[Technical discussions] Sofirn SP35 - SST40, 21700 size, buck driver

137 posts / 0 new
Last post
dropman
Offline
Last seen: 16 min 54 sec ago
Joined: 02/03/2019 - 11:04
Posts: 66

I was thinking about buying this light but this “advanced” temperature regulation is very dissapointing.
Thanks for review.

pennzy
pennzy's picture
Online
Last seen: 2 min 6 sec ago
Joined: 12/10/2017 - 19:45
Posts: 3186
Location: United States , Pa.

They regulated the temperature well but at the cost of steady lumen output.

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

I am just doing an analysis on High mode. As expected I see ATR is being triggered as well. I am going to post more details later on.

pennzy
pennzy's picture
Online
Last seen: 2 min 6 sec ago
Joined: 12/10/2017 - 19:45
Posts: 3186
Location: United States , Pa.

Is the max temperature fixed or adjustable?

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

pennzy wrote:
Is the max temperature fixed or adjustable?

You mean the stepdown threshold? No, it’s fixed. Sofirn says their new ATR uses three threshold values (45/55/65°C) that control brightness increase/decrease differently. In a nutshell: The hotter the light gets the more quickly the brightness will be decreased.

To my mind the average temperature kept during operation is too conservative with only 48-52°C. gchart revealed in his review of the no-ATR version that SP35 can maintain High mode (approx. actual 840lm OTF) stable at 56°C, provided that ambient conditions are probably around 18-22°C.

I hope gchart will not mind if I borrow his runtime chart of SP35 no-ATR on High mode...

Tendou
Offline
Last seen: 48 min 43 sec ago
Joined: 03/26/2020 - 03:49
Posts: 130

Darn! I was waiting for the ATR version for safety. But I should have gone for the non ATR version. Now I will only buy them if they rectify this

pennzy
pennzy's picture
Online
Last seen: 2 min 6 sec ago
Joined: 12/10/2017 - 19:45
Posts: 3186
Location: United States , Pa.

Thanks, it was that 45/55/65 thing that threw me.

Muto
Offline
Last seen: 2 hours 11 min ago
Joined: 09/04/2012 - 16:42
Posts: 2593
Location: Southeast, PA

So already Sofirn is offering 40% off new ATR version, being that we now know the ATR is whacko, how hard is it to disable?

I really like this company but they sure know how to shoot themselves in the foot over and over.
Why can they not send out like 10 pre-production units to veteran qualified testers to hash them out?

And no not me! I have no advanced testing equipment nor the tenacity that is takes to do in depth technical reviews but there are plenty of good people both here and on other forums that this should be a no brainer.

The difference between Hoarding and Collecting is the illusion of Organization
.

“History doesn’t repeat itself, but it sometimes rhymes,” Mark Twain

After the Apocalypse there will be only 2 things left alive, Cockroaches and Keith Richards

Fescron
Offline
Last seen: 3 hours 18 min ago
Joined: 10/17/2020 - 10:16
Posts: 148
Location: Belgium

Muto wrote:
So already Sofirn is offering 40% off new ATR version, being that we now know the ATR is whacko, how hard is it to disable?

I really like this company but they sure know how to shoot themselves in the foot over and over.
Why can they not send out like 10 pre-production units to veteran qualified testers to hash them out?

And no not me! I have no advanced testing equipment nor the tenacity that is takes to do in depth technical reviews but there are plenty of good people both here and on other forums that this should be a no brainer.

I have a feeling they just can make such mistakes. I suppose a lot of the people who buy their lights have little to no idea how they should really perform, so they have no reason to complain…

For instance, I recently got a SP36 4000k with firmware normally flashed on the LT1. I’ve made a forum post here and also put a lot of time in making a video to convince Sofirn what happened, but I didn’t really got a response on both. After messaging for some time with their AliExpress store they eventually send me a new one, this time with the correct firmware. I’m pretty sure I’m not the only one that got one with the wrong firmware but I suppose I’m one of the few who know something was wrong…

Don’t get my wrong, I really like the lights Sofirn makes, but I guess they have to cut some corners to keep the price down and/or make those mistakes and there aren’t really a lot of repercussions for them.

Muto
Offline
Last seen: 2 hours 11 min ago
Joined: 09/04/2012 - 16:42
Posts: 2593
Location: Southeast, PA

Fescron wrote:
Muto wrote:
So already Sofirn is offering 40% off new ATR version, being that we now know the ATR is whacko, how hard is it to disable?

