[Review] Sofirn SD05 Diving 21700 Flashlight

87 posts / 0 new
Last post
Macka17
Offline
Last seen: 1 month 1 week ago
Joined: 05/08/2017 - 02:31
Posts: 1025
Location: Queensland Australia

Sidney Stratton wrote: I initially ordered a SD03 (XML2 – 900lm) as didn’t have this timed throttling. Sofirn informed me this was no longer produced.

———————————————

I just now ordered the SD03 from Fasttech. Shipping date 20/5.
$28AUD with ins and tracking. L-M-H simple.

When you look at it. The 05, is basically a 1000? lum. Maybe. As it steps down after a coupla minutes.
I have a couple of divers already that do that and more.
and a lovely little red XANES 1600 lum. Divers. $22.60 AUD
Mates had it down to over 25mtrs outside the islands here when I first bought
Was very impressed for it’s size. Close work only, Basically.
But his prime light is a pro 4500? lum.
Just HMP grease the threads first. It gives a better seal.

I’ll Maybe order an ’05 next month for my Birthday. If reports on here improve.

It’s nice to get a fresh interest in Quality Budget torches. I was losing the acquisition glow.
What next?????.

I’m tending to go for divers in small/mid range.
Better built. better sealed. can bounce them around a bit more and similar prices to others.
PLUS. They seem to be getting away from that ugly water pipe look too.
A few curves in there.

All my good torches now have, in 18650.
Samsung 30Q. or Panasonic NCR 18650B’s
Protected and UN protected.
Other sizes, vary.

Sidney Stratton
Sidney Stratton's picture
Offline
Last seen: 1 day 14 hours ago
Joined: 01/28/2018 - 22:59
Posts: 415
Location: Canuckistan

To add to the confusion of this timed step down and the difference (lumens) in up / down switching to medium, Sofirn has a newer version without the timed step-down. And is advertised as 3000 lumens.

I presume they changed to the XHP50.2 – 3 volt variant and are using their SP33v3 driver adapted to the magnetic switch.

I asked to have the 5000ºK emitter. Read somewhere they can cater to special requests. Waiting for an answer…

edit (18:03): Sofirn Web Store

Funtastic
Funtastic's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 06/26/2014 - 02:14
Posts: 1005
Location: New Zealand

Thanks for your input, I’ll have to ask Sofirn on my next order

Texas Ace Lumen Tube calibrated with maukka lights

New Zealand store – https://www.piercingthedarkness.co.nz (NZ customers only)

d_t_a
Offline
Last seen: 16 hours 6 min ago
Joined: 08/04/2017 - 23:58
Posts: 2057
Location: Manila, Philippines

Sidney Stratton wrote:
To add to the confusion of this timed step down and the difference (lumens) in up / down switching to medium, Sofirn has a newer version without the timed step-down. And is advertised as 3000 lumens.

I presume they changed to the XHP50.2 – 3 volt variant and are using their SP33v3 driver adapted to the magnetic switch.

I asked to have the 5000ºK emitter. Read somewhere they can cater to special requests. Waiting for an answer…

edit (18:03): Sofirn Web Store

Decided to investigate this further, I have an earlier batch Sofirn SD05, I notice its brightest level is a bit less bright than the later version, however the early version SD05 has NoPWM on all modes (Low-Med-High), whereas the later batch Sofirn SD05 has fast PWM (detected by shining the light through a small portable fan blades).

Which seems to indicate that the early SD05 appears to use a similar boost driver as the early SP33v2 “2500” lumens version.
(also checked that the early SD05 “Medium” mode stays the same whether it was switched from Low-to-Medium or from High-to-Medium)

The later SD05 which has the different “Medium” mode (“Low-to-Medium” or Off-to-Medium is the regular Medium) [Off -to-Medium means the control ring set to Medium when battery inserted), has fast PWM on all levels. And the High mode is brighter (consumes more power).

The old SD05 (which likely uses XHP50.2 6-volt LED) seems to consume up to around 6 Amps max by measuring tailcap current.
whereas the new SD05 (likely uses XHP50.2 3-volt LED) consumes around 8-9 Amps max tailcap current.

