XTAR S1 (Production) Review (3 x XM-L U2 | 1-3 x 18650)

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
The XTAR S1 is a beast of a light and the flagship in XTAR's current lineup. They tout it as a "Rescue flashlight" with 2,350 lumens and 475m of throw to go along with it.

These are outstanding numbers to be certain and should serve those fulfilling SAR duties well. I am fortunate in that my most extensive SAR duty to date has involved looking for the TV remote trapped between the sofa but suffice to say that should I need to call upon it for more serious matters, I'm confident the S1 will serve me well.

I have received a finalized production version for review but given there are already excellent ones comparing this production version against the pre-prod one (done by candlelamp and selfbuilt), I will not be covering any of that here.


MFG SPECS
MSRP: $299.99 USD
5af50c4.jpg

Max Duration: 2350Lm/1hrs20Min (Tested by 3 x 18650 3100mAh battery)
Constant current circuit - maintains constant brightness
Smooth reflector, throws beam over 475 meters
Powered by 3x18650/3x18700 button-top battery.
Ø 83mm(Head dia)x Ø 47mm(Body dia)x 240mm(length)
888-gram weight (excluding batteries)
Made of durable aircraft-grade aluminum
Premium Type III hard-anodized anti-abrasive finish
Waterproof to IPX 8
Magnetic ring switch
Anti-roll, slip-resistant body design
Accessories included: Flashlight, Lanyard, Pouch, Spare O-rings, Manual, Warranty card


PACKAGING & CONTENTS
The XTAR S1 arrived in a nice thick cardboard box surrounded by ample foam cushioning:
2d0c641.jpg


Given I'm often asked whether 18700-sized batteries will fit in XXXXX light in my reviews, I decided to purchase six of their 2600 mAh batteries along w/the WP6II charger (these are included as part of the kit and the batteries are shown packaged with the S1 in last pic above):
f13a01b.jpg



DESIGN & FEATURES
As befitting its main purpose the XTAR S1, is a large light weighing in at 888* grams without batteries with a 3.26" (82.9mm) bezel with a nice AR coated glass:

(*which incidentally is a most auspicious number for Chinese people so I wouldn't be the least bit surprised if this was intentional ;))


The glass is surrounded by eight crenlations that allows one to see if the light is on if stood on it's head:


The two patterned rings provide some grip but not too much:


There are five cooling fins that aid heat dissipation, however it actually doesn't run all that hot (see runtime section for temp graphs):
9318851.jpg


These five fins alternate between eight flat and eight curved columns:



Just aft of these fins is a ring with the output and mode selection:


...and then the control ring itself which has the selection indicator engraved and is completely smooth save for these scallops machined in ladybug print:

They have been sanded so that while not sharp provides just enough traction to allow easy use.

The 3 x XM-l's are fitted into a deep, smooth, combo reflector that give it tremendous throwing capability:
ab07abf.jpg

Each LED is centered perfectly through the use of centering molds.

The head, body and tube can all be disassembled from each other:


The tube features machined grooves that run perpendicular to each other:

...that are pretty functional as well (see Size & Handling section for further details).

There are two "flat sides" on the tube with one featuring the company name and logo and the other beard the model and serial number:


The texture provides a good grip and offers a nice contrast to the smooth grooves:





SIZE & HANDLING
ade0934.jpg

L to R: XTAR 18700 | NITECORE TM11 | Niteye EYE30 | Elektro Lumens Big Bruiser | Sunwayman V60C | Niteye EYE25 | XTAR S1 (Production)

At the risk of repeating the obvious, this light is BIG! It's easily the largest one in my collection, however, as compared to some other lights (e.g. SR90, TK70), it's actually downright petite considering the output/throw it's capable of.

Given the three cell configuration, the size of the tube comes in smaller than say the four-cell TM11's:

XTAR S1 tube (L) @ 1.83" (46.5mm) vs. NITECORE TM11 tube (R) @ 1.96" (49.9mm)


UI
All modes and output levels are controlled by the selector ring which are (going clockwise from left to right):
SOS, Strobe, High, Stand By, Preset, Select (program)

When first turning on the light in High mode, there is a soft "ramp-up" feature. Check this video for an explanation of that:


This was intentional by design and is listed as no. 4 "Brightness gradually increasing to protect eyes and circuit." under the Features section on the S1 page on XTAR's website.

Programming is done by setting the ring to Select in which it will start ramping up and down from 15lms to 2350lms w/1 second stops at 25%, 50%, 75% and 100% respectively (albeit it's not always easy to spot this if you aren't paying careful attention). The total ramp time is about 45 seconds which should in theory allow sufficient time to dial in the exact level to one's preference.


