My 123 Test! Tekcell vs. Surefire

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
The other day I received my shipment of 100 Tekcells from Countycomm. Nice deal at $75 shipped! But when they arrived I noticed what looked like an orangish discoloration like some sort of leakage around the base of a few of them, and this prompted a test. Being sort of a thorough individual, my little test quickly became an afternoon of detailed examination worthy of inclusion in Consumer Reports! I had better things to do with my time, but the results were certainly interesting. Oh well. /ubbthreads/images/graemlins/icon3.gif

My single big insight on that day was that 123 cells behave a little differently from the alkalines we all know and love; namely, an alkaline cell's voltage will steadily decrease during its lifetime, while even a heavily depleted 123 might still deliver 2.8-3.0V. So while alkalines can be checked quite nicely by examining their voltage with a DMM, 123s cannot.

What does happen to a 123 is that its maximum current flow capability steadily decreases as it runs down, so a test of momentary unloaded current flow (let's call that MUC throughout the remainder of this article) using a DMM yields what appears to be very nice information about the condition of any given 123. Before getting to specifics in my test, note that this measurement can be taken using a DMM set to a high current range, by touching the probes to the terminals on the cell and holding *just* until the peak current reading appears. You'll see the current reading jump upward quickly, peak at some point, then gradually begin dropping. All this takes a fraction of a second, after which REMOVE THE PROBES! Longer than that could damage the 123 cell.

As I started going through my 100 Tekcells measuring MUC for each, I came to 'bin' the cells based on their readings. I assigned bins at each amp boundary, so 7.0-7.9A would represent one bin, 8.0-8.9A would represent another bin, etc. Here's what I found for my 100 Tekcells as shipped:

Bin (number of 123s)

0.x (2)
5.x (3)
6.x (4)
7.x (33)
8.x (58)

Given these results, I established 7.0A as the minimum MUC that indicated a 'fresh' 123, and as seen by my results, variance among Tekcells was only fair, with several not shipped fresh and two dead as a doornail!

Now I was curious, so I measured my other 123 cells, which consisted of 18 Surefires and 2 Sanyos. Here's how they binned:

Bin (number of 123s)

9.x (20)

So the first conclusion I drew was that Surefire cells really do give you something more for your money! In two ways, really; they ALL binned higher than the highest Tekcell, and the variance was MUCH tighter, suggesting stricter quality control. As for the two Sanyo cells, the results certainly were encouraging, although I don't feel comfortable drawing a conclusion from a sample size of two.

So in the first part of my analysis, I concluded that I didn't get all I bargained for with my Tekcells. But did I get enough? Let's look at the economics involved.

The Tekcells came with no shipping charge, and if I buy enough Surefires at once, they also come with no shipping charge. So looking strictly at the cost of the batteries, the Tekcells cost $0.75 each, whereas the Surefires cost $1.25.

For my money, I really only got what I'd classify as 91 'fresh' Tekcells, so they effectively cost me a little over $0.82 per 'fresh' cell.

Now being a little severe on the Tekcells (you'll see why soon), I'll estimate that I can only depend on getting 7.0A MUC from each Tekcell, whereas I'll assign 10.0A MUC to each Surefire cell. Let's further assume that MUC is directly proportional to the usable power capacity of a 123 cell.

This means those 91 fresh Tekcells really cost me $1.18 for the same amount of power capacity as I got from each $1.25 Surefire cell. This was calculated as $75 divided by 91 Tekcells, times 10A/7A (to account for the higher power capacity available in the Surefire cells). Even with the bad cells and the lower binning of the Tekcells, they still seem to be worthwhile.

So who wins? Well both, sort of, but it depends on your needs. If you simply want the most power capacity for the lowest unit cost, go for the Tekcells, which for all their problems still price to about $1.18 for what costs you $1.25 from Surefire. On the other hand, if you're going for the longest runtime before having to reload your flashlight, or for the greatest likelihood that the next cell you grab isn't a dud, go for the Surefire. It makes intuitive sense that the price you pay for these niceties is an additional $0.07 per cell for the Surefire 123 vs. the Tekcell.

