【Fresh Find】sofirn SP35T 21700 tactical light 3800lm SFN43 USB C Rechargeable

Thanks Jackie!

Ordered one. Came to $29 (pre-tax, no included battery) with the coupon code. Great deal. I’m pretty excited.

The C8L is an excellent light, and this having the same UI in a smaller, SP35-sized package, with an interesting new LED, and the buck+FET driver, it seems like a win.

Should be a great recommendation for those who like the SP35, but want a tail switch, or a very nice budget alternative to the Fenix PD36R.

I’m hoping to maybe get a bit more performance out of it with a P42A or 50S, but either way, the specs seem impressive.

Also grabbed another Yakorsei GD12 for $24, which is basically an 18650 tail switch SP35 with a UI very similar to this. SST-40, Buck driver, extra nice build quality easily on par with lights I have from higher end brands like Fenix, and comes in a nice retail box with accessories and an 18650 with built-in micro USB port. I grabbed a couple during the last sale for ~$22 each, and they make fantastic gifts for non-enthusiasts who just need a reliable, bright flashlight.

I think they were originally ~$75, but now that Yakorsei isn’t a thing anymore and they’re on clearance, they’re a steal for the features at this price.

thank you. GD12 yes it is a clearance price now, even much lower than our cost,Good choice, there are not many left

SP33S new color $6 code :SP33SCOLOR5
link: https://pl.aliexpress.com/item/3256804462497020.html?spm=a2g0o.store_pc_allProduct.8148356.1.5b947155A518pD&pdp_npi=2%40dis%21USD%21US%20%2491.98%21US%20%2441.39%21%21%21%21%21%4021365ec216633122849942017eb2fa%2112000029974317952%21sh&gatewayAdapt=glo2pol
LT1S new color $8 code :LT1SRU8
link :https://aliexpress.ru/item/1005004707621182.html?pdp_npi=2%40dis%21PLN%21402%2C09%20z%C5%82%21269%2C40%20z%C5%82%21%21%21%21%21%400bb47a1a16633119887708312e5922%2112000030178134359%21sh&sku_id=12000030178134339&spm=a2g0o.store_pc_allProduct.8148356.1.32156376UARXtb&gatewayAdapt=glo2rus&_randl_shipto=RU

Have had the light for 3 nights now, here are some pics and my initial thoughts.

It works with Wurkkos 20350 tube - I’m very happy about this, and will likely use it in this configuration after doing my testing phase.

Beamshots (please excuse the quality, I only have a cheap camera phone)


Runtime for High mode:

Runtime for Medium mode:

It’s not the performance I was expecting for high mode. On High mode - the light is very uncomfortable to hold, it’s too hot to touch. This shouldn’t be the case for a high mode, in my opinion. Due to the high heat, the ATR kicks in very quickly, 1.5 minutes - and then it’s a drastic drop. It basically nosedives off a mountain. This is a very poor performance compared to SP35 with SST40, which has a much more flat runtime for OVER 2 hours.


On medium mode it’s more reasonable, but still, the output doesn’t suggest proper regulation. Hotness in the hand is much more bearable and hand-holdable. The level of heat is what I expect from a High mode rather than a medium mode. My suspicion is that the driver for this light is not BUCK+FET. It doesn’t show the regulation one would expect from such a driver.

As I mentioned, the LED overheats very quickly. I don’t know why Sofirn chose this LED over some other proven LED because so far, I’m not impressed, I really don’t like the overheating. A youtuber called smothered chrome measures the turbo output at 2100-2300 lumens depending on the battery used. He also mentions the heat issue.

I would be happier if they just kept the original LED choice and reflector. I also have doubts that the driver really is BUCK+FET, because it’s performance doesn’t suggest the proper kind of regulation expected from that kind of driver.

The flashlight features and UI itself are excellent. I love the UI grouping of 2 modes, outdoor and tactical. I think it’s a very elegant solution for EDC+tactical option in one light. I also enjoy the ruggedness of the design. It’s utilitarian and practical. In build, it is the good quality that we have come to expect from Sofirn. My big issue with is with the LED heat and driver.

If anyone else has received theirs and are willing to conduct runtime graphs, please share them. I would like to know if maybe I just received a dud. I will also try to source another suitable LED to reflow, to see how much a difference it makes. I plan to use another more reasonable emitter. Maybe output is lower, or about the same, but I just want something that I can actually hold in my hand for more than a minute.