Sidney Stratton
Sidney Stratton's picture
Offline
Last seen: 1 day 14 hours ago
Joined: 01/28/2018 - 22:59
Posts: 415
Location: Canuckistan

Lan Lee gave me this link SD05 5000ºK w/out timed stepdown.

You did confirm what I suspected. The newer model is DD w/fet and has its downsides. But not having this timed throttling is a net gain.

I ordered last night – it’ll take some 8 to 12 weeks to get here. I have orders backed up since March.

One another note, Hank4U2 hank posted a discounted SD03 on AE link. $12 US.
http://budgetlightforum.com/comment/1657498#comment-1657498
Funtastic is looking into some 50 units, but I saw 60 down to 52 w/in a couple of hours.

Funtastic
Funtastic's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 06/26/2014 - 02:14
Posts: 1005
Location: New Zealand

Yeah I might not be able to get that many as Sofirn is offline. I’ll grab whatever is left.

Texas Ace Lumen Tube calibrated with maukka lights

New Zealand store – https://www.piercingthedarkness.co.nz (NZ customers only)

Sidney Stratton
Sidney Stratton's picture
Offline
Last seen: 1 day 14 hours ago
Joined: 01/28/2018 - 22:59
Posts: 415
Location: Canuckistan

@ Funtastic:

Should be more considerate. The idea is that if many people are interested in a 5000ºK version, they would make readily available.

hank
hank's picture
Offline
Last seen: 2 hours 36 min ago
Joined: 09/04/2011 - 21:52
Posts: 9357
Location: Berkeley, California

Deobfuscation: I’m not Hank4U2 and I’m not Hank Wang and I’m not the other hank-plus-a-number.
I did post the pointer to the sale on the SD03. Hope y’all got a chance before the reseller sucked them all away.

Sidney Stratton
Sidney Stratton's picture
Offline
Last seen: 1 day 14 hours ago
Joined: 01/28/2018 - 22:59
Posts: 415
Location: Canuckistan

@ hank:

So sorry. My bad! Corrected the OP.

arek98
Offline
Last seen: 14 hours 57 min ago
Joined: 09/06/2012 - 12:29
Posts: 96
Location: New Jersey, USA

Disclaimer: I’m not a prolific diver, I did about 200 dives, almost all in warm water (Caribbean like), few NJ shore and some quarry in PA.

I think considering a price this is a best recreational (take on the trip to warm place and dive) diving light available right now.
Timer sucks but it is not a show stopper, medium is enough for night dive, you don’t need to blind everyone and fish (on recreational dives you are in group 3-7 people – guide/dive master and 1 – 3 pairs), if you want to take a picture you still need flood light and here high CRI is nice.
High is nice for day dive when you look into holes checking if somebody hides there but then 3min timer is not an issue.

Another thing, for some time I was wondering about this thing fluoro diving, when you look for stuff that illuminates under blue/UV light. Blue is better, UV harms marine life and blue is causing better response, negative is that you see blue and you want yellow filter to get it out.
Anyway, I have Ferei W150 that I modded and put Royal Blue Cree XT-E in it. It is blue but it is not very strong.

I got 3 SD05s and I thought, lets put a blue LED in one.
Right, easy to say…
It is 6V LED, where do I get 6V blue LED, then….

it was like “wait a minute it already has a blue LED in it, it just has a phosphor on it”
So, I took some small pliers and removed dome from the LED, it wasnt gentle Smile and it already removed most of the phosphor from LED, I scraped a rest carefully using wooden splinter and now it is a blue light!!
It seems little less blue then Royal Blue but I think this is actually good.

LED Before

LED after

Close on low

Back in light with reflector

Next to Ferei W150 (guess which one is it Smile )

In real there are no violet rings visible just blue, smooth transition from spot to spill.

gchart
gchart's picture
Offline
Last seen: 7 hours 7 min ago
Joined: 03/19/2016 - 11:57
Posts: 2510
Location: Central IL

That’s some resourceful thinking there, arek98! Beer

Sidney Stratton
Sidney Stratton's picture
Offline
Last seen: 1 day 14 hours ago
Joined: 01/28/2018 - 22:59
Posts: 415
Location: Canuckistan

Hey, talk about modding a light!

