ran for very long. This time around, I was deep in the woods and had no choice - I had to run my lights to find my way out.

Fortunately, the ET light will take an 18650, so that is going to be ordered sometime this week as well.

Just an interesting tidbit to ponder. I was just looking at the datasheet for the cells you have there. The data-sheet claims 550mAH capacity, with a 500mAH minimum. The sales guys got ahold of it and called it a 680mAH cell, and we know that both the sales team and the people who wrote the data sheets are all going to be a bit overly optimistic... Funny how that works eh?

Since I have the Fenixs, and I'm not quite ready to ditch them yet (my next EDC will be a single cell neutral white (c'mon Quark, re-run the 123 in neutral white!)), is there any harm in buying a few protected LiCo 3V RCR123s and using the charger I currently have? I was thinking of these cells here:

http://www.batteryjunction.com/rc390reliba.html

On the site where I purchased my unprotected cells, they sell a different charger for the LiFePO4 cells, so they don't advise using the same charger for different 3V chemistries.

I am fairly confident that those "tenergy 900mAH" cells from batteryjunction will work with the charger that you have as they are using basically the same method of voltage bucking and the same LiCo chemistry. The "tenergy 900s" have that protection which would certainly make me feel better as far as the safety is concerned. (I would consider your current batch of "powerizers" to be unsafe at this point.)

Just as an FYI, just because the label capacity is higher does not mean that true capacity will be.. Interesting that more space could be consumed by the addition of the PCB and yet still come out with a higher capacity than an unprotected version... Unless the cell is over-sized by the size of the PCB, it's very unlikely to have any more true capacity than your powerizers did when they were new)

On another note - do you have any experience with the PrincetonTec Apex headlamp? They recently released an 8 AA cell version (I have the 4). I was wondering if the regulator of that headlamp is safe to use 14500s, and if I'd be any better off using 14500s over normal NiMHs that I currently am using. Total mWh are similar between batteries, but I'm not sure what is more advantageous - higher voltage and less mAh, or lower voltage and more mAh for the cells running the unit.

Thanks again for all your input (and thanks everyone else who has responded as well!),
H.

The 4AA and 8AA versions of the Apex both run on input voltages that are basically the same. The 8AA model runs 2 banks in parallel to double the capacity of the pack without increasing voltage. I have read that the unit has been tested by many as workable with 2 li-ion cells in series, but I'm not sure about any more than that.

If you were going to convert to li-ion, you should either re-wire the stock pack to 2S2P for use with 4x14500s resulting in ~7.4V operation, or build a totally separate pack that would take 18650s, (this would be the prefered choice, as 2 18650s would be about the same stored energy as 6 AA NIMH cells while at the same time being similar in weight/size (maybe less).

-Eric

 

[mdocod]

I have a question also about the 123's in general. Might be the wrong place or answered already.
I bought some Surefire 123's at $3.50/battery ( Canadian), yet I see Duracell and Engergizer 123 lithiums going for $8.99 to $9.99/battery in the camera section of various stores.
Is this just advertising money they are covering or is there a significant difference in these batteries.
Thanks
John

Hi John,

Duracell and SureFire CR123s are made in the same factory. Just different stickers slapped on there. Stick to USA made cells and most will be made in the same handful of factories. If I recall the Duracell/Panisonic/Streamlight/SureFire (and a few others) are all made in the same place and are solid performers.

Energizer cells have an excellent track record of safety however they don't perform as well when used at the really high discharge rates. (unless something has changed).

-Eric

 

[haplo]

(I would consider your current batch of "powerizers" to be unsafe at this point.)

-Eric

Eric,

I dusted off my multimeter and charged and tested the 3V powerizers I have. One of the cells appears bad, so I didn't attempt to charge it - it measured 0.82V, and since these cells aren't protected, I didn't think that was a good thing.

Either way, the cells are as follows:

Immediately after charge / 30 minutes after charge

1. 4.11 / 4.10
2. 3.98 / 3.74
3. 3.80 / 3.82 (odd)
4. 4.14 / 4.12
5. 4.08 / 4.06
6. 4.16 / 4.14
7. 4.11 / 4.10

I was a bit puzzled why the voltage was that high. I was under the impression that 3V cells were supposed to be 3.6V prior to load, and around 3.2V under load.

I assume cells 2 and 3 are probably (going/are) bad at this point, but the rest of the cells seem ok with that simple test.

Thoughts?

Thanks,
H.

 

[mdocod]

Hello haplo,

The buck circuit used in the cells needs a load present in order to "kick in" and start actually bucking the voltage down. A multi-meter does present a very very minor load, but the load varies from multi-meter to multi-meter so different people will get different results when measuring the voltage of a buck regulated cell. Most people who measure these cells will get a reading ~3.85-4.0V fresh from the charger when they are in good shape. With that in mind, I think it's semi-fair to conclude that your multi-meter may present and even smaller load than most out there resulting in access to a more true bare cell reading.

A LiCo cell would ordinarily charge to ~4.20V open circuit. Because of the buck circuit in there, it's impossible to know what the true state of charge is, but we can extrapolate your results to get at least a fuzzy picture of these cells state of health.

I would take a 3rd voltage test at 24 hours. At this point, I think it's fair to say that the #2 cell should *probably* be recycled. After 24 hours, anything that has dropped below 4.0V you might also consider on the way out.

The #3 cell may in fact be fine, but it's impossible to really know for sure... This is one of the reasons I really despise these voltage bucked cells, there's no good base-line standard way to measure them to get really useful information about their state of health. See, with most cheap electronics, there is usually going to be a tolerance in behavior of +/- ~5-15%. That tolerance applies to the buck circuit used in these cells and can cause variations in voltage readings that have nothing to do with the health of the cell itself.


-----

With LiFePO4 chemistry "3.0V" RCR123s, those do come off the charger ~3.6-3.8V and tend to run ~3.3V diminishing to ~2V under a load (depending on load).

-----

Under very light loads, like low modes on LED lights, both "3.0V" RCR123 types can actually wind up delivering 3.5+V.

-Eric

 

[fyrstormer]

A couple of thoughts:

1. While the very best AAs might have more mWh capacity than average CR123s, most AAs are not the very best, and either way, CR123s can discharge at a higher rate than AAs can, which was much more important back when xenon flashlights were all the rage among people who actually used them and actually needed them to run as bright as possible.

2. CR123s don't self-discharge nearly as fast as AAs, and the chemistry is much less likely to leak, which is important for the military, which popularized the format in the first place.

3. The CR123 form factor lends itself to being useful in single, double, triple, and even quadruple-cell configurations. Even double-AA is too long for pocket carry, requiring a holster of some sort, and quadruple-AA would just look silly. The thicker tube necessary to use CR123s fits more comfortably in the hand.

 

[Bullzeyebill]

The problem is that the very best rechargeable AA's are now the lower capacity AA's, the LSD cells, Eneloops being mostly at the top of the heap, with about 2000mAh. CR123's will have an advantage watt hour wise, and will have an easier time supplying high current than the best LSD AA. The single CR123 will look better, over time, as the efficiency of LEDs improve, and that short form factor, compared to an AA, is really nice.

Bill