I really like this company but they sure know how to shoot themselves in the foot over and over.
Why can they not send out like 10 pre-production units to veteran qualified testers to hash them out?

And no not me! I have no advanced testing equipment nor the tenacity that is takes to do in depth technical reviews but there are plenty of good people both here and on other forums that this should be a no brainer.

I have a feeling they just can make such mistakes. I suppose a lot of the people who buy their lights have little to no idea how they should really perform, so they have no reason to complain…

For instance, I recently got a SP36 4000k with firmware normally flashed on the LT1. I’ve made a forum post here and also put a lot of time in making a video to convince Sofirn what happened, but I didn’t really got a response on both. After messaging for some time with their AliExpress store they eventually send me a new one, this time with the correct firmware. I’m pretty sure I’m not the only one that got one with the wrong firmware but I suppose I’m one of the few who know something was wrong…

Don’t get my wrong, I really like the lights Sofirn makes, but I guess they have to cut some corners to keep the price down and/or make those mistakes and there aren’t really a lot of repercussions for them.

Agree!

Just like Lumintop got away with dangerous FW3A, most people who got burned just either returned it or threw it as far as they could.
With the horrible translation of quite a few manuals most customers have no idea if the thing is operating per the chart given (if they give a chart) and what passes for customer service today is “Would you like a Replacement or a Refund?) and that is only at Amazon.
If you bought straight from the Far East its “Send us a video of problem please” and then they give the same canned message back that there is nothing wrong, that is the way they operate, you know with Blue led of death or burning skin hot. “But we put Hot emblem of flashlight head”

Is what it is.
Sometimes the Flashlight Supreme Beings shine on you and you win the tint lottery or you get a newer version of Firmware.
That’s a good day right there:)

Later,
Keith

The difference between Hoarding and Collecting is the illusion of Organization
.

“History doesn’t repeat itself, but it sometimes rhymes,” Mark Twain

After the Apocalypse there will be only 2 things left alive, Cockroaches and Keith Richards

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

So, I made another analysis on High mode...

 

High mode is officially rated at 950lm. gchart measured 840lm with his calibrated equipment. That’s the number I now take for reference.

 

Observations:

Just like the version without ATR, the new SP35 with ATR can maintain its High mode output steadily for approx. 23 minutes or until 48°C head surface temperature is reached. After that I noticed that ATR was triggered and intervened very aggressively. The output then frequently dropped by about 30% in short intervals to approx. 580lm (calculated value). After 2h 55min the flashlight turned off automatically due to LVP.

 

Personal conclusion:

  • SP35 runs 5min longer without an initial start on Turbo. This is quite usual.
  • SP35 w/o ATR can maintain High mode continuously at approx. 56°C. SP35 with ATR runs about 7°C cooler but shows a mediocre output regulation, resulting in large brightness fluctuations.
  • The full potential of the light is not being used. I think that another 5°C on top would have been feasible. I get the feeling that this ATR is more focused to 50°C instead of 55°C.
  • Unlike more optimized flashlights like Acebeam L17, SP35‘s ATR prevents any hysteresis effect while heating up. Thermal regulation interacts too fast and too significant, so there’s no room left for short exceeding of its designed limits. On a positive side: The flashlight never gets any warmer than 50°C - possibly a business decision due to customer complaints about (too) hot flashlights?

I had a brief talk with Sofirn this morning, especially regarding my observations mentioned above. They want to discuss my feedback with their engineer.

JaredM
JaredM's picture
Offline
Last seen: 10 hours 21 min ago
Joined: 10/31/2011 - 13:33
Posts: 1803
Location: Pittsburgh, Pennsylvania

Its like someone forgot they were making a flashlight. Great thermal regulation though! Ha

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

Supplemental:

I do not mean to badmouth SP35 since it offers an extraordinary package of great form factor, constant current regulation, efficient buck driver, good allround beam profile and well spaced modes at an amazing bargain price. If only that thermal regulation was better...

I wonder if those fluctuations would disappear if they flashed Andúril onto the MCU of SP35. Maybe this could be an alternative to the ongoing investigation and improvement of their proprietary UI?

I also noticed quite some different runtimes on High mode. According to the specs it should run 3h 40min with a 4000mAh battery. My sample clocked in at 2h 55min (4.16V to 2.84V LVP) using a 5000mAh battery.

Tendou
Offline
Last seen: 48 min 43 sec ago
Joined: 03/26/2020 - 03:49
Posts: 130

Probably they would name those SP35 pro.

