MDs Lithium-Ion > Incandescent guide + compatability/comparison chart

ElectronGuru

Flashaholic
Joined
Aug 18, 2007
Messages
6,055
Location
Oregon
Does anyone know whether 2 x AW'S IMR26500 will instaflash a Philips 5761 or might I just about get away with it ...:devil: (2 C Mag - FM Bi-Pin adapter - No resistance fixes).


Tim.



Wit no resistance fixes you *might* get away with it..

My suggestion would be to let the cells rest ~8-12 hours to reduce the odds. In my experience, LiMn chemistry cells will settle from ~4.20V down to around 4.10V after that 8-12 hours resting.

$5 bulb, it's worth a try.



I've been keeping a log of my 26500 testing. This is with stock bodies and freshly charged cells:

2x - IMR26500 (7.4 - 8.4 volts)
3854 ROP Low - GOOD
3854 ROP High - :poof: (0.2 under?)
3853 ROP Low - GOOD
3853 ROP High - GOOD
Philips 5761 - :poof: (0.5 over)
WA 1111 - :poof: (0.1 over)
WA 1274 - GOOD (9.1 limit)
Osram 64250 - :poof: (0.3 over)
Osram 64275 - :poof: (0.3 over)
Osram 64430 - GOOD (10.9 limit)


3x - IMR26500 (11.1 - 12.6 volts)
WA 1331 - :poof: (0.6 over)
WA 1185 - :poof: (0.3 over) <- safe after 24hr rest
WA 1166 - GOOD (14.1 limit)



Comparing my results with mdocod's, 8.2 is still .3 over the 7.9 flash point for the 5761. Perhaps with more time...
 
Last edited:

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
IMR 18500 have just come out and im wondering what are some uses for these cells?

I mean I use IMR 16340 ALOT for lights that normally use Lico large cells...

The IMR16340 cell, since it lacks the wasted space of a PCB, and is such a small cell where a PCB displaces a large percentage of space, is actually on par capacity wise with many RCR123s. So in any case where you have a direct drive LED or incan where you can manually see the point where you need to kill the discharge to prevent over-discharge, the IMR16340 is a great replacement for an RCR123.

but other than LF IMR Lamps, what can I use IMR 18500 for? instead of liCO 18500

Depends on what results you are looking for. In a broad general sense, you can say that for 17500 cells and larger, IMR (LiMn) will have about 1/3rd less stored energy, however, will be able to safely deliver 2-3X the maximum safe discharge of the equivalent size LiCo cell.

Since the LiMn cell has lower internal resistance and no PCB to add additional resistance, some bulbs that work with protected LiCo cells are going to be closer to instaflashing on LiMn cells, especially bulbs that were already really having the **** pushed out of them...

Some people run the P91/MN11/MN16 for example, on a pair of protected 18650 cells. One can, for fun, drop down to a "2 cell" host and run any of those bulbs on a pair of IMR16340s just fine. Since the small 16340 size suffers plenty of voltage sag under the strenuous load, no new bulb related problems arise. However, replacing the protected 18650s with IMR18650s can translate to faster bulb failures, as that configuration was already hard on the lamp, the LiMn has just enough of an edge at this load to really increase the likelihood of popping the lamp.


safer chemistry?

Not only safer, but theoretically should have better cycle life under strenuous loads.

there is no PCB on IMR cells.

Defeats the purpose of low resistance high current cells. If you have an application that falls within the safe boundaries of a protected LiCo cell then the LiCo may be the better way to go. Regulated LED lights especially IMO.

I posted a thread on this question but it was shut down im told this is where to post such questions and MD your best qualified to answer anyway.

You might have a run through AWs LiMn sales threads, the original, when he released the IMR16340s, was heavy with questions of "what why where and how" and I spent the first few weeks of that sales thread helping to answer questions right in the thread.


----------------


MD, what do you think the runtime of 2x14500 and a P90 would be, I hope more than 30 minutes.