Also, an addendum to the manual. In tactical grouping, mode cycling only cycles through 2 modes - turbo and medium. This has been confirmed by Barry, saying it’s by design. The manual says that tactical group mode cycling goes through 4 levels - this is a typo error.

Low mode runtime graph.

It’s very odd to me that even on low mode there would be a dip. Maybe Jackie you misspoke when you informed us that it is Buck+FET driver.

Overally my impression:
Things I like:

- Good build and design

  • Great UI

Neutral:

- reflector is shallow, beam is floody. I think it’s personal preference but I prefer deeper reflector.

  • ‘ECO’ mode can be lower. 5 lumens is okay though. 1 lumen better.

Things I don’t like:

- Emitter choice. I don’t really mind the green tint or the cool white - but the overheating is unacceptable, in my personal opinion.

  • Driver. Poor performance esp compared to old sp35.

After all my complaints, do I still like this light? Actually, I still think it’s very usable. If you stay to medium mode you can still handhold for a few hours, and honestly speaking the medium output can be sufficient for duty-use. But the main point is I was led to believe that this driver was buck+FET which would have same performance as old sp35. Looking at the runtime graphs tells another story. I wouldn’t have clicked buy on this light, and instead gone for some other model with proven driver.

I am waiting for my emitter parcels to arrive to do some swaps and testing.

Darn, the runtime graphs sure make me wonder whether Wurkkos misspoke about the driver.
I recall communication issues in the past where a manufacturer said something similar, but meant the light was FET-driven, and the USB charging used a buck converter (or something like that).

Yes, I wonder if it was just a communication error. Something got lost in translation maybe?

But Jackie clearly says BUCK, FET.

It’s also possible that maybe the engineers or whomever Jackie received info from gave incorrect information.

Either way, I’m pretty disappointed about that fact.

I’ve never had issues with Sofirn products before, except one time they said the SC21 Pro came shipped with Anduril 2 but it turned out it was Anduril 1. Not that there’s anything wrong with Anduril 1, but there’s a lot of little improvements in A2 that I really like. Also, I don’t yet have the proper hardware for flashing new firmware.

You can never ever use zeroair’s graphs and state that high doesn’t step down. Zeroair uses a fan to cool the flashlights which is an incorrect method to test runtime.

I reviewed the SP35 and it steps down after a few minutes on High. Zeroair’s unit never hit the thermal threshold to stepdown because of the fan.

I forgot that. Thanks for the reminder.

Can you share your sp35 runtime graphs?

I think Zeroair’s graphs are still useful. He explained one time why he does the cooling and what setup he uses (it’s not a whole lot of air flow). It’s basically to try to emulate something besides testing in a still air head-up operation, which is not very realistic in itself. Breeze, ambient temps, blood circulation in your hand, etc. all contribute to moving more heat than a static test of a light. Sometimes he included non-cooled graphs along with the fan-cooled, and with some lights like the recent T3 titanium he didn’t use cooling at all.

All of these runtime graphs have some value but none of them really simulate real world use and run time…I think a lot of newer flashaholics don’t realize that and these graphs from various testers get a little more sense of importance than maybe they should. Always nice to see how quickly turbo/high step down and when a driver has good regulation down the line, but real world run times are often better than these graphs indicate since most people aren’t just running the cell from start to finish as they use the lights.

I deleted mine. Here’s another graph from another member here. Here you get 10 minutes, but I hate this behavior afterwards.

I’m confused. What’s causing it to act like that after 10 minutes? Is it something to do with the ATR? That looks horrendous

Poor coding. Sofirn and Wurkkos have been like that since they implemented that it try increase the brightness as the temp decreases. Cyansky and a few others do the same thing. A few recent models of Sofirn show they may have fixed it. The Sofirn C8L I reviewed was perfect.

I was actually this close to getting one based on i think your review and tactical grizzly’s. The only thing was I didn’t want something in the 45mm C8 head form factor. This SP35 size/form factor is near perfect for me for the usage I need it for.

What about the Convoy S21E with side switch? The issue with any Convoy model is that it won’t drop below High no matter how hot it gets. I think the S21E rose to 60°C in a review I skimmed through.

