Fenix E01 unpotted & driver efficiency data

[paulr]

Chimo this is great work you're doing. I'd guess it's possible to adjust the current by changing that resistor.

 

[chimo]

Thanks again folks,

I just tried one of my DS LEDs to replace the GS.

The output current jumped to 46.8mA and the output voltage remained constant at 3.2V.

I guess that puts a fork in the CV/CC debate.

Here is a photo of the board w/o the LED. I backlit the board so you can see the pcb traces. The +/- at the top are the LED solder points. You can see that the output resistor (labelled 30Y) does not provide feedback to the switcher chip.

In my opinion, using a CV driver in this type of light is perfectly acceptable as long as the current is kept reasonable to ensure the LED does not go into thermal runaway. A big well done to Fenix for doing this at such a reasonable price!

Modding this light, however, will require some additional work in determining the required resistor to ensure the above criteria is met for the chosen replacement LED. It will require determining the replacement LED's Vf at the desired current and then calculating the voltage drop required from 3.2V and calculating the required resistance.

Paul

 

[mighty82]

I have been trying to tell people the circuit is "cv" a long time. Finally someone can prove it.

 

[koala]

I have a feeling that markings on the chip relates to the output voltage?

3033 = 3.3v
3036 = 3.6v

 

[chimo]

I have a feeling that markings on the chip relates to the output voltage?

3033 = 3.3v
3036 = 3.6v

It's hard to say. Usually some character(s) in the top mark indicate a data code. There could also be some markings on the underside of the chip. I was getting 3.2V out.

Paul

 

[chimo]

I just did an LED swap for one of my E01s.

I put in a DS LED with a lower Vf than the GS. The LED current for the particular DS I put in is 46mA at 3.2V. Input current is almost doubled so runtime will be commensurately shorter.

The internal LED o-ring still seems to seal around the slightly more tapered DS LED.

One more observation:
I did not know if I liked how far Fenix set the LED back at first. After a bit of use, I think Fenix got the placement just right. The beam pattern of the E01 more closely resembles a high powered LED (hot spot with diminishing spill) than other 5mm LED lights that are very "ringy". I find the smooth, diminishing spill less distracting than the ringy beams and objects in the spill area are easier to distinguish.

 

[paulr]

Does the ringiness come from the GS led or its placement? My Arc AAA-GS beam is much ringier than the AAA-CS but I figured that was because of the further-forward led placement. The AAA-CS led is recessed quite far back, and the spill is wide and smooth. This_is_nascar has expressed a preference for the DS beam over the slightly brighter GS beam for this reason.

 

[Crenshaw]

wow, interesting thread, into subscriptions it goes...ive been wanting to replace my E01's led with a DS too...

Crenshaw

 

[chimo]

Does the ringiness come from the GS led or its placement? My Arc AAA-GS beam is much ringier than the AAA-CS but I figured that was because of the further-forward led placement. The AAA-CS led is recessed quite far back, and the spill is wide and smooth. This_is_nascar has expressed a preference for the DS beam over the slightly brighter GS beam for this reason.

I think it is mainly due to placement, however, I the different tapers of the LED likely plays a part as well. LEDs with smaller "lenses" (i.e. more tapered or smaller diameter such as 3mm LEDs) should have a wider beams. Since the DS is more tapered, the end "lens" is smaller thans the GS.

 

[Moat]

I've been playing around with shortening and lightly tapering a few of these GS's (installed in E0's), and the improvement in beam smoothness (lack of rings), spill, and especially tint (pure, even snow white!) has at times been remarkable. Definite loss of throw, however (oblong, violet/blue throw, that is).

I'm convinced that the overall shape and length of these LED's is far from being optically ideal for use in simple conical reflectored lights like the E0/E01 - especially for use as close-up/task lights.

Still need to determine the optimum length and taper, when I get the time - appears to be somewhere around 8.4/8.5mm (vs. stock GS @ 9.0/9.1mm, and CS @ 8.7mm).

Please excuse my addition to OT, Chimo.