Is it true 14500 LiCO are better than RCR16340 LiCo?


14500s usually come closer to rated capacity than RCR123s.

It would have to be a custom machined down body to be worth the smaller diameter cells, in that same length one can run 2x18500s....

Do you have a new body coming out/out already? hehe....

----------------


You can run the P91, the FM 1794 or a WA1111, the same things you use the IMR16340s for, only with the IMR18500 the bulb will be a little brighter and run about twice as long.

I think IMR18500s are likely to blow the 1794, possibly the P91, and possibly the 1111.



--------------------


Thats my question! Why not switch to 18500 LiCo in this setup, it will run 33% longer!

(except for safety.....I see 18500 as a step down in power, but no smaller than another 18500 LiCo...)

however it will run twice as long as 16340 if you do care about using IMR for safety reasons.

Protected LiCo 18500s shouldn't really be used to power up a 1111. In the case of the 1794 and P91, it's probably preferable to deal with the risk of the LiCo cell over the IMR18500 as the LiCo will have a natural voltage sag at these loads that would be less apt to pop the bulbs. The 1794 however is still going to be high risk either way. It's just a screaming bulb that was designed to be on the bitter edge from small cells.

----------------------------------

I've been keeping a log of my 26500 testing. This is with stock bodies and freshly charged cells:

2x - IMR26500 (7.4 - 8.4 volts)
3854 ROP Low - GOOD
I think the low bulb was designed by pelican to have a lot more bulb life on the table as the backup bulb. We see it handle much more overdrive than the high as a result.

3854 ROP High - :poof: (0.2 under?)
I'm actually really surprised here, so many people have run this bulb on 7AA NIMH cells... Very revealing how much voltage sag they must really be getting across the system.
3853 ROP Low - ?
3853 ROP High - ?
The high might be worth looking into, it should have an extra ~1-2V of overhead in the design over the 3864.
Philips 5761 - :poof: (0.5 over)
Shame there isn't a really good way to drive these without all the fuss. Such a great filament shape and everything. I can run it off a pair of old Emoli cells if I rest em over night.
WA 1111 - :poof: (0.1 over)
Poofed on my Emoli 18650s as well.... bleh
WA 1274 - GOOD (9.1 limit)
Hows the CCT look?
Osram 64250 - :poof: (0.3 over)
Emoli 18650s poofed this one too.. bleh
Osram 64275 - :poof: (0.3 over)
Can't say I'm surprised, the 18650s drive the **** out of it... Oh well, not the most efficient bulb anyways right-
Osram 64430 - GOOD (10.9 limit)


3x - IMR26500 (11.1 - 12.6 volts)
WA 1331 - :poof: (0.6 over)
WA 1185 - :poof: (0.3 over) <- safe after 24hr rest
You may find that more bulbs in that list above will survive after 24 hour cell rest. Especially the 3854H.
WA 1166 - GOOD (14.1 limit)


Comparing my results with mdocod's, 8.2 is still .3 over the 7.9 flash point for the 5761. Perhaps with more time...

Thank You for your many bulb sacrifices!

The trick is hoping that you have enough resistance in the switch, the spring, the bi-pin adapter, the contact from the bi-pin adapter to the switch tower, the contacts between cells, etc etc etc. C mags actually have lower resistance switches than D mags straight from the factory, and if it's a brand new flashlight then you have less oxidation on contacts than many people would.

Sometimes there's enough resistance sometimes not....

I was chatting with a friend on the phone, few weeks ago he was trying out his new IMR18650s in some various leefbodies and such. He was able to drive the 1111 and 1185 on the new IMR18650s without insta-flashing. Probably because he's had those flashlight bodies and switches and everything for a couple years now, so there's just enough extra resistance.