Unfortunately arek98, your links to pics come up 404.

arek98
Offline
Last seen: 14 hours 57 min ago
Joined: 09/06/2012 - 12:29
Posts: 96
Location: New Jersey, USA

Should be fixed now
Also I hopefully will try it underwater in September. With current situation no way to go on the trip right now. Can’t wait, hope to see something nice Smile

Funtastic
Funtastic's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 06/26/2014 - 02:14
Posts: 1005
Location: New Zealand

Sidney Stratton wrote:
@ Funtastic:

Should be more considerate. The idea is that if many people are interested in a 5000ºK version, they would make readily available.

I’m an authorized distributor mate.

Texas Ace Lumen Tube calibrated with maukka lights

New Zealand store – https://www.piercingthedarkness.co.nz (NZ customers only)

Funtastic
Funtastic's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 06/26/2014 - 02:14
Posts: 1005
Location: New Zealand

d_t_a wrote:
Sidney Stratton wrote:
To add to the confusion of this timed step down and the difference (lumens) in up / down switching to medium, Sofirn has a newer version without the timed step-down. And is advertised as 3000 lumens.

I presume they changed to the XHP50.2 – 3 volt variant and are using their SP33v3 driver adapted to the magnetic switch.

I asked to have the 5000ºK emitter. Read somewhere they can cater to special requests. Waiting for an answer…

edit (18:03): Sofirn Web Store

Decided to investigate this further, I have an earlier batch Sofirn SD05, I notice its brightest level is a bit less bright than the later version, however the early version SD05 has NoPWM on all modes (Low-Med-High), whereas the later batch Sofirn SD05 has fast PWM (detected by shining the light through a small portable fan blades).

Which seems to indicate that the early SD05 appears to use a similar boost driver as the early SP33v2 “2500” lumens version.
(also checked that the early SD05 “Medium” mode stays the same whether it was switched from Low-to-Medium or from High-to-Medium)

The later SD05 which has the different “Medium” mode (“Low-to-Medium” or Off-to-Medium is the regular Medium) [Off -to-Medium means the control ring set to Medium when battery inserted), has fast PWM on all levels. And the High mode is brighter (consumes more power).

The old SD05 (which likely uses XHP50.2 6-volt LED) seems to consume up to around 6 Amps max by measuring tailcap current.
whereas the new SD05 (likely uses XHP50.2 3-volt LED) consumes around 8-9 Amps max tailcap current.

I just tested mine and at 45°C it starts stepping down. That’s lower than the previous driver. The manual says 55 but it tests to be 45. That’s disappointing

Texas Ace Lumen Tube calibrated with maukka lights

New Zealand store – https://www.piercingthedarkness.co.nz (NZ customers only)

arek98
Offline
Last seen: 14 hours 57 min ago
Joined: 09/06/2012 - 12:29
Posts: 96
Location: New Jersey, USA
gchart
gchart's picture
Offline
Last seen: 7 hours 7 min ago
Joined: 03/19/2016 - 11:57
Posts: 2510
Location: Central IL

Great! I’m glad to see Sofirn / Wurkkos finally removing timed stepped down and letting their “ATR” spread it’s wings.

MrTick
MrTick's picture
Offline
Last seen: 2 days 20 hours ago
Joined: 09/05/2019 - 04:54
Posts: 2
Location: JO81

Tempted by the removal of 3min. timer I’ve ordered one directly from the Sofirn site.

I took it for some diving the very first day and I must say I was pleasantly surprised: it easily outperformed my old HiMax H7 clone.
I’ve tested it up to 30m(~100ft), it was working without any problems (no photos this time though).

While being a really nice (maybe a little to small for me) light, it has one serious flaw (at least for me): 6000K LED.
I know it’s mostly matter of ones preference so I won’t hesitate to say: I hate cold LEDs.

So next step looked quite obvious: let’s disassemble the light, replace the emitter with 3000K XHP50.2 and be happy ever after.
Unfortunately Sofirn has made some some further changes along with the step-down timer removal.