Each person will have their own reasons for buying, but now, at least with these brands, maybe you can make an informed choice. Hope this helps someone out there. /ubbthreads/images/graemlins/smile.gif
 

paulr

Flashaholic
Joined
Mar 29, 2003
Messages
10,832
You could also try the DP cells that Arc and Batterystation are selling for $1.00 each. They have tested pretty well.

You have to also remember that Surefire cells are supposedly optimized for very high current drains (P61, M6 HOLA, etc). If you're running lower powered lights, high current capacity may not matter.

Other people have tested the Tekcells and they're indeed not so great. The DP's are supposed to be better. I have some Tekcells that I'm using in my Arc LS and they work fine (relatively low drain) but I'll probably stick with DP or Surefire after I use my remaining Tekcells up.

What you're calling MUC is actually the short circuit current, by the way. It's not "unloaded", it's maximally loaded.
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
Scott_
Nice testing. I enjoy reading about such investigations. Your "MUC" parameter actually has a real name in the battery industry which is "flash amps".
Your assumption "Let's further assume that MUC is directly proportional to the usable power capacity of a 123 cell" is probably a fair assumption as written but from context I suspect that what you meant to say was "Let's further assume that MUC is directly proportional to the usable *energy* capacity of a 123 cell".

If my guess about your intended meaning is correct, I would say that assumption is going pretty far out on a limb.
 

e=mc²

Enlightened
Joined
Oct 2, 2000
Messages
537
Location
NJ - Land of malodorous \"earl\" refineries!
Doug_S: What would be an ideal setup or test for maximum energy capacity of a cell? Also, what params would have to be gathered during the course of the testing? I would assume that it would involve some sort of "runtime" test, maybe plotting the results (amps x volts) and then somehow integrating to find the area under the curve. I have numerous different "fresh" cells I would like to analyse. I do have plotting capability as well. Any suggestions would be greatly appreciated....

[added]

The only problem I forsee is the integration of the curve since it would be futile trying to arrive at a mathematical expression which accurately represents the curve as it would probably vary from mfr. to mfr. . I suppose I could make the sampling rate very frequent so I could "approximate" the area, that is, if this is the correct approach to calculate the energy capacity of the cells.

Ed
 

ChrisA

Enlightened
Joined
Feb 10, 2002
Messages
369
Location
Germany
Awesome testing Scott ! Thanks a bunch for that MUC idea. That's a nice and easy way to compare different manufacturers. Now I have to go and see how well my stash of Panasonic, Ansmann and Varta fairs...

Chris
 

MR Bulk

Flashaholic
Joined
Aug 12, 2002
Messages
6,059
Location
Hawaii
Milky, IMHO I think the Tekcells are actually costing you More than the SF's. This is because each time you receive a shipment of Teks, you will have to go through this whole thing again. Time spent sorting out the "good" Tek cells is actually money (poorly) spent. With most SF 123's (and I say "most" because once in a great while a SF may turn up bad too, although certainly not on the scale of your shipment of Teks), I can simply open a fresh box and use'em, no time-consuming testing and sorting required.

Just my 2¢...an' proud of it!
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
MR Bulk, it doesn't cost me anything to test 'em if I don't have paid work to do at the time! /ubbthreads/images/graemlins/grin.gif

Seriously, I see your point of course. I did the test this time as a learning experience, and do feel like the time was well spent. And once you get the hang of putting each cell on the probes, the whole process goes surprisingly quickly... maybe 10 minutes to get through all 100? Test, toss... test, toss... /ubbthreads/images/graemlins/icon3.gif

I also have a harebrained idea for making testing quicker, involving putting all the cells upside down in some sort of metal pan, clipping one probe to the pan itself, and just moving the other probe from battery to battery, poke, poke, poke... /ubbthreads/images/graemlins/smile.gif

Now regarding costs, that's part of what I was getting at in terms of what your additional $0.07 per battery buys you by going with Surefire. If your 10 minutes is worth at least $7, just buy the Surefire instead. It's a personal decision.

