My quick impressions of the SP33v3 after spending some time with it. One note, I ordered with the 26650 battery but it did not come with a battery.
Pros:
Build quality. It feels solid and looks good. The machining and anodizing are very pleasing.
UI. It’s simple and intuitive. I really like it a lot. The moonlight and turbo (and strobe…) modes being hidden as well as instantly accessible from off make it very versatile. Good usable mode spacing too for the stepped mode. Plus of course there is ramping mode.
Cons:
Beam colors. My first reaction turning it on for the first time: “Wow, that’s ugly!” It has a hotspot that is white enough, but there is a VERY noticeable green/yellow ring around it, which then turns to a noticeably blue spill. At close range it is very distracting because the same object will look multicolored! Truly the worst part of this light. The colors are so bad, I don’t even notice color temperature, CRI, or any other typical concern. At larger distances or for general lighting, it is ok… ish.
The switch is difficult to find in the dark or by feel. It always takes me time to find it, and the easiest way is to feel for the bump of the charging port and know that the switch is 180 degrees away.
Overall, I think it’s a great flashlight and it has a ton of brightness. I’ve been using it with only 18650 batteries that are not high drain, so I imagine it is capable of even more brightness than what I’ve seen. Very abundant on that front.
A minor note is that the cuts for the “fins” of the heatsink are slightly sharp, unlike the smoothness of all other edges. Not a huge deal, but I do notice it and wish they were deburred slightly better.
Generally, the SP33 feels really good in the hand. The 26650 tube is a nice size and every time I hold it, I feel happy about the construction quality.
The USB charging is a feature I appreciate a lot as well. I don’t know what the charge current is but by comparing with the SP40 which is known to charge at ~1A, I’d say the SP33v3 charges at a higher current since it is a bit faster. That also means it seems to charge the same batteries slightly less than the SP40’s built-in charger.
With it plugged in to a charger, there is a 1.8 mA parasitic drain. That goes down to ~26-28 uA with the cable unplugged.
Welcome to the world of the Cree XHP50.2 and XHP70.2. The older 50 and 70 had a much nicer looking beam with no yellow corona, but they also tend to produce a donut hole beam pattern (OP reflectors needed) and the max output was lower. I will say you tend to get used to the corona over time. Maybe we will get better leds in the future.
I think the very new Luminus SST-70 might make a good replacement for the 6v 50.2. Hopefully we can buy them soon. It should make a nice replacement for the version 2 light. I dont know of a good led for this version 3 since it’s 3v.
Um, thanks… I don’t know if I’ll get used to the corona. It’s especially noticeable when the beam is not static. The yellow ring gives everything a pee-stained look, especially light-colored surfaces like snow!
In contrast, the 5000K XP-L in the SP-40 also doesn’t have what I would consider a pleasant color due a greenish hotspot with blueish spill, but at least that is more homogeneous without a corona, making it less hideous and more tolerable.
Does that make me a tint snob??
To be clear I still think the SP33v3 is a fantastic flashlight. I don’t go easy on criticism.
No, it just means you can see. No one likes the yellow corona of the 50.2, 70.2, XP-G3 and XP-L2. These use Crees new flip chip design which has phosphor all the way across the top of the chip. Here is a comparison of old vs new.
The performance is good, but something about the design, maybe phosphor on the sides of the led, creates the yellow corona.
Tint snobs can be really picky always swapping new leds in all their lights looking for just the perfect tint.
There can be a lot of variety in the tint of any led and it also changes based on brightness. If you see green in the hot spot it may be a sign that it’s above the BBL. I dont know what model xpl Sofirn uses for the SP40, but you can certainly swap it for a different one. Most folks prefer tints below the BBL because they are closer to red than green and that’s more pleasing to most people.
Using a U2 USB C power meter the SP33v3 charged at 5.03v / 1.3A using premium thunderbolt 3 certified 18" C cable from the meter to the USB 3.0 port on the SP33v3.A high end A to C cable was used from a recently made custom build gaming desktop PC USB 3 gen2 port to the U2 meter..The very new ShockLi 26650 5500mAh cell inside the SP33v3 was at 3.6v when the USB C port charging began.When charging completed the flashing red power button turned blue.
An SP40 using a U2 USB power meters but using an A to micro USB cable from the meter to the SP40 charged at 4.97v / 0.898A.The new Panasonic NCR18650G in the SP40 was at 3.75V when charging began.When charging completed the flashing red power button turned green.
I wonder, are those LEDs expected to be used with specific optics? I don’t see how any professional application could tolerate those inconsistent colors.
Maybe I “spoiled” my eyes by the order in which I viewed LEDs. My first “fancy-pants” flashlight was the C01S with the 4000K CRI95 SST-20 and TIR. It blew me away by how nice it looks, even though I did notice the slight greenness at low brightness. Then I got the i3E (Luxeon TX + TIR) and i1R 2 (unnamed LED + TIR) which looked worse in comparison. Then came the SP40 (XP-L + OP reflector) and SP33v3 (XHP50.2 + OP reflector) and suddenly the previous two looked like beauty queens.
