Review: BLF-348 (SingFire SF-348) Nichia Special Edition Flashlight (Runtimes, Output, Etc)

You did hear correctly. I wrote the script days before I did the testing, and while thats true in some lights- its not true in this one. But the testing and my voice over during the runtime section was accurate. When I watched it before I uploaded it, I didn’t catch that. You’re right, it is contradictory. Oops! I added an annotation to explain that better.

I just have to chime in and say that none of the drivers in BLF-348s that I have taken apart included a 7135. All were a normal boost driver.

Ok then, do you know what chips ARE on the driver? It looks like 7135s to me, from the photos in this thread I can’t make out the numbers on the chips. The wording in that post lead me to believe that it uses 7135s. Then there’s the post #1174 in the BLF-348 thread where I was told 7135s employ thermal regulation…

The main thing is… the light appears to be thermally regulated on a li-ion battery. Do you agree? I was making an educated guess as to what was causing the brightness fluctuation. It LOOKS like thermal regulation. I have run this light several times with a 10440 for 10-30 minutes and it operated just fine.

Ah that could definitely cause some confusion. He was attempting to mod the driver by adding a 7135. But you are correct on all points. I’ll take a look to see if I can find which boost ic the driver uses.

I’ve been curious as to the fluctuations as well, although mine do it on alkalines and eneloops.

edit
Here is a photo of the driver. I haven’t been able to identify the crucial part in the middle that appears to be marked “J2NL” which is the boost converter, presumably similar to the “CFC2S” found on the nanjg 110

https://budgetlightforum.com/t/-/34918?page=17#comment-853953

Looking at that datasheet, it seems the PAM2803 does have a temperature regulation. So if that’s the chip used then it explains the behavior of the light. So I guess it doesn’t use a 7135 but another similar looking chip with built in thermal regulation. Does that sound like an accurate statement?

I certainly can’t argue haha. Honestly I didn’t take the time to read the data sheet. I got too focused on finding the exact product. Though our part is a 6 pin it does appear to be close in size to a 7135.

I have modded one with some stacked 10k resistors in place of the R50 and while output has dropped to my desired level I’m led to believe I’ve caused it to run out of spec. It will randomly jump to full (unmodified) brightness and stay for periods of time.

Has it improved your runtime?

I think it has. My daughter has been running around with it turned on for 5-15 minutes at a time for the past 3 days on a single eneloop. It’s been over an hour cumulatively and still going.

Awesome, good to hear!

Thanks for the heads-up! I have corrected my prior post to show Eneloop AAA rather than AA.

I apologize for the confusing wording in that thread. I was trying to relate that I know what a 7135 is and generally how it works. My goal was to add a chip with similar functionality to the driver. It worked but was horribly inefficient. I swapped the capacitor opposite the inductor with a 1uF (no idea what value was there already) and it seemed to increase runtime with no noticeable drop in output. If I recall correctly my light, with my eneloops, went from ~90 minutes to ~120 minutes before it was too dim to be useful.

After several days of reading about boost drivers my head hurt a lot. I came to the conclusion that the only real way to mod them is to replace the various components. Unfortunately, all the variables seem to interrelate. Without knowing the specs for that central chip it’s a guessing game.

I have variety of inductors coming so that I can swap them out and see how it changes output vs runtime.

No problem at all. Have you been using protected or unprotected cells?

I only use eneloops in my AA/AAA lights until my X5 arrives.

Gotcha, thanks!

That’s the same conclusion I came to. Its frustrating

Has anyone measured tail cap current on the cell types? Given that it survives 10440 it might be an ok driver for 2 X AA. What is current draw with 10440@ 3V?

Can I just thank you for the review mhanlen, top job…although you’ve cost me $6.99

I know the feeling, I’ve cost me $35!

Thanks for the review.

What the PAM2803 has is not what we’d normally refer to as temperature regulation. It has “Over Temperature Protection” which kicks in at 150c. The recommended operating temp range is –40c to +125c junction temp and an absolute maximum ambient temp of +85c. What you are seeing is not the chip proactively trying to reduce the temperature, it’s the chip reactively trying to prevent an outright failure while already operating above the absolute maximum temperature. The reason that it goes back and forth slowly rather than flickering is due to the 15c hysteresis. “Over Temperature Protection” should trip at 150c and not un-trip until 135c due to the hysteresis. The datasheet isn’t the be-all and end-all, maybe it will survive forever like that. My intention with the above is simply to help you better understand what the datasheet says.

See above RE: “survives”. That said, I think that 2xAA or 2xAAA will be fine. After quickly examining the pics over in this thread: Фонарик BLF 348 Nichia NVSW219BT - V1 (post#11 by MSS) we see that the driver is using a 0.5ohm sense resistor. That’s perfectly reasonable (about 0.19A output - exactly as one might expect with the output measurements). While the driver may be unregulated for part of the discharge, the amperage should not be high enough to cause any temperature problems. Swapping in a smaller sense resistor could help maintain regulation (more boost).

Does anyone have any opinions on these BLF-348 units w/ the Nichia emitters vs the previous GB lights from _the_’s GB over here: STARTED: GBGB: Stainless steel AAA clicky with a clip (CQG S1 “clone”) ?

Any possibility of a new group buy? I would definitely buy a couple or three. I did post a WTB in the forum if anyone has an extra they’d like to sell!