That said, there's no guarantee that all your Surefires will be good in all cases; we've seen plenty of reports here on CPF about people receiving bad batches of SF 123s. For that matter, it's entirely possible that every once in a while a cell that tested great might still fail for some reason once in your flashlight. So you could take the approach that you just grab some, pop 'em in, and if the light doesn't work, start over.
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
[ QUOTE ]
paulr said:
What you're calling MUC is actually the short circuit current, by the way. It's not "unloaded", it's maximally loaded.

[/ QUOTE ]

It's just an issue of semantics. The 'load' I'm talking about is a resistive load, so 'unloaded' to me means no resistance in the circuit. But I agree that without further clarification, my terminology could be confusing.

Hopefully what I meant was clear by the context in which it appeared.
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
[ QUOTE ]
Doug S said:
Scott_
Nice testing. I enjoy reading about such investigations. Your "MUC" parameter actually has a real name in the battery industry which is "flash amps".
Your assumption "Let's further assume that MUC is directly proportional to the usable power capacity of a 123 cell" is probably a fair assumption as written but from context I suspect that what you meant to say was "Let's further assume that MUC is directly proportional to the usable *energy* capacity of a 123 cell".

If my guess about your intended meaning is correct, I would say that assumption is going pretty far out on a limb.

[/ QUOTE ]

Doug S, First of all, many thanks for the compliment. /ubbthreads/images/graemlins/bowdown.gif

Now regarding my possibly flawed assumption, I'm hesitant to get into much of an argument here because frankly, I think you're likely to kick my butt. /ubbthreads/images/graemlins/grin.gif However, based on real world observations, I still believe the MUC (or "flash amps"... thanks) is a reasonable indicator for the remaining energy in a 123 cell. This comes with the disclaimer, though, that I don't think the same could be said about batteries in general. In the case of the 123, the relative stability of the output voltage, I think, makes my assumption more true than it might be elsewhere.

Also, in practice what I've seen is that a drop in MUC does seem to occur in conjunction with the life cycle of the 123 cell; I'm seeing fresh cells in the 7.x-9.x range depending on brand, then testing around perhaps 6.0A after some usage, and testing around 1.0A or less just at the point where the cells can no longer drive my Space Needle II at more than a trickle.

Meanwhile, even when the cell tests at 1.0A, it's still delivering something like 2.98V.
 

MR Bulk

Flashaholic
Joined
Aug 12, 2002
Messages
6,059
Location
Hawaii
Hi Milky,

Yeah, I guess it's all a matter of self-imposed time constraints. Mine are self-imposed because I elected to take on mod projecys, and after my job, child-rearing duties, and some limited "play" time with friends and co-workers, I really don't have yet more time to come home and test batteries! Add to that the fact that I must test each and every Luxeon when I embark upon any mod project, and I think you can see the very real value to me personally in being able to simply pop'em in and use'em.

But this is weighing things against batteries with a *slightly* iffy track record. Now DP's at a buck each on the other hand...
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
[ QUOTE ]
milkyspit said:
I still believe the MUC (or "flash amps"... thanks) is a reasonable indicator for the remaining energy in a 123 cell.


[/ QUOTE ]

Scott_
I think that you are quite right about this. This is a *much* better test than trying to estimate remaining capacity by measuring voltage. The part of your earlier statement that I meant to caution against taking too seriously was the "directly proportional" . An interesting test that you may wish to do sometime is to discharge a cell in equal increments and check the flash amps after each increment. You can then build your own table for the correlation between flash amps and remaining capacity for that particular cell type and brand. I would strongly caution against drawing conclusions about relative energy capacity by comparing flash amps between different brands. Each manufacturer makes their own choices in balancing the various design parameters in their cells. It is quite possible that they may have to accept lower capacity to get higher flash amps. This is certainly the case for NiCd cells. As a side note, be aware that flash amps are somewhat temperature dependent [will measure higher at higher temps].
 

Doug Owen

Flashlight Enthusiast
Joined
Jan 30, 2003
Messages
1,992
Scott,

What the other Doug just said!