 

[chimo]

I've been playing around with shortening and lightly tapering a few of these GS's (installed in E0's), and the improvement in beam smoothness (lack of rings), spill, and especially tint (pure, even snow white!) has at times been remarkable. Definite loss of throw, however (oblong, violet/blue throw, that is).

I'm convinced that the overall shape and length of these LED's is far from being optically ideal for use in simple conical reflectored lights like the E0/E01 - especially for use as close-up/task lights.

Still need to determine the optimum length and taper, when I get the time - appears to be somewhere around 8.4/8.5mm (vs. stock GS @ 9.0/9.1mm, and CS @ 8.7mm).

Please excuse my addition to OT, Chimo.

No problem, I look forward to seeing (and perhaps duplicating ) your work on ideal dome shape. I wonder if there could be some benefit in a non-round shape to the dome to counter the rectangular die shape? The die sits perpendicular to the flat surface formed by the leads.

I have seen a photo of one of these turned to a 3mm size. Not sure of beam output.

 

[Illum]

Re: Fenix E01 unpotted & driver data

Now...put a red LED in there and I'll pay hundreds. Well, I'm just joking about that...kinda

If someone manages to put a red [not orange] LED in one I'd buy it for around $30:thumbsup:
<Chimos got it :thanks:>
I need one for astronomy

 

[paulr]

Re: Fenix E01 unpotted & driver data

If you want an orange or red led 1aaa light, you can get a Peak Matterhorn for something like $25 during their memorial day sale, on for the next week or so. They are nice little lights with several led colors available.

 

[mighty82]

Re: Fenix E01 unpotted & driver data

Do anybody know what current a cree r2 or a q5 would draw at 3.2V? (this being a constant voltage circuit). Witch of them have the lowest vf? I'm thinking about modding one into one of my E01's. Just for fun.

 

[chimo]

Re: Fenix E01 unpotted & driver data

Do anybody know what current a cree r2 or a q5 would draw at 3.2V? (this being a constant voltage circuit). Witch of them have the lowest vf? I'm thinking about modding one into one of my E01's. Just for fun.

In my experience, the current at 3.2V will likely be a lot more than the cct can handle. You would have to profile the specific LED you want to use to determine a current (say 50mA) and Vf and then work backwards to determine the dropping resistor. You may also want to start with a "colder" tint as you will be under-driving the Cree and the colour will shift to the warmer side.

 

[paulr]

Re: Fenix E01 unpotted & driver data

Unless you want a wide flood beam, you'll want more of a reflector behind the Cree, or some kind of optic in front of it.

 

[Tohuwabohu]

I made some measurements of input current and battery voltage with an USB-oscilloscope.
I used a 10 milliohm shunt to measure the current.

First I measured average off-load battery voltage Vo, then average battery voltage Vin and input current Iin of some different types of AA-size batteries (scope-setting: 2ms/div, 200k points/track, 10 MS/sec, average of 32 tracks):

                                  V0(V)    Vin(V)    Iin(mA)
fresh Energizer Lithium AAA         1.80    1.71     73.8
fresh Energizer Alkaline AAA        1.59    1.55     81.7
used Varta Alkaline (1) AAA         1.28    1.05    107.3
used Varta Alkaline (2) AAA         1.10    0.86     52.5
Sanyo Eneloop AAA                   1.46    1.44     82.8

Then I took the following graphs at a setting of 20 microseconds/div:

fresh Energizer Lithium

fresh Energizer Alkaline

used Varta Alkaline (1)

used Varta Alkaline (2)

The voltage under laod is well below 1V. Chimo stated in the first post of this tread that the head drops out of regulation at around 1V. I think that is what can be seen here.

Sanyo Eneloop

Battery voltage shows smaller ripple than with alkaline batteries.

Just for comparison one measurement with an AA sized Eneloop:

Input current is nearly identical to AAA sized Eneloop but battery voltage shows even smaller ripple.

 

[jirik_cz]

Great post. Thanks. :thumbsup:

 

[Probedude]

Anyone have any info on the chip used? (manufacture, pn and datasheet?)

TIA,
Dave

 

[Illum]