Hearing that this SP35T gets hot, I doubt it’s hotter than some Convoy models like the 4X18A hitting over 80°C

Specifically for the SP35T vs c8L it’s the dual-switch tactical and the UI that I was interested in. Convoy doesn’t offer anything really close. Sure you can stick a fwd clicky on one of their lights but it’s not really the same. This particular sofirn UI is really just perfect for me.

For the money I paid for this, I’m honestly willing to live with the negatives to have this combination of UI, tactical switch and form factor.

I don’t particularly need a light that can hotrod for long runtimes, I’m just surprised and really disappointed that this one does so poorly.

Still, I can’t stop carrying it even though it annoys me how fast turbo and high heat up. That’s how much I like everything else about this light.

Ah, i forgot it was you who owns the SP35T. I just remembered the SP35 being mentioned as better and thought about the S21E.

I think every other Sofirn model has the thermal regulation issue like the graph I posted.

I’m also disappointed because it’s on its way for review and I thought it was using a buck driver.

Looking forward to your testing.

The “thermal issue” is because people (including engineers) don’t really understand control circuits except superficially.

The sensor is usually on the driver, not right next to the LED (ie, right on the star), so the control circuit/program is simple: dim the LED when the driver gets too hot, brighten it when it’s cool enough.

That works great to keep the driver at a nice constant temperature, but it turns into a bang-bang controler for the LED.

Think of a steam-heat system in a house. Turns on and off, but keeps the house (well, the part near the thermostat) at a relatively constant temperature. The room/thermostat is the driver, while the LED is the boiler, or at the very least, the radiator.

Worse, the time-lag between when the LED snaps on full-hot until the driver starts feeling the heat, exacerbates the bang-bang nature of the LED’s current, ie, oscillation.

BIG thanks to darosk for sharing those runtime graphs.

I am VERY disappointed in this situation. The WHOLE REASON I bought one is because Jackie clearly stated that it was Buck+ FET driver. I would NOT have bought one if I knew it was only a FET driver.

I know Jackie isn’t an engineer, I know they wouldn’t deliberately lie to us and it was likely some kind of miscommunication. But there WAS a lot of subsequent discussion and excitement about the Buck+FET driver after Jackie said that, and it wasn’t corrected. Maybe Jackie didn’t read the thread which is also understandable.

Still, I feel misled. I was expecting it’d maintain the 1500 lumen high mode fairly well, and of course be able to achieve it even on low battery, or at least have flat output on the 500 lumen high mode, but it doesn’t - it drifts down with the battery.

It doesn’t mention the driver on the aliexpress page, so I’m not sure I’ll be able to return it.

It is quite bright. Just my observation comparing against other lights, but with a P42A I think it’s hitting somewhere close to the 3800 lumen rating. Throw might be 280m at startup on a fresh battery, for a split second, but really it’s a ~240-260m light.

Gets very hot on turbo, like you’d expect from an enthusiast hot rod, but also gets hotter than I feel like it should on high.

It’s SUPER floody. Almost like an XHP50.2. I don’t think you’d guess it was a domeless LED. It is a very nice beam, but I do prefer a bit more balance towards throw. It’s spitting so many lumens up close that it’s hard to use it for looking in the distance if there are things in the foreground. Of course it is very nice for lighting up entire areas at close range.

Tint is better than XHP50.2 and SST-40, which is a low bar, but moving into useable territory for me. Definitely above BBL, but not overly green even on lower levels. Some slight tint shift around the corona but not offensive to me. I suppose it’s a better choice than if they had an XHP50.2 in this with similar lumens, and it probably throws further than an XHP50.2. It’s fine, but I have seen better tint with other new chinese LEDs.

My main complaint with the UI (present on the C8L and Yakorsei GD12 as well) is there’s nothing between the 5 lumen eco and 100 lumen low. It needs a 30 or 50 lumen mode in between those two. I can live with it but it’s not ideal. Shortcuts and mode groups work well.

Build quality is great. Beefy springs, nice and thick without being too heavy, nice switches, well-protected lens. Side switch is very clicky on mine, but that’s a bit of a lottery - my C8L has the same side switch and it’s a little mushy, but they’re both good.

The light is just weird. It’s like a light with an identity crisis - it’s a floody enthusiast hot rod pocket rocket, stuck in the body of a tactical light.