Issues in powering a Malkoff M30 with one RCR123?

[jaundice]

I'm putting together a lego pocket rocket out of an Aleph 1x16340 body, a VME head, a Malkoff M30, and an Aleph switch with an McE2S switch. Since I'm using an M30, I would like to have full output, which requires using a 3.7v Li-Ion battery. Using a primary 123 cell will reduce output to ~160 lumens from ~235 lumens at full power.

My concern is that at 1.0 amp draw, an AW RCR123 has a capacity of 419 ma. (from Silverfoxes Li-ion shootout thread here: http://www.candlepowerforums.com/vb/showthread.php?t=117117 )

The M30 has approximately a 1.1 amp draw. Therefore, an RCR123 would be running at approximately 2.5C, or over the 2C practical limit.

Is this an issue? Are there any problems with running an AW cell at this output?

One thing I noticed is that at a 1.0 amp draw, the AW IMR 16340 has a similar capacity as the LiCo version (500 ma). This would argue for the use of an IMR cell. However, I strongly prefer to use a LiCo cell for the protection circuit, to prevent over-discharge.

What should I do?

Thanks,

-John

 

[Kestrel]

Whew, a number of things to cover here.

I highly recommend reading the master Malkoff M60 & M30 thread by gsnorm, if you haven't read it yet. It is the largest single collection of M60 & M30 info on CPF, and has a lot of information about cell configurations, outputs, and resultant runtimes.

It is somewhat common for folks here to use an AW RCR123 with the M30. I think that it is generally considered to be at the 2C discharge rate, although as you indicate it difficult to ascertain what capacity the small RCR123's have, depending on the drain rate, so it's hard to know if you're really at 2C or not. The IMR16340 should have more usable capacity at that drain rate as you suggest, ~500 vs ~420 mAh or so. If I already had either cell, I wouldn't sweat it, but I would slightly prefer the IMR16340 (which I have), as it will age more gracefully over time. The 2C load on the RCR123 will cause a relatively-early demise (100 cycles or even less, just a guess, trying to recall what I've read here), beyond which the overall safety of the cell is compromised - i.e. the LiCo RCR123 would need to be retired & disposed of to maintain overall safety. Perhaps mdocod or Silverfox might chime in, they have hands-on experience with this (end-of-life) LiCo behavior.

Also, you will definitely notice the considerable drop in output near the end of the discharge of the IMR16340 before the cell becomes overdischarged. I run my IMR16340's much harder (direct drive P7, ~12 minute runtime on 'high') and have complete confidence in them, easily noticing when they are low.

The other thing to keep in mind is that there are other modules (either on the market or upcoming) that provide/will provide higher output than the M30, and will require the IMR16340 cell as they will utilize higher currents to operate at their full potential. So if you think there is a possibility of upgrading your configuration in the future, IMR16340's might be the way to go. You might want to look into the Nailbender XP-G modules for starters, this can get you to a P7 or an XP-G drop-in for more output.

One last thing, there is also a LiIon 18340 option, offering greater capacity, although this size is not offered by AW. Others have more knowledge about this option.

Hope this helps,
K

Edit: BTW my pocket-rocket is the SF E1 / SSC P7 / IMR16340 / McE2S two-stage tailcap, so I'm pretty familiar with the configuration you're talking about. If I was doing it in my 1xCR123 host instead of the E1 host, I'd wait for the XP-G Malkoff M30 (should be bombproof) or the Moddoo / Electron Guru XP-G triple (should be bright) to become available. Although the XR-E?? M30 was a great module (I owned one for a while and liked it a lot - check the link in my sig for more info), there are some new options coming out which are potentially very exciting - however, patience may be a virtue here, neither product has a concrete release date as of yet. ;-)

 

[SilverFox]

Hello Jaundice,

Keep in mind that higher current drains = fewer charge/discharge cycles.

Tom

 

[Kestrel]

Keep in mind that higher current drains = fewer charge/discharge cycles.