TheIntruder
Offline
Last seen: 3 weeks 2 days ago
Joined: 03/09/2018 - 02:11
Posts: 257

Muto wrote:
So already Sofirn is offering 40% off new ATR version, being that we now know the ATR is whacko, how hard is it to disable?

I really like this company but they sure know how to shoot themselves in the foot over and over.
Why can they not send out like 10 pre-production units to veteran qualified testers to hash them out?

And no not me! I have no advanced testing equipment nor the tenacity that is takes to do in depth technical reviews but there are plenty of good people both here and on other forums that this should be a no brainer.

Businesses can be good and some things, but not others. Those strengths and weaknesses determine their level of success, or lack thereof.

Obviously, the company possesses the level of competency required to design, build, and market their lights to a good degree of success, certainly with the enthusiast crowd, and likely with the online consumer market at large as well.

Product marketing is not among its greater strengths, and the product management for the SP35, ranging from the confusingly recycled name (Why not the unused SP34? Perhaps a numerology thing, but the company’s model nomenclature have never had much logic, so who knows?), technical bugs that escaped the development phase and into production, and ultimately, performance shortcomings in the final “final” product exposed the weaknesses in more than one area.

One can only guess at why that is, without knowing the company structure and its development and production processes, but such issues are solvable. But, only if the company recognizes them and desires to improve. Personally, I’m with those who like the products, and the company, but with some frustration, and see some unfulfilled potential.

Fescron wrote:
I have a feeling they just can make such mistakes. I suppose a lot of the people who buy their lights have little to no idea how they should really perform, so they have no reason to complain…

For instance, I recently got a SP36 4000k with firmware normally flashed on the LT1. I’ve made a forum post here and also put a lot of time in making a video to convince Sofirn what happened, but I didn’t really got a response on both. After messaging for some time with their AliExpress store they eventually send me a new one, this time with the correct firmware. I’m pretty sure I’m not the only one that got one with the wrong firmware but I suppose I’m one of the few who know something was wrong…

Don’t get my wrong, I really like the lights Sofirn makes, but I guess they have to cut some corners to keep the price down and/or make those mistakes and there aren’t really a lot of repercussions for them.

I’ve beat on this horse before, but the flashlight market hasn’t been hurt by products that are marketed in misleading ways, and don’t live up to their inflated marketing specs, even amongst enthusiasts who should know better. So what incentive is there to improve or do things any differently? Especially if it costs more and eats into profit?

Whereas the firmware issue with your SP36 could reflect a careless production worker using the wrong files when preparing drivers, the first non-functional ATR, and then underperforming ATR in the SP35 reach further up the chain.

Fortunately, I’ve never really had to deal with a complex customer service issue, technical or otherwise, and those of a minor nature have been handled well by the company, as long as the “right” party is contacted. But, knowing who that is can be confusing, and frustrating if that party is non-responsive.

One thing that piques my curiosity, and to me, really deserves more transparency, is the web of players involved with some of these companies, what their true relationships are, and how they operate. Some, like Convoy, are simple. It’s Simon’s business, his shop, and nobody else speaks for or represents him. Others, like Sofirn, have multiple sales outlets that operate independently, and multiple people speak for the company, with it not entirely being clear in what capacity.

But, back onto the topic at hand, the irony is that whatever resulted in the first SP35 production run, lacking a major feature, actually resulted in an arguably better product, at least in the right hands.

Unheard
Unheard's picture
Online
Last seen: 11 min 16 sec ago
Joined: 01/16/2019 - 11:38
Posts: 1814
Location: Germany

What was it? 30% amplitude difference at high mode? Who would notice that without luxmeter? How much time should Sofirn invest in flatten beautifying the curve?

Smile, you cannot kill them all.

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

Unheard wrote:
What was it? 30% amplitude difference at high mode? Who would notice that without luxmeter? How much time should Sofirn invest in -flatten- beautifying the curve?

I could easily notice it with my bare eyes even on High mode, no luxmeter required. Especially on lower brightness the human eye becomes more sensitive to output fluctuations. As always: YMMV!

My aspiration is to help them becoming better, not worse. I am sure they can improve things here without the necessity of redeveloping everything from scratch. It probably takes some tweaks on their algorithms, that’s all.

Unheard
Unheard's picture
Online
Last seen: 11 min 16 sec ago
Joined: 01/16/2019 - 11:38
Posts: 1814
Location: Germany

Lux-Perpetua wrote:

I could easily notice it with my bare eyes even on High mode, no luxmeter required. Especially on lower brightness the human eye becomes more sensitive to output fluctuations. As always: YMMV!