1. Disassembly

I must say this was second most difficult screw I’ve dealt with this week.
I’ve 3D-printed few head adapters but all failed to remove the glued ring.
Eventually I’ve made hybrid printed key with four M3×8 screws that fit nicely into head’s cuts:

I’ve bought proper 5kg vise, screwed it into my desk and after a brief struggle I’ve managed to remove the header.
For my surprise Sofirn has exchanged blue thread locker to red one, that’s much more difficult to break.

2. PCB

This was the biggest disappointment here. The boost converter is gone, now we have most simple MOSFET driver that passes PWM-modulates battery connection directly to 3V LED Sad

Two ICs aboard are PIC12F1882 (datasheet) + AON6512 MOSET (datasheet).
Three hall sensors are TSH251 (datasheet)

3. Some scope dumps

PIC generates 20kHz frequency that looks like thins (measured on the LED connectors):

LOW:

MEDIUM:

HIGH:

And there doesn’t seem to be a difference in duty cycle when powered from 3.6V (left) vs. 4.2V (right):

4. Conclusions

Switch to new driver is a step-back for me, I realize this solution is 2x cheaper than boos converter but I feel such nice light deserves something better…
The good side of new design is SOIC8 package of PIC controller, so it should be easy to connect to and reprogram.

Anyway I now wait for 3000K XHP50.2 to arrive (I’ve ordered 6V version so I guess I’d have to reflow it to 3V pcb taken from this light…) I’ll let you know if when I succeed Smile

Btw. this seems to be my very first post on BLF so: Hi everybody! Smile

gchart
gchart's picture
Offline
Last seen: 7 hours 7 min ago
Joined: 03/19/2016 - 11:57
Posts: 2510
Location: Central IL

Welcome to BLF! What a great and informative first post.

Funtastic
Funtastic's picture
Offline
Last seen: 1 day 17 hours ago
Joined: 06/26/2014 - 02:14
Posts: 1005
Location: New Zealand

No, the 6V XHP50.2 is different than the 3V version, you can’t reflow it to a 3V MCPCB.

Boost driver would lose Turbo at 3.7v which was around 8 minutes. The only good thing about the boost driver was 2hrs of High 920 lumens.

Texas Ace Lumen Tube calibrated with maukka lights

New Zealand store – https://www.piercingthedarkness.co.nz (NZ customers only)

MrTick
MrTick's picture
Offline
Last seen: 2 days 20 hours ago
Joined: 09/05/2019 - 04:54
Posts: 2
Location: JO81
Funtastic wrote:
No, the 6V XHP50.2 is different than the 3V version, you can’t reflow it to a 3V MCPCB.

Thanks for that info! I’ll look for 3V XHP then.

EDIT:
The worst new of the day: CREE doesn’t make 3V 3000K XHP50.2, just two cold colors are available Sad

rdhood
Offline
Last seen: 52 min 4 sec ago
Joined: 09/24/2020 - 09:15
Posts: 40
Location: Dahlonega, Ga

I’ve been holding off waiting for the 5000k version to come back in stock. I don’t buy much of anything anything cooler than 5000k or less than 90 cri anymore.

flashburn
flashburn's picture
Offline
Last seen: 1 day 3 hours ago
Joined: 09/21/2020 - 14:18
Posts: 48
Location: Italy

Hmm. I wonder if the glass is thinner as well. If I compare the recent photo with the photo in the first post, I get the idea.

A proper diving light should be at least 100m proof, also more for safety, even if this depth is normally not reached with scuba gear. Imagine a light filling with salt water… boom, that hurts.
I do diving actually quite some time and am looking for a good modern LED (non-incandescent) dive light, because I do not want to take two of them down just in case the bulb burns out. Up to now, LED divelights did not match my taste.

One very important issue is, the dive light needs to be a neutral-warm light, to fill up the view with more than just deep blue colour.
The deeper one gets, the more blue is the light, until it is all deep blue and getting darker.
Ideally, the light has a tint ramp, because in a cave situation, one could use a bit more of cool light as well, to not tint everything in yellow.
A part of the actinic blue spectrum could reveal some fluorescent colours.