This is often written regarding LiCo (i.e. LiIon), but would this behavior be less pronounced for IMR? I guess that there are two different hypothetical situations here:

• Repeated 2C drain on LiCo, these cells exhibit much shorter overall life than repeated 2C drain on IMR cells
• Repeated max C drain on LiCo (2C), will these cells exhibit shorter overall life than repeated max C drain on IMR (8-10C)?

 

[jaundice]

Thanks, guys, for the responses.

Kestrel; Excellent advice. I plan on getting new Malkoff drop-ins once they become available. For now, though, I already own an M30. I guess there's really no right or wrong answer. The output between the two cells is about the same at this discharge rate; the cost of the cells is about the same; it appears to be safe to use LiCos in this case; and finally, I don't use these cells enough to sweat the decline in usage. As I said before, I prefer over-discharge protection of a LiCo, but you've sort of convinced me to go with IMR.

Silverfox;

When you mean higher drain rates, is that linear throughout the published range of acceptable values? As in, would a cell being discharged at a .1 C rate last longer than a cell being discharged at a 1 C rate (both rates being well in the acceptable range)? Also, is there a rule of thumb to indicate how much this declines? So, as in Kestrel's case, a constant, 2C discharge would mean 1/5th the lifespan versus a lower discharge rate?

Thanks,

-John

 

[SilverFox]

Hello John,

Sorry, the decline in cycle life involves more than just the current. It also involves the "mix" used for the electrolyte, and we have no control over that.

When I refer to higher current draw, I mean outside the normal rate of 1C.

And to address Kestrel's question, I have had both LiCo and IMR cells turn to crap after as little as 20 charge/discharge cycles at high discharge rates (about 5C).

Tom

 

[Justin Case]

I measure about 1.5A draw at the tail for my M30 when using 1xIMR26500. When using a freshly-charged, lightly-used AW16340, I get about 1.4A at the tail.

 

[jaundice]

Based on Justin's data, we can extrapolate that the AW16340 is running at >3C.

Does anyone have a discharge/capacity graph for an IMR 16340? Basically, does the capacity of an IMR cell decline under extremely high current like a LiCo cell? Would a 500 ma IMR cell decline to, say, 350 ma, under a 1.5 amp load?

Thanks,

-John

 

[Kestrel]

Based on Justin's data, we can extrapolate that the AW16340 is running at >3C.

Do you mean the AW IMR16340? I'm a little confused by your post. :thinking:
BTW, comparing the behavior of a 16340 to JC's 26500 is sort of like comparing a mouse to a moose. Yes, they are both herbivores, have brown hair, are very good to eat, and are extremely dangerous when wounded. However, one generally has to work much harder than the other and as a result exhibit different performance characteristics.

Does anyone have a discharge/capacity graph for an IMR 16340? Basically, does the capacity of an IMR cell decline under extremely high current like a LiCo cell? Would a 500 ma IMR cell decline to, say, 350 ma, under a 1.5 amp load?

The IMR16340 should be ~500 mAh at a wide range of loads. Here is a post which includes discharge data of the AW IMR18650, it basically holds its capacity up to its rated max discharge rate of 10C, although the voltage that it is able to maintain is lower in the higher-drain conditions.

Again, in the master Malkoff M60 & M30 thread, JC posted data here (in post # 124), showing that the runtime will be significantly less on the IMR cell, although the output will be somewhat greater. You will find a lot of useful information in that thread BTW. As you've said earlier, it's a tough call between the two cells for your application. FWIW mdocod has posted that he now always recommends the IMR over the LiCo chemistry in the RCR123 size these days. Edit: whoops, [...] for incandescent applications., I just came across two good posts of his on this topic (posted below).

 

[jaundice]

Kestrel;

Thanks for the illuminating post. I'm not sure how eating a mouse is, but I'll take your word for it! :sick2: On a more serious note, point taken.

On your advice, I began re-reading that thread last night. I first read it early last year, when I was accumulating various M60s and M30s. Re-reading it again has proved useful, now that I'm a year further down the road to flashaholism. I saw Justin's post, and emailed him directly about it.