I suggest you're in just the right spot to do a very graphic A/B test. Put one of each in your candle! This way you can draw them both down at the same 30 mA and put some real numbers on remaining energy.....

Doug Owen
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
[ QUOTE ]
e=mc² said:
Doug_S: What would be an ideal setup or test for maximum energy capacity of a cell? Also, what params would have to be gathered during the course of the testing? I would assume that it would involve some sort of "runtime" test, maybe plotting the results (amps x volts) and then somehow integrating to find the area under the curve. I have numerous different "fresh" cells I would like to analyse. I do have plotting capability as well. Any suggestions would be greatly appreciated....

[added]

The only problem I forsee is the integration of the curve since it would be futile trying to arrive at a mathematical expression which accurately represents the curve as it would probably vary from mfr. to mfr. . I suppose I could make the sampling rate very frequent so I could "approximate" the area, that is, if this is the correct approach to calculate the energy capacity of the cells.

Ed

[/ QUOTE ]
Ed, it pretty much sounds like you know what to do; record volts, amps, and time and then do a piecewise integration. I have done quite a bit of this type of testing. By far the easiest method is to build yourself some constant current loads from LM317 ICs for various currents of your chosing. Be aware that these will need a minimum of 3V across them to maintain regulation so you will need to operate a power supply in series with your cell under test. Since the current is fixed, the only data you need to record is cell voltage at regular intervals. While I encourage you to test to your heart's content, be aware that the battery manufacturers do this testing too and the data is readily available on their websites.
Your asked about maximum energy capacity. For almost all cell types this is obtained at very low discharge rates. It is worth a look at the Duracell 123 datasheet
Duracell 123
Look at the graph of available Whr vs discharge rate. Note that this is for constant power discharge which is a more complicated test than what I propose for you to do. The highest total energy is available at the lowest discharge rates.

EDIT It's data like what is shown on this graph that makes me cringe when people start talking about trying to deliver 5W to a Luxeon from a single CR123 cell using a 60% efficient convertor. /ubbthreads/images/graemlins/icon15.gif
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
[ QUOTE ]
Doug Owen said:
Scott,

What the other Doug just said!

I suggest you're in just the right spot to do a very graphic A/B test. Put one of each in your candle! This way you can draw them both down at the same 30 mA and put some real numbers on remaining energy.....

Doug Owen

[/ QUOTE ]

Doug, thanks. /ubbthreads/images/graemlins/smile.gif Sounds like a test I'll have to do, eventually. I assume you mean measure how long each type of cell runs the Milky Candle until it falls out of regulation? Or perhaps I'm just missing the point entirely, as I'm occassionally apt to do...
/ubbthreads/images/graemlins/icon3.gif /ubbthreads/images/graemlins/grinser2.gif
 

Shadows

Newly Enlightened
Joined
Nov 10, 2002
Messages
72
Location
Singapore
er.. correct mi if i'm wrong... but i thought lithiums can keep 10 years IF untouched... once you use them thou they will drain faster /ubbthreads/images/graemlins/jpshakehead.gif ... so 120 123s....??
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
[ QUOTE ]
Shadows said:
er.. correct mi if i'm wrong... but i thought lithiums can keep 10 years IF untouched... once you use them thou they will drain faster /ubbthreads/images/graemlins/jpshakehead.gif ... so 120 123s....??

[/ QUOTE ]

In the past, there has been some discussion here by CPF members who I regard as thoughtful and knowledgible that there may be a theoretical basis for this. I however have not seen anything from the battery manufacturers to support this concern. In the absence of any data confirming that this is a real problem, I personally would not worry about it.
 

Ginseng

Flashlight Enthusiast
Joined
Feb 27, 2003
Messages
3,734
MilkyScott,

I don't have any way to measure the cells I buy, but I can assign another cost to buying the Tekcells. There is 2% chance of having a dead cell in a batch of 100. If I use these exclusively in my 3-cell SF9P, then there is at least a 1/33 chance of loading in a set of batts with 2 dead cells. There can be up to a 2/33 chance of loading in a single dead cell. Thus, the failure rate jumps to 3% and 6% respectively.