-Eric
 
Last edited:

Timson

Enlightened
Joined
Mar 5, 2004
Messages
526
Location
Leeds - England
Electronguru,

Useful numbers.....Thanks for sharing those test results. :twothumbs

That's saved me a few blown lamps. ;)


Tim.
 

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
Eric - thanks for that great post (#182). A wealth of useful tips there, and I for one have certainly learned some new info.
 

Jay T

Enlightened
Joined
Dec 8, 2006
Messages
372
I think IMR18500s are likely to blow the 1794, possibly the P91, and possibly the 1111.


In all my lego playing the only one of those I have blown is the 1794 on some IMR18500s. The others run just fine.

All of my tailcaps are new and all have recieved a spray of Deoxit.

18650s
DSC00473.jpg


+1111 = no blow
DSC00474.jpg


+P91
DSC00475.jpg
 

ElectronGuru

Flashaholic
Joined
Aug 18, 2007
Messages
6,055
Location
Oregon
Thanks Eric,

I'm actually really surprised here, so many people have run this bulb on 7AA NIMH cells... Very revealing how much voltage sag they must really be getting across the system.

The trick is hoping that you have enough resistance in the switch, the spring, the bi-pin adapter, the contact from the bi-pin adapter to the switch tower, the contacts between cells, etc etc etc. C mags actually have lower resistance switches than D mags straight from the factory, and if it's a brand new flashlight then you have less oxidation on contacts than many people would.

Reminds me of my very first hotwire. An ancient 2D (no boring needed) running ROP high off 8AA (yes 8) eneloops in a Sandwich adaptor. There was so much oxidation, sometimes it felt as if it was creaking to life, but it it didn't pop. Until one day, it didn't not pop. It was as if the oxidation had "burned" off and the bulb was finally getting near spec levels of juice.


The high might be worth looking into, it should have an extra ~1-2V of overhead in the design over the 3864.

Yup. I have a set on order and will report in after testing.


Shame there isn't a really good way to drive these without all the fuss. Such a great filament shape and everything. I can run it off a pair of old Emoli cells if I rest em over night.

As noted, I got good results resting with the 1185, but the over limit volt difference is almost double with the 5761. Still, your experience is encouraging. I don't know how similar IMR's are to Emoli's, but if they are - we should be able to apply the same lessons.


Hows the CCT look?

I'm working on a photographic technique to show color temperature. When ready, I'll add profiles for all my bulbs (and a few LEDs) to my incan guide.


You may find that more bulbs in that list above will survive after 24 hour cell rest. Especially the 3854H.

This brings to mind a question: when cells rest, is any energy lost? Specifically, does a cell that is charged to 4.2v and rested to 4.1v have any more energy than a cell that is charged directly to 4.1v? Even so, how tough would it be to change the charging stop point to 4.1?


Thank You for your many bulb sacrifices!

Sure, though it wasn't a grand plan. When the first one popped, I just had to see if the second one would, then the third. Before I new it, I had a small pile of glass trinkets. In some cases, I only had one example bulb, so I'll have to wait (rest!) before trying again.


He was able to drive the 1111 and 1185 on the new IMR18650s without insta-flashing.

Frustrated with the battery life from 3xIMR16340, I swapped the 1185 for a 1331 and am super happy. I'm looking to do the same thing for 2xIMR16340. I have a 1274 on the way to replace the 1111.


:thanks:
 
Last edited:

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
I don't know how similar IMR's are to Emoli's, but if they are - we should be able to apply the same lessons.

Emoli cells are made by Molicel Canada. They were, as I understand it, the first to come to market with a viable LiMn cell. Which also means that they soaked up a lot of the R&D costs while most other manufactures piggy-backed on that design and had cash-flows remaining to improve on the design. So actually, many of the other brands of LiMn chemistry cells, including AWs, are of higher capacity and as good or better performing under a load.

Molicel was the original implementer of the prefix "IMR" to define a LiMn cell.