MEASURED PERFORMANCE
As part of a new standard for my reviews, I measure the performance on my PVC LMD by first calibrating the sensor's positon w/a Xeno G10v2 (immediately perfectly regulated in med. mode). All readings taken @31 seconds (to conform w/ANSI FL) and ambient temp of 75F (24C).

0f9b802.jpg

Left: Given I needed to hit the 4K range on the Extech HD450, I calibrated the sensor with the G10v2 on high for 480lm
Middle: S1 ready for lumens measurement and runtime testing
Right: lux reading

And here are the measurements: High: 2560lms (60.5K lux @ 1m)

Here is a video explaining a little of the process and my new method for conducting runtime testing based on measured lumens rather than relative output. The goal of this is to provide an idea of real world lumens as an average for non-regulated lights.


Sorry I'm still new at these so it's not particularly good but it'll get better. :)



BEAMSHOTS
Beam angle

[to come]

Indoors (5m)
Max


Min

For details of the above indoor shots and comparo vs. many other lights, please check Epic Indoor Shots Trilogy

Whitewall Hunting
Exposure settings sequentially from top left: 1/25, 1/100, 1/800, 1/1600 @ f2.9 on AWB (light is ~.4m to wall / camera ~.59m):




RUNTIME & BATTERY CONSIDERATIONS
The relevant battery stats are provided above each runtime graph along with:
- Voltage of the battery at the start and end of the test
- Current draw as taken right before the test
- Actual runtime using ANSI FL1 (first in HR and then in M so for the XTAR 2600's read this as 1.2hrs OR 71min)
- For High, captured the temperature: ambient, the head at start and the max it reached (fan was used for all bats)

Before we dig into the run time, using XTAR 18700 2600 mAh cells I measured the following current draw:
1 cell - 4650mA
2 cells - 5200mA (2600mA each)
3 cells - 5820mA (1940mA each)

Keep in mind that that the cells are arranged in parallel so the load will be reduced beyond one cell with three ultimately only drawing roughly 2A each.

With that said, I was able to achieve 1hr and 11min with the XTAR 2600mAh batteries (note the decently regulated run). I fully expect to meet (actually beat) the claimed 1hr 20min runtime w/RL 3100's (which I will test later).

Also, with a light capable of this output, heat is always a potential concern but as can be seen from the graph above, the S1 doesn't get all that hot on three cells. Here is a clip covering the temp check abut 10 minuites into the runtime as well as additional details about the PVC LMD:


However, if dealing with 3 batteries are a hassle, given there is ample room in the core of the tube, one can use non-flat top 26xxx/32xxx batteries (provided they are able to withstand 5A sustained). Careful alignment of the negative side of the battery w/the prongs in the tailcap are required to ensure proper contact, otherwise there may be risk of overheating due to poor contact points:

(26650 battery shown above)

[NEW 4/23: I've now conducted runtime testing on KK 26650 and RL 3100's. So the trade off for simplicity in dealing with just one battery is seemingly just 10 minutes of total runtime but as the chart shows, also a whole lot of output.

As for the RL3100's these nickel-containing cells have a lot of staying power allowing it to reach 2.4hrs before output dropped below 10%. Albeit the latter half of the run is not regulated but again, now that these runtime charts are conducted in real measured PVC LMD lumens you can have a real understanding of just how much lumens is lost during non-regulated runs.

I'll be testing some IMR's next and will also add 32xxx cells when I pick up some in the future.]

[EDIT 4/24: Given the S1 is well regulated, it doesn't receive any added benefit by running IMR's. As a matter of fact, the low capacity of these cells really hurt the runtime in this 1S3P setup. The light only draws less than 2A for each cell. With that said, it should get interesting though on 1-2 cells runtime testing which I'll do in the future.

As a side note, the temp on the KK testing is off since the thermocouple was in direct line of the fan thus causing the erractic line. In reality, it should be closer to the other runs but again, overall the S1 doesn't really get unreasonably hot.]


PHOTO GALLERY



=======
S1 provided by XTAR for review
 
Last edited:

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
Size comparo pic and runtime testing with 1 x KK 26650 and 3 x RL3100's added.
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
(and heavy!) I've incorporated it into my weightlifting routine... ;)

AW IMR runtime and commentary now added.
 

BWX

Enlightened
Joined
Aug 18, 2007
Messages
359
Location
N.E. USA
Yeah it's big, but compared to the SR92 and SR90, the S1 puts out substantially more light, and it's lighter or similar weight and smaller than both.