Important is thermal regulation! Underwater the light is cooled very efficiently, so there is absolutely no use for a timed step-down, actually this could be a dangerous failure, but if someone uses the light on land, it should not burn out, so a thermal regulation that handles this as a safety feature is relly important.
For example, the halogen light can be tested only for a few seconds if not submerged in water, otherwise it will just burn out. Procedure even says to submerge the light first to test it and keep it there for a minute to cool down propagating heat. With LED this should not be necessary any more.

Also, there should be some end-of-battery warning, to not let someone in the dark without notice. This could be made with down-blink (dip) that is left on until user action as confirmation. Let’s say, two dip 3.2V, four dip 3.0V, nine dip 2.9V and forced medium as max. Also a battery voltage readout should be done, as not not being forced to open the light to test the battery. This could be done with fast changes on the ring from medium to low and back, calling up voltage readout.

Also, a simple humidity sensor would be fine, because incoming water could be detected long before the light fails and leaves one in darkness (and possibly in vicinity of an underwater explosion of the lithium battery). For that, a non-stoppable slow dip could work, slow because it should not disturb the diver too much.

I think the SD05 host could handle that all, if the driver is made accordingly, and even a tint change could be done with the ring, could be like:
1st round: neutral-warm, very little actinic blue
2nd round: cool, with partly actinic blue
3rd round: warm, no actinic blue

Just some ideas. I could help with developing that.

Edit:
forgot something also important: buoyancy! The light should not sink on it’s own, nor should it shoot up. Ideally, it would just stay if left alone. For this, a buoyancy addon could be done on the tailcap or on the bezel, or as a sleeve. Ideally, the light keeps even direction pointed at.

garrybunk
garrybunk's picture
Offline
Last seen: 1 day 11 hours ago
Joined: 10/31/2011 - 09:25
Posts: 6093
Location: Johnstown, PA
Funtastic wrote:
The only good thing about the boost driver was 2hrs of High 920 lumens.

No, another good thing (and perhaps best thing) is the extremely well regulated consistent output of the boost driver. As a mountain biking light for night riding, this is a big plus! I’ve been tempted to pickup the new version without the 3min stepdown, but I’m turned off by the lack of regulated output and only 6000k tint option.

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques

, MTBR’s Lights & Night Riding Forum
NOTE: Now hosting my photos from my Google account. Post up if you can’t see them. Older photos hosted on Photobucket or Flickr may disappear (PM me if you want access to them).
Lightbringer
Lightbringer's picture
Offline
Last seen: 3 hours 14 min ago
Joined: 08/30/2016 - 14:12
Posts: 12447
Location: nyc

Kewl, I knew none of this about diving. LOL

Only suggestion I’d make is to have the low-voltage warning come on a lot sooner. Below 3.6V, capacity is pretty much circling the drain. By 3.3V, it starts dropping like a rock. I was just watching this last night doing a capacity test of a coupla 18650s.

If you need more “cushion” to prepare, as in getting to the surface vs just pulling out a backup flashlight, much more notice/warning would be essential.

09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0

flashburn
flashburn's picture
Offline
Last seen: 1 day 3 hours ago
Joined: 09/21/2020 - 14:18
Posts: 48
Location: Italy

Yes, a longer notice/warning would be good.
Values given were just examples, giving enough room to reach or inform the buddy. But if he/she also has a light already at minimum, both be at the bad F soon. But keeping headroom for return to surface would mean at least 30 minutes spare runtime, depending on stops needed, plus extra headroom to consider aging battery or different chemical, summing up to 45 minutes or more. Warning voltages should then be tested, or calculated by constantly monitoring battery voltage change and extrapolating runtime based on prior measuremnts. Very EEPROM-heavy.

edit: A buck/boost driver is also essential to keep a constant light output over battery runtime. It can not be a technically sophisticated light fades just because the battery leaves the full-level voltage and enters the more linear part of the discharge curve. I expected that as a given fact, so a FET-PWM driver is a no-go-item.

Pages