Short Review of Sofirn SP33 v2 with XHP50.2

From post 177.

I was told by Sofirn that their 5000 and 5500 are basically the same cell, but some gave higher capacity and are labeled 5500mah, while the lower ones are labeled 5000mah.

The Shockli 5500 is a superior battery. A real 5750 or so mah and good at delivering high amps. I think it came out in either 2017 or 2018 so it’s quite a modern cell. I use it in my SP33 and get longer battery life than any of my other 4300-5000mah cells.

Thanks guys. that’s interesting to me at least :slight_smile:

Does the xhp50.2 version come in neutral white or is it a typo? I think they copied and pasted from the xpl version but not sure. https://www.aliexpress.com/item/Sofirn-SP33-Mini-Powerful-Led-flashlight-18650-torch-light-CREE-XPL2-LED-Mini-Handy-flashlight-6/32831075588.html

It says 5350K-5700K, but it’s more cold white. More towards 5700K or maybe even 6000K. I would not call it neutral white.

I just received mine from Sofirn official shop and it's indeed cool white but that's no surprise as Sofirn advertise it as 6000-6500K.

I would be carefull about what is said in your link as the only color temp "5350k-5700k" mention is in the "clicky order thingy" and the pictures that show the led aren't from an xhp50.2 (obviously smaller led)

I don't like cool light neither and my idea was to order the one from Sofirn shop and change the led for a warmer one if i really dislike it.

Got it only two hours ago so i have no tested it for real yet but my first thought when powering it for the first time was "uuuuh what a ugly green ring" wich seems to be typical for xhp50.2 with reflector

Interesting, 5700K is considered Daylight White. My flash units fire at that temperature for my camera’s. These are $400-600 flash units. I’ve always found it interesting what Kelvin people find “comfortable” when in real life the actual Kelvin of ambient light is not what most people think it is.

Are you confusing green with yellow? The 50.2 and 70.2 typically have a yellow corona. I’ve never seen green.

To me, the flip chip technology from Cree (XP-G3, XP-L2, XHP-50.2, XHP-70.2, XHP-35.2) has a greenish yellow sickly aura around the hot spot. I’ve found that slicing the phosphor off the substrate surrounding the die pretty much eliminates this, the only way to do it though is also with a slice of the dome. There are no bond wires but it’s fairly easy to mess up and get a bit of phosphor off the die itself and reveal the blue light.

@ Kame Sennin;

Changed the 5700º to 5000º and see no yellowish / green halo. But the tint is somewhat ‘softer’.

Arrow Electronics

Well yeah, Sidney, that one’s a 90 CRI version. :wink: Nice find! It won’t put out anywhere near what the cooler white ones will do but still, beautiful light…

@ DB;

I did a ceiling bounce measurement w/ phone app and got 2579 lm (instant on). My measurements may be off some as my reference light is a Dorcy 3s “C” cell (alkaline). And I do downplay the data (throwing out the highs and lows on some 5 tests).

I believe their boost driver has something to do with this value - I was also very surprised. And the price / shipping was excellent!

What i meant is are there two versions. The one in the sofirn official store on AliExpress says 6000-6500k but there’s another seller on AliExpress which shows a warmer tint. I think it’s just a mistake tho. Btw do you think this is better than the convoy S11. I need something which is floody but can do 1000ish lumens continuous whilst still being comfortable to hold.

Perfect description imho ! ;)

@Sidney Stratton : Thank you

The SP33 v1 is the discontinued unit with the xpl. All SP33 sold now are v2 and have the 50.2. The description in your link is incorrect about the led tint. A mistake.

I have zero experience with the S11. It is rear clicky only, which I don’t care for.

This is a great thread! Thanks Jason and all you reviewers/contributors!

I ordered one kit through the Amazon group buy and one by itself through Ali.

With the Amazon Sp33, I got approx. 6.3A from the Sofirn 5500 and with my Shockli 5500. The beam is a little yellow within the white hotspot.

With the Aliexpress sourced Sp33, I got 6.45A with the Shockli. The hotspot is solid white and perhaps, tighter.

I haven't done any spring mods.

I suppose the beam differences are due to reflector positioning.

I don't mean to relate beam appearances and current reads.

I don't have an S11 but for approximately the same price, I prefer the Sofirn.

Interesting comparison between the VG10s and SP33.

The VG10s has 1100 lumens vs 2500 but does not appear that much brighter because of the hot spot. You actually have to pay attention at the spill and the larger/smoother hot spot to see where that lumen difference goes.

I always preffer floody to throwers for an EDC. However, in an urban enviroment, floody “waste” light because it’s competing with ambient lighting. So a thrower with half the lumens appear to be be just as bright.

Two instances….in a dark area standing in between two buildings, the SP33 shows off its high lumens.

When i pointed across the street at a building 350 feet away, hard to tell because both covered the building about equally and spill was irrelevant.

Another interesting thing was the tint made comparison more difficult.

In the end….i stll preffer the SP33 at twice the price….even though we are talking about 15 vs 30 bucks.

Floody is a relative term. The SP33 is floddier the lets say a VG10. Larger, smoother hot spot and more spill.

The more lumens, the better . i would love the Thrunite TC20 with an xhp70.2 but at $90 it’s just too expensive. At $30, the SP33 is a better buy. Of course there is cheaper with less lumens, but 15 vs 30 bucks is not much of an argument.

I compared all 4 at the same time.

The SP33 is whiter then the Thorfire Q8/VG10s/TK15.

I would say it’s in the 6000k plus range.

OK, quick question for the driver folks. I’ve messaged Dan but he’s not around much lately so here goes. Is there any reason this driver wouldn’t work with IOS/D4 ramping firmware? I was hoping it would work but it won’t even give a flicker except when it rarely seems to ramp a bit or sometimes just gives up turbo for a second. Using IOS ramping on an 85 MCU from RMM, Decent FET, a single 7135, diode on D1, 10uf cap with zener piggyback on C1, 36k pulldown resistor on R4 over by the FET. I can build a working driver on RMM’s FET+1 board and these components so what gives? I know it’s a hack, don’t laugh too hard…

This is out of my knowledge zone, but just to confirm, you switched to a 3v led and loaded a 2 channel driver firmware, right?

Hopefully someone with more driver design knowledge will answer you soon.