I wonder if the big difference is TIR vs OP reflector, with the former presumably mixing any colors to produce a homogeneous beam that is invariably more pleasant and less distracting than bands of colors.
As a side note: the human eye and brain are strange. What looks good one moment can look terrible depending on the contrast with something else. 4000K SST-20 looks yellow compared to cold CCT lights, but then looks crisp white compared to incandescent bulbs. My brain gets so confused!
Thanks for these measurements! What are the final voltages of the cells after charging has ended?
Yes. The FT03 is offered in both SST-40 and 3v 50.2 and they use the same mcpcb. Just a matter of reflowing the led. Same package size so the reflector and centering ring are all the same. Keep in mind that output is a lot less, maybe 1800-1900lm at most due to smaller die size (2mm square and 3mm square)
Just did some testing with the SP33 v3 and EC01. Set both to about 190 on my lux meter (max stepped mode for the sofirn). My FT03 with a fully charged shockli measures 387 so 190lux is somewhere close to 1000 lumens. Both lights were at 26°C at the start of the test (room 22.5°C).
Beginning of test: about 190lux.
SP33:
13 mins (46°C head) step down to around 70-80 lux
Once it cools down a bit it starts stepping up again. The temperature management doesn’t let it become too hot to hold. I’m sure it could run near 1000 lumens without stepping down if the temp was set higher. Yesterday I tested it whilst holding it (170-180lux 900ish lumens) and it didn’t step down after 30+ mins.
EC01: temp set to 65°C
13 mins 57°C negligible drop in output, body too hot to hold
18 mins 153lux
22 mins 118lux
At the end of the test the cell was 50°C, body was about the same and the head was 59°C max.
It seems like the longer head and cooler (slightly more efficient) LED helps the SP33 dissipate heat better.
What optics might help? Has anyone used them with TIR lenses?
Maybe a diffuser film or lens would be an easy fix. As a proof of concept, shining the light through two layers of clear bubble wrap (just because I had some nearby) shows dramatic improvement: the corona is almost totally gone, and the spill is less blue. Everything looks closer to “white”.
Nice. Based on that I prefer the SP33v3’s behavior. Where did you measure temperature?
Your observations match my own with a thermocouple held on the head by the rubber charging port cover’s flap (not under the cover, just the pull tab).
On Turbo, it stayed in the low to mid 40s C range. The brightness did ramp down and up noticeably, but the changes were made sufficiently smoothly not to be annoying.
However I was not using a high drain battery.
Temp was measured using a cheap infrared thermometer at the head wherever I could find the hottest spot. Tested it tailstanding indoors at max stepped setting which is about 1000 lumens with a shockli 5500. This drops as the cell voltage declines but it doesn’t seem to drop as much as the EC01. Could be cause the battery has more capacity. The EC01 will also ramp up once it cools down. This would’ve happened if I left it for longer or set the temp lower. I tested it before with temp set a bit lower at 60°C and it stablised at around 90-105lux which should be about 500 lumens. This doesn’t make sense to me cause my MH20GT could do that whilst being smaller and having a less efficient LED. Maybe it’s more than 500. I highly doubt it though cause I’m assuming the FT03 is 2000 when it’s probably a bit less…so it might actually be less than 500…makes even less sense.
If I understand you correctly, your directly associating lux with lumens? These are completely different measurements. Are you integrating all the light? You have me quite confused.
Thanks. Those voltages are right in line with what I observed too. Maybe also try the same 18650 cell in the SP33v3 for one less variable, but I’m guessing the result will be similar.
Well it already has an OP reflector. Perhaps it needed stronger OP. Blending the corona and reducing the blueness of the spill make it much much nicer.
On a side note, I’m finding I’m not the biggest fan of strong hotspots. Blended, homogeneous, wide beam profiles seem more pleasant and useful to me, certainly for shorter distances. Also, hotspots seem to give a tunnel vision effect, even (especially?) after turning off the flashlight.
Astrolux EC01 has an ever so slightly slightly stronger OP reflector than the SP33 V3.0 (identical reflector and glass, o-rings and SS bezel) just enough that you can see it in the beam pattern (bit smoother - actually the reason I've noticed it in the first place)
For high tint shift emitters like the XHP 50/70, XP-L2, XP-G3, etc. a narrower reflector with the smallest emitter hole as possible helps much more than an OP texture (and gives a cleaner, sharper beam) - the lower the angle between the base of the reflector and the emitter, the less tint shift - the higher the angle, the stronger the tint shifting in the corona.
No, I don't have a beamshot comparison, as I don't have a good way of taking such pictures. The difference is marginal and the tint between the two would be much more noticeable than the beam pattern itself. Being the same geometry reflectors, the tint shifting is also the same, just a little bit smoother beam on the EC01, but actually looking worse because of the overall cooler tint vs my SP33, thus more tint shifting into green.
I only have the version 2, but the reflector on it is just barely OP. It’s almost SMO if you look at it closely and look at how sharp the hot spot is. If they made it more OP it would definitely help the appearance.
No one likes the yellow corona of the 50.2, 70.2, XP-G3 and XP-L2. These use Crees new flip chip design which has phosphor all the way across the top of the chip. Here is a comparison of old vs new.