I can't really compare the S1 to the Tiny Monster because the S1 has way more throw, and a little more output, but can deliver full power output for much longer without it getting too hot/ switching down to lower output. It's a good size I think, for what it can do.
 

jake25

Enlightened
Joined
May 13, 2007
Messages
864
Location
Santa Rosa, Ca
Xtar S1 outthrows SR90?? I'm soon goin compare how the Dry 3x XML vs Xtar S1 vs Spark SP6 turns out. The Dry is pretty much all flood
 

candle lamp

Flashlight Enthusiast
Joined
Jun 16, 2010
Messages
1,572
Location
South Korea
Excellent review. Thanks for your time & effort as always. Tim! :thumbsup:

I'm surprised to see the shape of the runtime graph between RDLST3100 & XTAR2600.
RDLST has much more cpapcity than XTAR and gives a longer(appr. 15~17mins) runtime. But XTAR gives a longer flat regulation than RDLST3100 than I expected.
Why do you think the main reason?
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
Thx guys!

@candlelamp, same here, that was an excellent comparo reiview you did! As for the curve, this is perfectly normal. The RL3100's contain nickle which helps increase capacity but at a slight hit to nominal voltage (3.6 vs 3.7), this is why in all my runtime testing involving these cells, you'll note that around midway into testing for non-regulated lights, LiCo cells based output will be slightly higher for a while and then suddenly drop.
 

BWX

Enlightened
Joined
Aug 18, 2007
Messages
359
Location
N.E. USA
Xtar S1 outthrows SR90?? I'm soon goin compare how the Dry 3x XML vs Xtar S1 vs Spark SP6 turns out. The Dry is pretty much all flood

I don't think it out throws the SR90, but I think it puts out more light overall... and still throws pretty damn good for a multi-emitter light. That was my impression after reading all the reviews and comparing all the beamshots I could find anyway.
 

jmpaul320

Flashlight Enthusiast
Joined
Mar 31, 2012
Messages
3,805
Location
CT, USA
what a great light... definitely on my watch list for upcoming lights i might get... for now im still playing with the tm11
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
All, it's been a bit hectic but I've managed to add material/pics to the Design & Features section. Anyways, more to come...

@BWX, the S1 will not out-throw a S90 as it's less than half the lux (60.5K measured by me vs. 126K for the SR90). However to put things in perspective, it is a multi-emitter light after all and considering that, the throw is pretty darn good. It has replaced the T40CS as the best thrower in my current collection and with over three times the total output to boot.

@jmpaul320, the TM11 is a great light but geared more towards flood w/the overall compact size and shallow reflectors. If and when you do finally get the S1, you should likely be impressed with its throw.

Cheers,
Tim
 
Last edited:

BWX

Enlightened
Joined
Aug 18, 2007
Messages
359
Location
N.E. USA
Xtar S1 outthrows SR90?? I'm soon goin compare how the Dry 3x XML vs Xtar S1 vs Spark SP6 turns out. The Dry is pretty much all flood

I don't think it out throws the SR90, but I think it puts out more light overall... and still throws pretty damn good for a multi-emitter light. That was my impression after reading all the reviews and comparing all the beamshots I could find anyway.

@BWX, the S1 will not out-throw a S90 as it's less than half the lux (60.5K measured by me vs. 126K for the SR90). However to put things in perspective, it is a multi-emitter light after all and considering that, the throw is pretty darn good. It has replaced the T40CS as the best thrower in my current collection and with over three times the total output to boot.

Yeah that's exactly what I thought.. That's good.

Hey I'm a little worried about this! (POST #41)
.....................
...................So all I can conclude is my light is not regulated like the production versions Selfbuilt and Candle Lamp received for testing.
..................
..................

http://www.candlepowerforums.com/vb...rsion-Review&p=3934191&viewfull=1#post3934191

*EDIT* post #49 in that thread makes me less worried...
 
Last edited:

jmpaul320

Flashlight Enthusiast
Joined
Mar 31, 2012
Messages
3,805
Location
CT, USA
@jmpaul320, the TM11 is a great light but geared more towards flood w/the overall compact size and shallow reflectors. If and when you do finally get the S1, you should likely be impressed with its throw.

Cheers,
Tim

its going to have to wait, after buying 5 lights in april i was forced to buy my wife an iphone to stop all of the nagging! (collateral damage lol)... but for now i am playing with my deere light aspherical for a thrower :)
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
@BWX - Regarding two of 357mag1's concerns:

- slow ramp-up on high: As per my update in the UI section I just added, this was intentional by design and is listed as no. 4 "Brightness gradually increasing to protect eyes and circuit." under the Features section on the S1 page on XTAR's website. Also check the video I uploaded demo'ing this effect on the S1 as well as their WK21.