The Emoli cells I have were actually manufactured back in 2005! They still work quite good though, just probably not as good as AWs, which is actually a good thing, as they don't pop bulbs as easily. Probably why I am getting away with driving an MN21 off a pair of them rested and others are popping MN21s on AWs cells.


I'm working on a photographic technique to show color temperature. When ready, I'll add profiles for all my bulbs (and a few LEDs) to my incan guide.

Sweet :)




This brings to mind a question: when cells rest, is any energy lost? Specifically, does a cell that is charged to 4.2v and rested to 4.1v have any more energy than a cell that is charged directly to 4.1v? Even so, how tough would it be to change the charging stop point to 4.1?


I believe that the natural tendency for LiMn cells to settle to ~4.10V probably does represent a loss in stored energy. I am not sure how that would compare to a cell charged to 4.10V and then immediately discharged. Interesting question, might be a fun experiment for someone with precision battery discharge testing equipment to try out some day.

Modifying an existing charger to have a lower termination voltage could be as simple as changing a passive component that tells the charge controlling IC what termination value to aim for. Such a modification would only work on some chargers and not on others. It's also theoretically possible to use some sort of diode on the output of the charger to have a voltage shunting effect.. However, this is getting into an arena that's a few hair-do's over my head. There are hobby chargers out there that have manual adjustable termination values for their li-ion charging algorithm, that would certainly be the easiest way to go about it.


-Eric
 

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
Quote:
Originally Posted by Nite
MD, what do you think the runtime of 2x14500 and a P90 would be, I hope more than 30 minutes.

Is it true 14500 LiCO are better than RCR16340 LiCo?

MDOCOD
14500s usually come closer to rated capacity than RCR123s.

It would have to be a custom machined down body to be worth the smaller diameter cells, in that same length one can run 2x18500s....

Do you have a new body coming out/out already? hehe....


Well, MD, Since you asked Ill assume its safe to answer. Yes I am half sold out of a FiveMega 1,2,3xAA body SF C head host..with momentary built in clicky.
Its a special edition limited run, only thing ive done outside the standard FM 18mm bodies and extensions.

My OWN personal light that you are asking about. with a G3 nitrolon head...
LIGHT02.jpg


You may have missed it as I know you dont hang out at CPFM where I just moved BTW, as much as you do here.lovecpf
 

Jay T

Enlightened
Joined
Dec 8, 2006
Messages
372
:tsk:

A 14500 has a rated capacity of 750Mah.

The Max discharge for these cells is 2C.

2 X 750 = 1500.

Thus the Max currect draw is 1500mA or 1.5A.

From the Lumensfactory specs page the EO-4 draws 2.35A

So :tsk:
 

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
:mecry:
:oops: I just killed an Unprotected but old...14500....

I wont trust it now....

was getting old anyway.


Must recycle it...:caution:
 

Jay T

Enlightened
Joined
Dec 8, 2006
Messages
372
the single cell model should only use an LED?

You might be able to use a Strion bulb in a FM bipin D26, HOWEVER, this bulb will draw 1.67A. Just a bit over 2C.

So before concidering this you must go back to post one and read the paragraphs following "Lithium-Ion cells become more dangerous as they age."


Edit to add:

Not too sound mean, but, your questions are answered in post #1, please review the information there.
 
Last edited:

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
a good link I got from LuxLuthor

some info:
http://www.carleylamps.com/faq.htm

Mean Spherical Candlepower - MSCP (or CP candlepower for short) is the total light output of a lamp in all directions.
Foot Candles - Total light output of a lamp when the light is compressed by a lens or reflector. Foot candles is a "directional specific" light measurement.
Filament Temperature - The operating temperature of the filament measured in degrees kelvin.
Lumens = MSCP x 4. 1 MSCP=12.57 Lumens

thanks Lux

I read Post 1 all the time, but anyone know offhand whats the smallest cell a LF EO-4 Can use..1x18500?
sorry if answer is in the chart....Nevermind, found it. I was right, 1x18500: LF D26 EO-4: 9W, 120 - 67 lumen in 34 minutes
 
Last edited:

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,654
Location
MS
You're welcome. This was prompted from my trying to find out the actual specs on the FM1794 custom made Carley bulb which Nite is going to send me to test. He was told by FM that this bulb puts out 50 MSCP at 7.0V & 3A (21 Watt bulb).

