[Available again! - BLF special edition light] new Sofirn AAA twisty high CRI 5mm LED

Oh, weird that they potted some of them but not the others!
Wonder if the factory didn’t get the memo when they manufactured certain colours.


or, I speculate

factory staff says: Boss, were running out of potting compound.

Boss says: those guys on BLF are really cheap… Im not buying another barrel of potting compound, too expensive, and too much labor. Finish the rest without potting…

factory staff says: Yes Boss :slight_smile:

factory staff says: Boss, the red 5mm LED has beam artifacts

Boss says: those guys on BLF are really cheap… Im not buying another bag of LEDs, too expensive, too much labor. Finish the rest with the LEDs you have.

factory staff says: Yes Boss!

One more here to confirm the above list, my colors are potted the same

Already did a first mod, one I was having in mind since the desire for a host version. Simplest travel nightlight ever for my son, for which application 7 lumen is too bright.

I used a 2300K 95 CRI led from rngwn (I cherrypicked one that was duv –0.0017) , which is shorter than the stock led so probably a littlebit light is lost in the head (looks like not much). And I changed the 220 (22 Ohm) resistor for a 221 (220 Ohm) resistor. The battery current went down from 70mA to 10mA (good for runtime) and the light went down from 7 lumen to 0.7 lumen (some of that reduction is caused by going from a 3400K led to a 2300K led). At 10mA, a AAA alkaline has about 2700mAh so that should be about 270 hours runtime (as opposed to about 40 hours for the stock light).

Scotchbrite vs stock

C01 670nm

Genius move by Sofirn in my opinion to pot some of the lights - what a fun surprise. I had ordered one potted color and one not-potted color already (plus two C01S).

Just ordered two more not-potted and planning to adjust output like djozz. What size resistor is used?

Maybe I missed it but….

Are the C01S’s potted?

All colors?

None of the C01S’s are potted, but some of the new C01’s are. (The original C01’s were all potted, though.)
I’m still waiting for mine in the mail to see what I got, haha.

I think the stock resistor was 0402, but I replaced it with a 0603 size one. Had to fiddle a bit with solder iron and tweezers to make both solder ends connect.

I won`t go below 1206, so I admire your work there! :wink:

I must admit that while hobbying I have done a lot of this small stuff by hand-soldering so I developed some skills, but it is still easy to get frustrated and fail.

djozz, would a pencil mod be ok for Rangwans high power LEDs? Not sure what the resistance would be.

It’s very hard to control the resistance of a pencil trace. It might work for a digital enable line, but not for this purpose.

Ok, thanks.

For a lower current the resistance must go up, so it starts with removing the stock 220 resistor. Removing a resistor often is the hard part for me, one reason could be that industrial solders may be lead-free and have a high melting temperature. (my method for heating up the resistor is loading the solder tip with a large solder blob and cover the whole resistor with that blob, once the resistor unsolders launch it with something pointy like the tip of your tweezer, or maybe a tooth pick).

If you want to increase the current the resistance must go down, Noctiluco posted a while ago that a complete bypass a bit more that doubles the current (so about 50mA to the led), that is easier, a simple solder blob on top of the resistor could do that, electrically connecting the two sides. Or a bit more subtle is solder an extra resistor on top of the old one. Or you could still go for removing the old resistor and replace it.

Removing it is, as you said, the hard part. I’d try a small copper wire, but 50mA is not so challenging for the 32K-C-EX.

Is it me or is this light basically a fenix e01 with slightly deeper “reflector”?

The fenix “reflector” seems to be a little cleaner beam imo, but i only have a handful of them to compare 3 fenix / 4 sofirn. I had put rngwn sourced in the fenix without soldering for easy to change leds for tinkering. Has worked well.

Btw, on fenix one had no oring, and all were potted. The potting makes a little shelf for the next led if you get the orig led removed cleanly.

One fenix had gooey uncured or false cured epoxy.

Two of my sofirn (black) lights are potted, i recommend “cracking” the pot seal with channel lock or carefully set visegrips (and isolating layer) before trying to push out led/ driver, especially if pulling from driver side, if not, see earlier pics of drivers and their bits separated from failed extractions (my first fenix at to much $$$…was torn up in likewise manner)

Hey… is trump the potting compound splitting (tearing) us all apart, leaving the bits scattered and broken on two sides…

Dang it…Couldn’t “resist” that. I should’ve kept it “light”.

What a circus! Minus the entertainment factor. Glad i have a hobby to keep “focused” on…


Anyhow, back to the first point, the inside bits look very similar to my fenix e01. Output may be more on the fenix lights?? (Estimation based on non scientific testing)

The only obvious difference on the driver is what i think is called an “inductor”. Of course that is just a flash visual comparison.


Lastly a question… was the old blf yuji 5mm a single mode also? And how were they side by side to fenix e01? Now my genius is kicking in, the names are similar! c01, e01…whaaaa

I would really like a driver like nitecore tube or klaurus mi2 (not sure model#) in the form factor of aaa tube light like c01/e01.

Btw2, the fenix heads and sofirn heads appear to be interchangeable, in my extensive one light testing.

I was closely involved with the design of the C01 and indeed the Sofirn C01 was 100% started as a Fenix E01 clone, which was at the time the E01 was discontinued by Fenix and many people still loved it, the dimensions of the light and the driver design were copied. During the design stage there was a lot of discussion here in this thread on what what could be improved on the E01 and that resulted in a two-way clip, a centered keyring attachment that combines with tailstanding, and holes in the tail for a magnet and a tritium vial. It must be added that the Nichia GS led of the E01 was cool low CRI and “throwy”, while the Yuji’s of the C01 were warmer (5600Kband 3200K option was available, even 5600K was significantly warmer than the “purple” Nichia GS), better tinted 95CRI, and very floody (60 degrees leds). The current batch with 3400K Sophia high CRI leds is very similar to the 3200K Yuji version.

thank you to everyone that has helped to make this project a success, including the customers,
the head instigators in charge:
djozz, slowtechstef,

and Barry from Sofirn
for your parts in bringing a High CRI Cockroach to the world

I hope for a Titanium and Copper version to follow :slight_smile:

And Noir from Germany contributed with the idea for the novel tail design that combines a centered keyring hole and tailstanding.