- regulation: I have some thoughts on this but want to run additional testing to confirm my theories before I post my reply.

Cheers,
Tim
 
Last edited:

BWX

Enlightened
Joined
Aug 18, 2007
Messages
359
Location
N.E. USA
I'm much less woried after selfbuilt chimed in on post #49 anyway, but yeah I didn't realise the ramp up was that long before 357's post. If it's same on all of them, looks like that's a non-issue as well. Damn it sucks waiting for the post office..
 

357mag1

Enlightened
Joined
Nov 21, 2009
Messages
677
Location
Chesapeake, VA
- regulation: I have some thoughts on this but want to run additional testing to confirm my theories before I post my reply.

Cheers,
Tim

Tim,

I'm dying to see if you can shed any light on why we don't see tailcap current increase as the batteries lose voltage. I don't have a light meter but just did a comparison outside against my TK70 and could not see any difference in lumen output of the S1 when using freshly charged batteries or ones that had a resting voltage around 3.85volts.

This would seem to indicate the light is regulated and I know we can't change the fact that P=IE so current draw has to be increasing as the batteries drop in voltage. Now the question is why aren't we able to read this at the tailcap?
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
I've just whipped this up and should have something for you by tomorrow (mind you, I wasn't going for design awards here, so yes it's not the prettiest thing but it works) ;)



For now, my gut is that the S1 has a relatively high threshold (voltage-wise) for maintaining regulation and once it drops below that threshold, it no longer attempts to boost and perhaps goes into direct drive. Keep in mind that the light was designed to run from 1-3 cells and since XTAR couldn't possibly ensure that one would not attempt to run on just one cell, they may have set this threshold higher for safety purposes. Obviously the less cells you use, the more severe the voltage sag so perhaps this was XTAR's way of keeping tabs of how many cells are in use since it'd otherwise be impossible to tell given the 1S3P arrangement.

Another thing that supports my line of thinking is the respectively decreasing current draw on 2 and then 1 cell (XTAR row):

If the driver was programmed to continually boost regardless of v threshold, then the current draw should have gone up instead of down due to increasing voltage sag.

Also note the single King Kong cell run. It initially runs regulated but only for a very brief time, given the single cell and regardless it's pretty robust, the voltage sag must still be pretty high for it to fall out of regulation so fast.

Of course this is just conjecture on my part but with the above contraption (given your years of electrician experience I'm sure you know what it does ;)), I'll be able to graph V/A AND output for you (on 1, 2 or all 3 cells) but just not right now. More later tonight.

Cheers,
Tim
 
Last edited:

BWX

Enlightened
Joined
Aug 18, 2007
Messages
359
Location
N.E. USA
It looks like you're making it so you can switch between 1,2, or 3 cells on the fly and measure current/voltage independently on each?
 

turboBB

Flashlight Enthusiast
Joined
Jan 16, 2008
Messages
1,032
Location
NJ, USA
@BWX - Correct but more importantly is the ability to confirm that each cell is in fact making contact which would normally be very difficult when all in the tube given the 1S3P config.

@357mag1 - I graphed the runtime w/current and voltage and will convert the raw data into a chart later but for now, here's a quick vid of the results. As you can see there was no regulation with this run given the huge voltage sag and resistance introduced through all the connections in my contraption:


This would've been a whole lot simpler if I had a desktop power supply (coming soon) but it was also good to see what happened to the current draw/voltage/output when I cut one and then 2 cells out which a voltage sweep can't realistically simulate (unless one knew what the voltage sag points were which this was able to show).

Anyways will post graph later today but for now, hope that helps alleviate your concerns re: the regulation and dispel notions that reviewers got special lights :)

I'll try another run w/IMR and additional cells tomorrow to offset the voltage sag from all the resistance to see if we can graph a regulated run with this setup.

Cheers,
Tim
 

357mag1

Enlightened
Joined
Nov 21, 2009
Messages
677
Location
Chesapeake, VA
Tim,

That is an awesome amount of work you have done. I appreciate you going to all the trouble and look forward to seeing the results of your continued testing.

I have come to the conclusion the light is regulated but appears to be extremely sensitive to any resistance introduced when trying to take tailcap current measurements. You and selfbuilt both proposed that was probably the case and I was having trouble believing it. It appears you guys were right and looks like the old guy (me) can still learn something new.
 
Top