Rechargeables in the Gladius - The Final Word

In my quest to obtain free lumens for my Gladius, I have spent and wasted nearly $100. These are my stories...

Batteryspace.com - 0300 Hours 15 December 2005

Late at night or early in the morning, I was typing away, researching and hunting for "free lumens." Little known to me at the time, harnessing the sweet power would not come easily. That fateful night I found myself scouring in the darkest corners of the internet, hoping to see 85 lumen lights at the end of the tunnel. I purchased some RCR123As with a charger from Batteryspace.com, and waited days upon days for the $40 package to arrive. Once it did, I glanced inside the box, hoping to see the little Lithium-ion cells that would end my lifelong quest! I charged up the batteries, but there was something in the air... I dove for my multi-meter, and tested the cells! I found only what I was fearing, that the so-called 3.0V regulated cells were indeed not regulated! The 4.2V monster was among us! I could not place these in my dear Gladius, for the 7V circuit limit would not have it! Chapter one of my hunt had ended unsuccessfully, and I was hoping that the 4.2V monster would never again be seen, but I was most definitely mistaken...

e-lectronics.com - 1400 Hours 27 December 2005

With more rest, I expected success. I typed as I griped, and came upon an interesting scene! I looked and looked, and found something clean! I ordered more cells, that were supposed to work! On the website it did say, will work with Gladius! I bought them and then got them soon after, however the cloud of distrust was still above my head! Again I dove for the meter, expecting 3.0V! But what did I see? Oh my! 4.2V! The monster lunged, trying to destroy my beloved $250 light, but I restrained it! I slayed the dragon! Those cells would not enter my light!

So I bought some primary batteries.

THE END!!

You can use a single cell, like a 17670 Li-Ion. The light will blink, but it will be fully functional otherwise. The blinking isn't really that annoying to a lot of people, myself included. I never minded.

Scott

Joshua,

Either of these cells should work fine, and will allways be less than 3.4V each, installed in the Gladius. The problem is measurement technique. The "regulated" cells need some kind of minimum load (about 10 micro amperes) on them to regulate. No load (<1uA) equals no regulation, equals ~4.2V fully charged, as you have found. A 10 MOhm digital volt meter by itself only presents a 0.42 MICRO ampere load.

Try this: Lick your thumb and forefinger so they are really wet, and use the wet thumb and forefinger to press the metal DMM probe tips on to the cells. See the voltage fall .25V or more, vs. holding the insulated test leads? See it vary, wet skin, vs dry skin? You will NOT see this effect with a standard unregulated Li-Ion, or a fresh primary lithium cell.

Powered off, the Gladius loads the cells at about 8-9uA, regulating these cells to below 3.4V per cell or less, IME. Running, the cells are 2.7-3.2V, depending on their state of charge, 20% to 100%:
http://candlepowerforums.com/vb/showthread.php?p=1194965#post1194965

To get an accurate idea of what the cell voltage would be in the Gladius, place a ~330K Ohm resistor ( the exact value is unimportant) across the input of your DMM when you take the voltage reading. The cells will see a ~10uA load (~330K Ohms/3.3V) and almost "fully" regulate to about 3.3V each, at full charge. A new lithium primary is also about 3.25V, light to no load. Try it! Or, simply try measuring the voltage on your regulated cells with an old passive analog meter, it you can find one. You will see 3.2V, because of the light load of the passive (resistors and a coil) meter's loading effect.

Sorry you are having such a tough time. I know it can be scary running a $250 life saving (or life taking) tool on the edge of the unknown. The operation of the regulated cell's circuit is not obvious.

NextLight-

I checked the voltage of the batteries (which have been sitting around for about 2 weeks out of the charger) with a 100K ohm resistor that would simulate about 36.7uA of resistance, which is much more than the Gladius puts on when off. I did this test with both kinds of batteries, and the voltage on the batteryspace cells went down to 3.7V and the voltage on the e-lectronics cells went down to 3.68V. This is still far to high for the Gladius to handle, even if it put over three times the load on the cells.

-Josh

Wow, I am shocked! I wonder if the vendor(s) have changed the circuit in the batteries? I did characterize my several sets carefully before I used them in my Gladius.

It just occurred to me that maybe the Gladius has changed too; (though that would not account for your DMM results above) My SNs are in the link I posted earlier, and I have used the cells in a 0601X light with the new Black HA-III and larger diameter battery tube, but I don't remember right now if I measured it for standby current. EDIT> (Yes, I did) <EDIT There HAVE been several changes in various Gladius components recently. I still have 066XX Gladius on loan (which I need to return) I'll measure it for standby current on standard primaries this weekend.

I will check my cells again from scratch, as if I did not know anything about them, and report here. I may end up ordering a new set to see what I get. I definately don't want to be touting this solution if it no longer works CONSISTIENTLY.

Are your Batteryspace 3.0V cells the older and common baby blue on white label that has been around so long, or the new orange-brownish on white labels (used on the special 3.0V cells?)