I think I'm going to go with an IMR cell, or just primaries. I agree with Mdocod that it appears LiCo 16340s aren't really worth it, except at less than 500 ma draw.

Thanks again for your info. I really appreciate it.

-John

 

[Kestrel]

OK, more info for your reading enjoyment, from one of CPF's two experts:

Regular protected 3.7V LiCo chemistry RCR123s are often rated between 750 and 900mAH capacity. Realistically, they are more like 550-650mAH under most loads. The maximum safe load for these cells is going to be ~1.2A if used in continuous discharge applications. In short bursts, they can be safely used up to 1.5A and higher. In either case, the absolute limits of the cell are being pushed, and the cycle life will be dramatically reduced.

For reliability, cycle life, and safety, I always recommend LiMn chemistry (IMR) RCR123s over LICo chemistry cells in incandescent applications. The IMR cell when operated at that same 1.2-1.5A range, is not stressed to any significant margin. The cells will maintain good cycle life and maintain better average useful capacity and performance over the life of the cell compared to LiCo cells.

IMR16340s are not protected, however, this should not be considered a negative.

The reason regular LiCo cells need to be protected is for safety reasons. When Lithium Cobalt chemistry cells "go-off" (combustion) they produce their own oxygen during the burn, which accelerates it, while at the same time, produce harmful toxic excrement, (hydrofluoric acid in gas and liquid form as I understand). The odds of having this happen are pretty low, but are exponentially increased when the cell is repeatedly abused. (over-discharged, over-charged, or drained too quickly). The protection circuit in your 3.0V (buck regulated LiCo cells) are preventing you from using the P61 as a safety precaution. The P61 would be at least a 4C load or higher on the cells, and the voltage-bucking diode would generate major heat at these discharge levels, which would be further cause for concern.

LiMn and LiFePO4 chemistry cells are often categorized as "safe lithium" chemistry cells. They can handle much more abuse (in the form of fast discharge rates and over-charge/discharge) than ordinary Lithium Cobalt chemistry cells before venting. When they do "go-off," they do not produce their own oxygen to fuel the burn and produce far fewer toxic excrement.

Fast discharging and fast charging and over-charge and over-discharge are all going to reduce the useful cycle life of a LiMn cell, but it is very rare (pushing non-existent) to experience a catastrophic failure with LiMn chemistry cells when used within reasonable perameters. They can be thought of as ALMOST as safe as a NIMH cell (NIMH cells are NEVER protected).

With all of this in mind. The need for a protection circuit is drastically reduced. A protection circuit adds resistance, which will reduce performance in direct drive applications (most incans). A protection circuit is also just another failure point. If the circuit fails, it can render a cell useless very suddenly. By eliminating the circuit on IMR cells, we are actually improving overall reliability.

 

[Kestrel]

On your advice, I began re-reading that thread last night. I first read it early last year, when I was accumulating various M60s and M30s. Re-reading it again has proved useful, now that I'm a year further down the road to flashaholism.

LOL, I had the exact same experience as you did, I began posting in that thread quite a while ago, and a few times since then have re-read it. I've gotten a lot more out of it after obtaining more contextual knowledge from CPF.

I think I'm going to go with an IMR cell, or just primaries. I agree with Mdocod that it appears LiCo 16340s aren't really worth it, except at less than 500 ma draw.

Unfortunately, when I paraphrased his post I had forgotten that he tied it to incandescent use - with incans you really have to drive the 123's hard in the smaller lights. At least with LED's, we can have a usable low level where the LiCo chemistry can demonstrate their higher capacity at the much lower drive levels.

Thanks again for your info. I really appreciate it.

No problem, glad I could be of help. It does take some effort to go into this much depth, I really understand why SilverFox & Mdocod don't seem to write long informative posts as much as they used to - they are a great help as always, but now it's our turn to carry the torch for a while - this stuff can be time-consuming.

BTW, you have read Mdocod's two LiIon guides, right? Pretty much the lithium-rechargeable bibles here on CPF. IIRC they should still be linked in his signature.