You sure be right, I might underestimate that.

Smile, you cannot kill them all.

gchart
gchart's picture
Offline
Last seen: 50 min 17 sec ago
Joined: 03/19/2016 - 11:57
Posts: 2986
Location: Central IL

Lux-Perpetua wrote:

Supplemental:


I do not mean to badmouth SP35 since it offers an extraordinary package of great form factor, constant current regulation, efficient buck driver, good allround beam profile and well spaced modes at an amazing bargain price. If only that thermal regulation was better…


I wonder if those fluctuations would disappear if they flashed Andúril onto the MCU of SP35. Maybe this could be an alternative to the ongoing investigation and improvement of their proprietary UI?


I also noticed quite some different runtimes on High mode. According to the specs it should run 3h 40min with a 4000mAh battery. My sample clocked in at 2h 55min (4.16V to 2.84V LVP) using a 5000mAh battery.



If it’s anything like their SP32A V2.0 (one of their other buck driver lights), it’s using a 10-pin PIC chip which isn’t so easy to retrofit Anduril onto. Once I do my comparison of ATR vs No-ATR, I might have to take one apart and get a closer look at the driver.

As far as the runtime on High goes… notice how my No-ATR version “only” lasted 2 hours and 15 minutes? I feel like that’s pretty respectable, but nowhere nearly the claimed runtime. If ATR were even more aggressive (not that I want that!) it would indeed last longer. So perhaps that’s how they arrived at 3h 40m? Dunno.

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

gchart wrote:
If it's anything like their SP32A V2.0 (one of their other buck driver lights), it's using a 10-pin PIC chip which isn't so easy to retrofit Anduril onto. Once I do my comparison of ATR vs No-ATR, I might have to take one apart and get a closer look at the driver. As far as the runtime on High goes... notice how my No-ATR version "only" lasted 2 hours and 15 minutes? I feel like that's pretty respectable, but nowhere nearly the claimed runtime. If ATR were even more aggressive (not that I want that!) it would indeed last longer. So perhaps that's how they arrived at 3h 40m? Dunno.

Thank you for taking a deeper look inside the driver specs, Gabe. Thumbs Up  Beer

On a side note regarding Andúril, I was told that manufacturers may soon run into a major bottleneck for MCUs compatible to Andúril, probably even the old ATtiny85 chip. It is uncertain how severe the impact of this bottleneck will be and how flashlight makers will address this issue.

Concerning SP35's runtime: On TLF I was tought a lesson today by another member. He said the entire runtime specs in the ANSI chart did not add up. Provided that Luminus SST-40 consumes about 2.5A at 1000 LED lumens (excluding losses in driver efficiency) it seems impossible to reach 3.5 hours runtime at rated 950lm on High mode. Taking losses in efficiency into account (approx. 15-20%) the real output of the ATR version is apparently far lower than this. I don't want to argue about his assessment, also because I cannot measure any actual lumen readings.

My layman assumption i.a.w. the datasheet (6500K) is:

650lm @ 1500mA or 1040lm @ 2500mA forward current

I measured 2h 55min runtime on the ATR version with a 5000mAh battery. Without losses in efficiency the power consumption should be less than 1750mA. This would be equivalent to approx. 780lm constant output on High mode. This assumption does not take into account: full capacity by discharging the battery down to 2.5V, losses in driver efficiency, losses due to reflector and glass lens. If I deduct a loss of 25% in total for these variables it would result in actual 585lm OTF on average on High mode. Due to these fluctuations the output could be jumping between 670lm and 470lm. IDK, maybe my calculation is too pessimistic or maybe Sofirn got a higher flux bin, much better than average specs on the datasheet. I would feel much better if someone more experienced in this were to verify and comment on this. LOL

SpinRay
Offline
Last seen: 2 weeks 5 days ago
Joined: 03/28/2021 - 16:42
Posts: 15

I received my non-ATR SP35 yesterday and won the battery lottery (5000 mAh) as well. I think this is a great light overall and a great value.

gchart
gchart's picture
Offline
Last seen: 50 min 17 sec ago
Joined: 03/19/2016 - 11:57
Posts: 2986
Location: Central IL

Lux-Perpetua wrote:
On a side note regarding Andúril, I was told that manufacturers may soon run into a major bottleneck for MCUs compatible to Andúril, probably even the old ATtiny85 chip.