EDIT> Anyone else have a set of these regulated cells, a DMM and 100K resistor to compare to Joshua's and my results? <EDIT

NextLight has it right the 3.0 batts need to be placed under load to activate the
regulation. I bent a large paper clip to hold a discarded bulb and touch the
negative end using the bulb (3.6 volt krypton) to load the cell. Using this all sixteen of my (Batteryspace.com powerizer brand)
cells test under 3.28 and most are under 3.2. And yes unloaded my cells read well over four volts. I
do not have a Gladius but I have used the Powerizer cells in a Pelican 3watt,
Surefire U2, SF kl3 on a 6z, a 3 123 2c magmod w/kpr118 (really bright),
Streamlight 2l 3watt, both 3 and 5 watt TnC lights, Electrolumens blaster six,
a E-bay cheap 5 watt pobably Nuwai. I'm probably forgetting a few but I have
not cooked one yet through many charge cycles YMMV though. If I had a Gladius I would'nt hesitate to use these.

Dave

Dave, do you have a 100KOhm resistor to use for a minimum load while measuring voltage with a DMM?


OT, but my 3.0V Powerizer cells have also worked well in several LED lights. The do nstaflash a MN03, however.

This is a discharge curve of the 3V regulated and protected R123 vs Panansonic CR123 under 300mA load :

I do realize that these batteries must be under load to regulate, and knew this all along, however the Gladius is a unique light and even when batteries are in the light, even when the light is not on, they are in physical contact with the circuit.

The Gladius uses a magnetic switch, unlike many other lights. Most lights turn on when the user presses a button or switch, physically connecting the battery to the circuit. The Gladius, however, is always in contact with the batteries and only turns on when the magnetic switch sends a software signal to the light to turn on. When the light is off, it requires a little bit of power to be able to notice when the button is pressed, apparently about 8-9uA. But, as we have figured out, this is not enough to kick in regulation in the cells. So, although it seems to make little sense, just putting these batteries in the light could destroy it, whether you turn it on or not. In fact, it would be better for the light to be on as it would force the regulation in the cells to activate.

However, I thought that perhaps this 7V circuit limit applied not to what the batteries are always in contact with, but only what receives power once the light is activated and the regulation kicks in (if this were the case, putting in almost any RCR123A would be acceptable). So, I checked in with Ken Good, and he told me that this was unfortunately untrue, that every part of the circuit in the light had a limit of 7V.

So, my friends, it seems that we need some kind of cell that will regulate itself entirely even with a load of only 8-9uA.

As for the Pila type batteries, I considered them, but decided that two flashes every fifteen seconds would be quite distracting, not to mention the fact that you would have to explain it to everyone you'll ever show your to.

Also- I have the blue and white powerizers.

-Josh

Cheers NextLight !!

That was a very concise and helpful explanation of the situation. I have yet to try these cells but had read that the open circuit cell voltage is identical to 3.7V cells. I had been reluctant to try these cells because of their lower capacity but it seems that for certain 6 volt lights, LED in particular, they would likely perform better (i.e. full brightness) compared to a single 17670 (Pila 168S) cell. I usually use the 2x 17500 solution for 9-volt lights calling for 3 CR123s, but notice the 17670 just dosent quite cut the mustard when substituting for 6-volt lights.

I can understand the reluctance of risking damage to an expensive light like the Gladius, but I really think the 3.0V batts are safe to use. The only light I'm afraid to substitute a rechargeable cell in is my vintage SF L1, the one that started my LED obsession. Lighthound and others have said its a no-no, although I've run my 2-cell Nuwai Q-III with R123s for dozens of cycles without a hitch and its also on the "forbidden" list. All that aside, I would have faith in the regulation circuit and hope to get some decent runtime from the Gladius. NextLight's key point is the voltage dropping immediately under load, so its not as scary as it might seem. Thanks NL!

As much as I would like to believe that using these batteries is okay, all my tests prove otherwise. Using a DMM and and a resistor is a very accurate way to test these types of things out, and when they give you results, you have to pay attention to them; if the regulation doesn't kick in with a resistor placing three times the load on the batteries than the Gladius does, then they aren't going to regulate in the light.

Keep in mind that as soon as the light turns on, it will place an incredible load on the batteries, forcing the regulation to kick in. This means that if you have a light where the switch physically connects and disconnects the light from the batteries, any of these cells should work for you. However, the Gladius, as well as many "high-end" lights do not rely on this physical connect/disconnect system, but rather on an internal computer that decides when to allow power to the light and when not to. To be able to accomplish this task, the computer must be in constant contact with the batteries; and if this little computer has a voltage limit of 7V and only puts a load of 9uA on the batteries, then you had better be careful about what you put in that light.

I would really love to do some "field testing" with all different kinds of batteries in the Gladius, but the risk of severe monetary loss makes that idea rather unappealing.