I won't do the math for the low-cap cells, but 5.x and 6.x cells would certainly result in sub-spec performance in my lights. Thanks for doing this very informative piece of work.

Wilkey
 

Doug Owen

Flashlight Enthusiast
Joined
Jan 30, 2003
Messages
1,992
[ QUOTE ]
milkyspit said:
[ QUOTE ]
Doug Owen said:

I suggest you're in just the right spot to do a very graphic A/B test. Put one of each in your candle! This way you can draw them both down at the same 30 mA and put some real numbers on remaining energy.....



[/ QUOTE ]

Doug, thanks. /ubbthreads/images/graemlins/smile.gif Sounds like a test I'll have to do, eventually. I assume you mean measure how long each type of cell runs the Milky Candle until it falls out of regulation? Or perhaps I'm just missing the point entirely, as I'm occassionally apt to do...


[/ QUOTE ]

Not quite, I wasn't clear enough. I'm suggesting you start with two 'new' cells, one of each type, in your candle. At the same 30 mA level, drain them both down until one goes flat (and the candle drops out of regulation), this gives you the lifetime number of the weaker cell. Then replace it (leave the other used one in) and restart. Add the time the second gets before it dies and you know the relative capacity of both brands. You can say 'brand X has 20% more capacity than 'brand Y'.

Doug Owen
 

radellaf

Flashlight Enthusiast
Joined
Apr 10, 2002
Messages
1,098
Location
Raleigh, NC
In playing with my new A2 and having killed two sets of Sanyo cells in it so far, which didn't take long so I presume they were used more than I thought in the Inova X5s they were in, I noticed that the voltage open-circuit was misleading. The new cells read 6.5 or so and the old ones 5.5. Both old ones, I have one set that is really dead, but still runs the X5, and the other set from the A2 just as it drops out of regulation.

I notice that a low draw, 100-250mA that I've seen from the X5s, the voltage goes down a bit (half volt or so) and then holds there, which is why the X5 still works even on the "really dead" cells.

Loading a pair with a MagNum Star (~800mA) lamp, the voltage on the really dead ones drops quickly to about 1.5 before plateau-ing out. The just-out-of-A2 cells drop to about 4.4 volts over 5-10 seconds. After that the voltage does creep down, but very slowly. New cells drop from 6.5 to 5.8 or so and hang there.

Just speculating here, you might be able to check the cells under a less extreme load (and thus discharge them less in testing) than with the sub-ohm value of your meter.

Under a certain load, I agree, it's hard to see if they're old or really old. Above that, though, the voltage on the old ones falls off a cliff. Now I see why the A2 will still light for 5 sec or so with the cells that dropped out of regulation a minute before. It takes about that long for the voltage to drop past 4.5 or so.
 

this_is_nascar

Flashaholic
Joined
Mar 29, 2002
Messages
8,341
Location
Gloucester, New Jersey
[ QUOTE ]
paulr said:
You could also try the DP cells that Arc and Batterystation are selling for $1.00 each. They have tested pretty well.

You have to also remember that Surefire cells are supposedly optimized for very high current drains (P61, M6 HOLA, etc). If you're running lower powered lights, high current capacity may not matter.

Other people have tested the Tekcells and they're indeed not so great. The DP's are supposed to be better. I have some Tekcells that I'm using in my Arc LS and they work fine (relatively low drain) but I'll probably stick with DP or Surefire after I use my remaining Tekcells up.

What you're calling MUC is actually the short circuit current, by the way. It's not "unloaded", it's maximally loaded.

[/ QUOTE ]

Those DP cells suck. First off, they come in putting out only 3.24 volts maximum, whereas SureFires put out 3.26/3.27 volts out of the box. A bigger issue to me is the runtime (even in an ARC LS) of those DP cells. In my experience, the SureFire cells yield approximatly 20% more runtime. There's something to be said for the "you get what you pay for" phase.
 

Latest posts

Top