L2D-CE High mode dims after a few seconds?

I have a L2D-CE Q5. When I enter High or Turbo mode with somewhat used (2x1.3V) alkalines, the flashlight enters "high" mode for a second or so, then dims to ~medium in a rather abrupt manner. This isn't the same behavior I had with my L2D-CE P4, which would gracefully degrade. Is this normal behavior?

EDIT: Additional information
The new L2D-CE Q5 doesn't have the "L2D bright flash" when turning it on, which suggests new regulation circuitry.

When the L2D-CE is on High mode, it gets warm, whether it's actually outputting "high" or "medium". When the L2D-CE is on medium mode, it does not get warm. Therefore, after the light goes refuses to deliver "high" output, I can get better battery life by switching it into medium mode. I suspect a defective light.

Welcome to CPF evilr00t.

Sounds like it is because the batteries can't put out enough power, and the driver is stepping down to a level they can support. I don't have a Q5, just an older L2D-CE [P4?] and this is the behavior I get with alkalines when using high or turbo, unless the batteries are very fresh. Works much better w NiMH rechargeables or lithium when using high or turbo. The newer low self-discharge NiMH, such as Sanyo Eneloops, are tremendous with the L2D.

LED_Thrift said:

Welcome to CPF evilr00t.

Sounds like it is because the batteries can't put out enough power, and the driver is stepping down to a level they can support. I don't have a Q5, just an older L2D-CE [P4?] and this is the behavior I get with alkalines when using high or turbo, unless the batteries are very fresh. Works much better w NiMH rechargeables or lithium when using high or turbo. The newer low self-discharge NiMH, such as Sanyo Eneloops, are tremendous with the L2D.

Click to expand...

Haha, I'm using my L2D to kill off alkalines actually. (I've got some 18650 flashlights that are way brighter, I mainly use the Fenix on low and medium)

Sorry, I added the second part of my post later. I just tested this a second ago, and this is what makes me really suspicious:

I let the flashlight run down set at "high", and when it switched into the lower output mode, I switched it off and dunked it in cold water to cool it off. Then, I set it back on "high" and it switched into lower output mode right away, except it's cold. About ten minutes later, the flashlight is really hot again: the head and the batteries are hot.

I dunk the flashlight in water again, then fire it up in medium, which outputs I think, exactly the same amount of light as high-stepped-down**. This time, the flashlight doesn't get hot anymore. **(I just verified this with my camera.)

From what I know about switching power supplies, this doesn't make sense... it's as if something's stealing the LED current at high-stepped-down-to-medium; I get the same light without excessive heat when I set it to plain old medium.

EDIT: I think I'm going to do some adjustable PSU work in a minute. Testing theory.

It seems to me I recall seeing that most lights still use up more battery power if set to high compared to medium, even if the circuit has stepped down the output to the same level as medium. I know I once went through a new set of lithiums pretty fast in my L2D that had stepped down from high.

You guys are going to like this post a lot.

Doing voltage measurements from L2D-CE
Not testing turbo, probably not safe (I'm using a 3A 18V adj PSU)
Might blow the head anyway though, we'll see. *BOOM HEADSHOT*


Preliminary results:
L1D users (or L2D users like me who use an aluminum foil "battery") should be interested in the <1.5V results!

Methodology:
1. Constant voltage, variable current. Current limit set to 2.1A. (See note 2).
2. Note that these are not battery measurements, these represent infinitely awesome batteries with no internal resistance.


Bold entries show out-of-regulation behavior.
Non-bold entries are in "High Efficiency Mode" [HEM].

Observations:
1. High mode goes out of regulation at 1.3V, and begins to squeal audibly as well as go crazy on the power consumption. Medium mode goes out of regulation around 1V, and begins to go crazy on power consumption, but doesn't squeal. Low mode does not go out of regulation at all, but instead, the flashlight refuses to turn on below about 800mV.
2. The Vf of the LED at 350mA (assuming high = 350mA) is around 3.5V. I suspect it's actually lower, but Fenix's circuit probably 'inflates' this measurement of the LED 350mA Vf to be higher because if it were 3.3 as the Cree datasheets say it should be, we'd have problems driving the LED at low with fresh Alkalines (1.5+1.5 @ 60mA is quite plausible).

Note 1: I didn't measure Turbo mode, but... at 1V, it drew over 2.1A before I realized I was probably cooking the circuitry!
Note 2: I found out that the circuit may need to draw over 1.3A when switching modes at low voltage. If you don't feed it enough, it'll remain stuck in low-efficiency mode, explained below. I observed this when testing at 0.9V. The flashlight drew 1.3A (the current limit!) and depressed the PSU's voltage output to 400mV! This yields... 520mW, which is incidentally the same power consumption of medium!! But it was not as bright as medium, apparently because it's also much less efficient!)