As I added in my PM to you, their statement on that linked page is not accurate, regarding their equating MSCP to CP:
[FONT=Arial, Helvetica, sans-serif]Mean Spherical Candlepower - MSCP (or CP candlepower for short) is the total light output of a lamp in all directions.[/FONT]
Candlepower is only a cone shaped beam of light in one direction. Carley should know better than to make such an error. Once you measure the total light output in all directions (in an "Integrating Sphere"), that is also how "Lumens" are determined, so conversion between an accurately measured MSCP and Lumens is possible.

Again, I highly doubt that Carley would go to the expense for a custom order bulb to put it through an expensive, calibrated Integrating Sphere test to certify the MSCP (or lumen) rating. It was a number of these claims by various bulb manufacturers that prompted me to do the independent destructive testing comparisons.
 

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
Confirmed...

Even with an AW soft Start switch in use....2x18500 IMR will flash an FM1794 as soon as 100% is activated!

I think I got that crazy Idea from JayT.

MDCOD was right!

Unless someone can tell me why I need IMR 18500 for anything but LF IMR bulbs I wanna know...

18500 IMR are a great substitute for 18650 IMR as they take up less space in my Cars EDC, a 1000 LF lamp.
However they are NOT a great substitute for IMR 16340 as they wont sag as much.

Man this post just cost me 10$

and I just sent 2x FM-1794 to LuxLuthor..my hobby grows more expensive by the day!

It will be worth it....Im sure many here including myself are aching to see the destructive incan tests run on those FM-1794..halogen..

thanks in advance, lux.
 

Jay T

Enlightened
Joined
Dec 8, 2006
Messages
372
Confirmed...

Even with an AW soft Start switch in use....2x18500 IMR will flash an FM1794 as soon as 100% is activated!

I think I got that crazy Idea from JayT.

What was your voltage?
Were the cells hot off the charger, or did they sit a while?

It seems as if these cells act more like Nimh rather then LiIons and they relax a bit once you take them off of the charger. The way I charge my cells is I take them off the charger where they sit until I carry them to the where the flashlight is. I don't really need a bright flashlight at that exact moment so I don't sit there waiting for the light to go green. When I get around to loading the light it gets loaded. Could this be why some others are having trouble that I am not?

It would be nice if when someone blows a bulb they would post some real information about the voltage and the cells state of charge.

I'm sure you have heard of a build called the mag85, it's a popular build traditionaly done with 9AAs. The thing is that a good build will blow the bulb hot off the charger. I blew several 1185s before I learned to take my time and not hurry. I'm sure that the total number of 1185s blown would fill a nice little box. Does that confirm that the Mag85 is a bad idea?

Here is a little video I just made featuring the IMR18500/1794/AW soft start combo and a 9 eneloop mag85. The eneloops just came off the charger and were still warm.


http://www.youtube.com/watch?v=oKAVoQn3170

Edit to add: Disclaimer - bulbs were slaughtered in the making of this video.
 
Last edited:

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
ahhh help

Advice needed

Just over-discharged a pair IMR 16340 while talking to FM, i didnt notice the time pass...

despite stepping down brightness on my soft start I got two imr 16340 down to, after a short 20 min rest..to measure 2.96 volts and 3.12 volts...before changing cells.:oops:

I am recharging them immediately.

do I have to trash them?:confused:
lovecpf

I want MDOCODs opinion on this, and Maybe AW as well.

FM said recharge them see if they come back..

im guessing they lost capacity but wont explode and are more tolerant to this kind of abuse.
 
Top