I also believe this to be true. I’m nearly out of attiny1616; I think I have ~5 left. The usual suppliers are out of stock. I’ve got an order in at Mouser, hoping to receive some in the next few months. Most of the other attiny’s are also out of stock. It looks like there’s a few of the attiny85 to be had, but not many. I figure if the big automotive companies are shutting down their production facilities (such as those used to build the F-150) then we truly are in a crunch and the “little guys” don’t stand a chance at getting the chips they need. Sad
zoulas
zoulas's picture
Offline
Last seen: 1 hour 40 min ago
Joined: 06/01/2020 - 08:35
Posts: 1017
SpinRay wrote:
I received my non-ATR SP35 yesterday and won the battery lottery (5000 mAh) as well. I think this is a great light overall and a great value.

For $24, no reason not to get it.

SpinRay
Offline
Last seen: 2 weeks 5 days ago
Joined: 03/28/2021 - 16:42
Posts: 15
zoulas wrote:
SpinRay wrote:
I received my non-ATR SP35 yesterday and won the battery lottery (5000 mAh) as well. I think this is a great light overall and a great value.

For $24, no reason not to get it.

The ATR version comes in a 24 with the 40% off coupon. There is a coupon code for the non-ATR, but it is not that great. I paid 32 I think… or I got the wrong code. Either way – worth it in my book.

Lux-Perpetua
Lux-Perpetua's picture
Offline
Last seen: 8 hours 46 min ago
Joined: 03/01/2018 - 04:39
Posts: 2964
Location: between 365nm and 750nm

For those of you who like to see the difference between the first 100pcs of SP35 (w/o ATR) and the latest revision (with and w/o ATR) with improved reflector, I took two whitewall shots (WB: 5400K) today. Personally, I think Sofirn did a good job here. The revised version sacrifices some peak intensity (and range) in return for a more appealing beam shape mostly free of artifacts which also suits to the form factor of this flashlight much better.

zoulas
zoulas's picture
Offline
Last seen: 1 hour 40 min ago
Joined: 06/01/2020 - 08:35
Posts: 1017
Lux-Perpetua wrote:

For those of you who like to see the difference between the first 100pcs of SP35 (w/o ATR) and the latest revision (with and w/o ATR) with improved reflector, I took two whitewall shots (WB: 5400K) today. Personally, I think Sofirn did a good job here. The revised version sacrifices some peak intensity (and range) in return for a more appealing beam shape mostly free of artifacts which also suits to the form factor of this flashlight much better.


Nice, the new one looks better.

gchart
gchart's picture
Offline
Last seen: 50 min 17 sec ago
Joined: 03/19/2016 - 11:57
Posts: 2986
Location: Central IL

My updated SP35 with “working” ATR showed up yesterday. I still think the SP35 is an amazing light & value. That said… my first runtime test is underway and (as expected) I’m seeing the same swings that Lux-Perpetua observed. I wish a little more focus was put to improving the ATR algorithm to help smooth this out.

The peaks are, on average, 1 minute and 45 seconds apart. The ramp up and down is very smooth; swinging between ~500 and ~1000 lumens. So in real life, it’s really not as noticeable as this graph might make you believe.

Once the tests are done, I’ll overlay graphs of the ATR and non-ATR versions and post those up.

JaredM
JaredM's picture
Offline
Last seen: 10 hours 21 min ago
Joined: 10/31/2011 - 13:33
Posts: 1803
Location: Pittsburgh, Pennsylvania

I wonder how it would behave with fan cooling? Just thinking about which PID parameters should be adjusted. My guess is the fan wouldn’t change the amplitude much, just the mean output. Could be wrong..

On second thought. It might ‘raise the valleys’ by cooling the light faster once current is pulled back. This is of course based on an assumption that light to air conduction + radiation is much slower (ie limiting) than internal conduction.

gchart
gchart's picture
Offline
Last seen: 50 min 17 sec ago
Joined: 03/19/2016 - 11:57
Posts: 2986
Location: Central IL

I’m with you on the second part. It would likely “raise the valleys”. With regards to PID, I doubt that is relevant here. I imagine it’s more of a “ohh, too hot! now too cold!” algorithm without a bunch of extra math/prediction.

JaredM
JaredM's picture
Offline
Last seen: 10 hours 21 min ago
Joined: 10/31/2011 - 13:33
Posts: 1803
Location: Pittsburgh, Pennsylvania

I bet you are right. In that case. There could be a simple two stage method. Let’s assume its 50c target. Set a ~2 degree window with a low gain value, then beyond that a high gain. If these two gain values are tuned, oscillations would reduce like a pid initially and then reach a continuous low amplitude ring.

Pages