Circuit behavior:
The circuit requires 800mV to start; no "run-down" test was done to determine shut-off voltage. (These are different; you can run a L2D-CE down enough that you can then turn it off, but not turn it back on again; the LED would be dimly lit for this to happen)
The circuit proceeds to bootstrap in "Low Efficiency Mode" to hit a target output current, and will draw incredible amounts of current to do so; the flashlight may even SQUEAL!! Once it reaches the target output, it will go into "High Efficiency Mode" and reduce power consumption while staying in regulation.

So what happens when I have my flashlight set to "High"? (Note that old, used 1.3V alkalines are probably below 0.8V/cell at 700mA)
1) 2.7V to 1.5V: the circuit goes into HEM, I have High mode etc. and I'm happy.
2) 1.4V to failure: the circuit goes out of regulation. LEM kicks in as the circuit tries to regain regulation. The batteries are drained at incredible currents (see note 2), and if the batteries can keep up, the light remains as bright as it was in HEM. But that is sadly not ever going to be the case, so output drops from "high" to "medium" for me, while consuming a lot more power than HEM-medium.

Going to do 1.5, 2.0, 2.6, 3, and 3.6V measurements.

Huh, bumped because I finished the post. Woah that was long, and I can explain the behavior now. Wonder if that's how it really supposed to be.

And the answer is no.

Stolen from Flashlight Reviews.

Notice the graceful degradation from full output at 1.5V to zero. I see a step from super-bright to medium, then from medium to low (just saw it go a second ago!). Fenix has changed their circuit, or I have something that's VERY different from what Chevrofreak tested.

A continuous discharge behavior is very different than intermittent. At 1.3v open circuit, an alkaline cell is dead. If you try to pull the high load that the L2D demands on high or turbo, it will quickly drop in voltage and increase in resistance, and the light will not be able to maintain output. I'm talking ~1v here.

That is the typical behavior any high power light will display when using a dead alkaline cell.

I would further warn that using dead alkaline cells in a high power light is an excellent way to ruin a nice expensive flashlight. You are just begging for one to spew its guts.

Marduke said:

A continuous discharge behavior is very different than intermittent. At 1.3v open circuit, an alkaline cell is dead. If you try to pull the high load that the L2D demands on high or turbo, it will quickly drop in voltage and increase in resistance, and the light will not be able to maintain output. I'm talking ~1v here.
That is the typical behavior any high power light will display when using a dead alkaline cell.
I would further warn that using dead alkaline cells in a high power light is an excellent way to ruin a nice expensive flashlight. You are just begging for one to spew its guts.

Click to expand...

The forum's eaten my post three times.

1. Thanks Marduke! 1.3V = dead!!? *surprised*

I did some calculations and got around 1.1V for 1.3Voc cells drawn at High. Then again, it doesn't account for the cells discharging, which they will rapidly do at such a high load.

I used the naive equation: (so tell me if I'm missing something)
Vcircuit^2 - Voc*Vcircuit + Rinternal*Pcircuit = 0
and solved for Vcircuit, from which I can look up the Pcircuit in my measured data.

Using Energizer's published data for internal resistance, I get non-real solutions for Vcircuit for Voc < 1.25V/cell (Voc = 2*[1.25V/cell], Pcircuit = 1.2W, Rinternal = 2*[0.35ohm/cell]). If I'm interpreting this correctly, this means that it's theoretically impossible to draw over 1.2W from an alkaline cell that's below 1.25Voc, which is in line with observations. In fact, internal resistance goes up like crazy below 1.25Voc to above 0.7ohm! Kinda cool actually!

I get the feeling that Fenix has changed the circuit in the new L2D to take advantage of the flatter NiMH discharge profile (which incidentally is what should be used in the light anyway (or Lithiums)) as opposed to the sloping Alkaline discharge profile, because I do recall my old L2D not stepping down modes (I see high-medium and medium-low on the new L2D) but gradually reducing in output. This is interesting because the old L2D also did not get hot when it ran down the batteries in High.

If you start with a fresh alkaline, you will see gradual declining output. But you are starting with a dead alkaline (yes, 1.3v open circuit is dead for alkaline), so you will only get that first burst before it falls to the appropriate spot on the constant discharge graph, which should be around the same level as medium.


You can also think of it this way. When the alkaline cell's internal resistance goes way up when you turn the light on, the circuit is working harder in an attempt to pull the current it requires but can never achieve, forming this nasty little equivalant to an endless "do while" loop that does nothing constructive other than produce a bunch of extra heat.

The L2D is not a light designed for alkalines (in any generation), and can only really use them in low or medium for any length of time. If you want to use high and turbo, NiMH is your most cost effective way, being MUCH cheaper than alkalines, and with better performance also.

BTW, this runtime graph of the Photon Proton Pro using an alkaline cell will give you an idea of what your light is trying to do:
Source:
http://cpfreviews.com/LRI-Proton-Pro.php