-Josh

I'm pretty simple, I guess. I put a Pila 168S in it and forget it. It gives a quick blink every 15 seconds. The only time I notice it is when I'm looking for it. I don't notice it if I'm really paying attention to what I'm shining the light on.

That being said, the Gladius could scrap the battery indicator. I've gone through 2 sets of CR123's playing with it and the battery indicator came on about 15 seconds before the batteries were dead. Not very helpful. If they scrapped this functionality then we would have no issue with Pila 168S.

Lightedge said:

the Gladius could scrap the battery indicator. I've gone through 2 sets of CR123's playing with it and the battery indicator came on about 15 seconds before the batteries were dead. Not very helpful. If they scrapped this functionality then we would have no issue with Pila 168S.

Click to expand...

Yes, that would be really nice.

However, at what level did the blinking start? Was it on high and the battery started blinking, or did it start on low? I am thinking you would have more warning from the high level. In using the light with a Li-Ion, the light died gradually as brightness levels were lost in sequence from bright to dim.

Scott

e-lectronics do have an "IC 3.0V" version that should work.

Can it works with a protected 18650? good performances?

NextLight said:

Dave, do you have a 100KOhm resistor to use for a minimum load while measuring voltage with a DMM?


OT, but my 3.0V Powerizer cells have also worked well in several LED lights. The do nstaflash a MN03, however.

Click to expand...

Nexlight

No I have not used a resistor the 3.6 volt bulb seems to place enough of a
load on the batts. By the way the same method works on disposable
123's. All the disposables I have tested (Surefire, Duracell, Panasonic) have
been under 2.8 volts under load brand new out of the box. I also have instaflashed an SF p60. I limit my use to leds or 6 to 7 cell incadescents for instance a three 123 m@g mod with a kpr118. It is MUCH brighter on the Powerizer.

Dave

SUMMARY

NextLight says, WEARING EGG MAKEUP:

I have been very busy, but had a chance to quickly duplicate Joshua's 37uA (standby) load test on my Powerizer 3.0V regulated cells. Unfortunately, my results essentially track his, ie. the voltage only falls (regulates) to 3.7-3.74V per cell, at a load current of ~37uA. THIS SUGGESTS THAT IN STANDBY MODE (what looks like "Off" for the Gladius) THE VOLTAGE FROM THESE REGULATED 3.0V CELLS WOULD PLACE ABOUT 7.5V ON THE GLADIUS ELECTRONICS; THIS IS 0.5V ABOVE THE MANUFACTURER'S DESIGN MAXIMUM. Though I have done this for many weeks, on multiple Gladius lights, with no problems, I need to correct my previous (apparently erroneous) statements, so others can make a well informed decision for themselves.

===================================

My initial (October, 2005) tests showed these cells regulate down ~80mV/uA of load current. Recent tests show that this is only partially true. From 4.2V no load, to about 3.7-3.74V, the cells down regulate ~80mV/uA as I earlier reported. Then, the regulation curve encounters a "knee." From ~3.7V, the cells requires almost 1mA load per mV voltage requlator output decrease, for the circuit to down regulate to 3.5V. (Arguably, our ultimate goal here is to get to 3.0-3.2V/ cell, but this thread addresses specific limits of these cells in the Gladius.)

In other words, my well worn but freshly charged (4.2V open circuit) REGULATED 3.0V cells did not regulate to below 3.5V/ cell (the maximum design voltage rating for a Gladius) until the load resistor was decreased to 180 Ohms. This is almost 20mA, and decades above the standby (commonly called "Off") current of the Gladius.

Later this week, I will connect the cells to my original Gladius externally, and look again at actual voltages and currents in all modes again. I will report my results here.

I apologise for my previous error(s), and any problems it caused anyone.

Lightedge said:

I'm pretty simple, I guess. I put a Pila 168S in it and forget it. It gives a quick blink every 15 seconds. The only time I notice it is when I'm looking for it. I don't notice it if I'm really paying attention to what I'm shining the light on.

Click to expand...

I'll second that, I've been my Gladius on 168's for about 8 months now as my primary duty light without any problems other than having to take the pila wrapper off the battery to make it easier to get in and out.

Nextlight-

I certainly would like to see your results. CPF member "LuxLover" and I have been doing lots of these experiments ourselves as well, with the batteries outside of the light. We are calculating how much these batteries actually regulate, the voltage (in various load levels), the amperage, and the wattage that the LED is actually getting. For now, I'm using a single RCR123 in my light, using a "dummy" made out of aluminum foil and wrapped in electrical tape. The light is still very bright, although it really doesn't compete with 2 primaries. I'm only doing this with the so-called 3.0V regulated and protected batteries (which is all I have at the moment). Perhaps doing this with some unprotected cells would result in a brighter light, as they will provide a higher voltage and current output.

When LuxLover and I are finished with our tests, we will also post our results.

-Josh