Osram Synios DMLN31.SG LED test - (not very good, flaw found after testing)

NOTICE A testing design flaw has been discovered that is artificially inflating the cd/mm^2 numbers of the Synios LEDs. Due to the concave shape of the LED surface, additional light is being reflected into the luxmeter that is not coming from the die. The true cd/mm^2 needs to be measured either using a lens, or by blocking all areas surrounding the LED. Koef3 is rerunning some of his Synios tests to obtain proper numbers, which are closer to 200cd/mm^2. Ther Black Flat is still the best at the moment. My testing results, which used the same method that koef3 originally did, are unfortunately also incorrect :frowning: NOTICE

I wanted to double check for myself that the results from this test were correct: [LED-Test / Review] Osram SYNIOS P2720 DMLN31.SG - der neue Leuchtdichten-König! | Taschenlampen Forum
They found 297.8 cd/mm^2 which is the highest of any LED we currently know of, even higher than the Black Flat.

My test was done slightly differently, I used a fan cooled copper base and heatpipe CPU cooler:

The Synios was mounted on a Noctigon XP20 copper DTP MCPCB.
The PCB was screwed down tightly to the heatsink using two M3 screws and holes tapped into the heatsink.
Thermal Grizzly Conductonaut liquid metal thermal paste was used between the MCPCB and the heatsink.

Luxmeter was placed at 1m away, current and voltage was measured with a UT61E multimeter and UT210E clamp meter.

AMAZING results, peak of 330 cd/mm^1 likely due to the better heatsinking.
Even though it is only 4 watts, the lower the junction temperature is the more light the LED can produce.
Similar Kof3’s test (with the peak at ~0.9A) the peak for this test is at exactly 1A, any more and it starts dropping fast, as you can see from the grey dot on the graph.
This definitely confirms the values measured in Kof3’s test are true!

You can look forward to seeing my next flashlight build using this LED :slight_smile:

Nice test! Although I knew that Köf3s results were pretty accurate (he goes to a lot of effort).

Have you compared the emission profile to that of Black Flat?

Thanks for the test. - And the confirmation of my measurements, too :wink:
To be honest, this LED isn’t suitable for many flashlights, not only because of the center rings needed to get the LED die exactly in the center of the reflector. The quality of reflectors or lenses must also be as high as possible, otherwise there are lot of artifacts (rings etc) in the light pattern.

Could be interesting for little keychain pocket or low current throwers. Because of the very high thermal resistance of typ. 20 K/W I wouldn’t give the SYNIOS more than 650 - 750 mAmps.

imo this led is suitable for compact thrower, like keychain AA light with 200m throw
i would want to see some ridiculous pencil beam, maybe modding it to BLF GT

Nice test guys (Enderman and Koef3)

It is so tiny :frowning:

What I don’t like about that OSram leds is small size and annoying emitter angle looks like bitten off .

In my test linked from Enderman in OP post are pictures from a modded Tillytec light with this mentioned pencil beam.
Measurements: 11 kLux at 101 lm, LED current max. 500 mAmps.

The handling of the SYNIOS is somewhat difficult, but is possible. At this point the Luxeon Z ES is much more complicated in reflow…

In your post in the german forum:
You should replace “schwarzen Wohnung” with “Black Flat”
Google translate doesn’t know its LEDs

I tried to do a measurement of the black flat but for some reason the LED I reflowed turned blue and died so I basically spent the rest of my time re-reflowing another black flat.
Maybe I should have let it warm up a bit more since it was the first time turning on, or maybe the reflow was bad and the driver provided all 12v.
Not sure if I will have more time to test it again.

Thank you :slight_smile: your tests are super useful, I was just unsure because the lumens/mm^2 didn’t match up with the intensity when compared to the black flat.
But nope, it definitely is as good as expected :slight_smile:
I may put one in my P5r.2 to see how much throw I can get from my EDC flashlight. 750mA sounds appropriate, the cooling in a flashlight is significantly worse.

My P5r.2 produces 200lm and has 240m of throw, I wonder how much I can get with the synios!
Maybe like 300-400m which would be amazing from a tiny 14500 flashlight.
That’s a mod on my list of to-dos :slight_smile:

Thank you :slight_smile: If you use a wavien collar it makes the notch not visible :wink:



Thanks so much. :slight_smile:

In the next weeks / months I will post more tests of some LEDs here. In the german forum I’m not longer active anymore. But - the next Osram LED is waiting on my heatsink. :slight_smile:

I think you mean the Ledlenser P5R.2?

I have a lot of experience and knowledge in modding of Ledlenser flashlights. Some time ago I experimented with very small dies (XP-E / XP-C dedomed, SYNIOS red) in Ledlenser flashlights with bigger lenses (P7.2/P14.2 and MT14). In almost every case the light pattern is - especially in case of the P14.2 - bad and not really useful. There are visible rings and dark spots if the light is half-focussed.
The only really helpful solution I got is to frost the lens, so the artifacts aren’t so clearly visible. Unfortunately with this technique the throw is significantly reduced, too. :frowning:

The wavien collar projects an inverted image onto the die. That means the opposite corner of the LED is being reflected onto the notch, and since the substrate is white it is very reflective and the spot turns out to be a perfect square :slight_smile:

Yay more LED tests! Thanks!
Yeah my EDC is the led lenser P5r.2, it’s the farthest throwing flashlight of that size that I could find, and has very long runtime at max (due to the low lumen output)
I never use the light half focused though, it is always either fully unfocused or fully spot.
I love the P5r.2.
I considered putting the TIR from the led lenser into a MH12GT and then modding it with a black flat, however I don’t know if I will be able to open the MH12GT without damaging the head because they use red loctite :frowning:
That would be an amazing flashlight IMO, more throw than anything else that size with 1” head. Making a focusing system for the MH12GT would be hard though.

Interesting, thanks



Please find us decent replacement for old XP- G2… You are great loss for TLF and great win for BLF :+1:

Yeah, I will do my very best. :slight_smile:
@Enderman: Please insert the Synios Type DMLN31.SG in thread title, because there are some other Synios LED available in market, to avoid misunderstanding.

One of them (DML*Q*31.SG) I have tested yesterday - with a surprising high luminance (approx 330-350 cd/mm^2) at 450 lumens. Test coming in a few days…

Thanks for the work in the OP Enderman, I had read köf3’s test on TLF and it is great to see a confirmation. My next RS order will include some of these leds.

@koef3, I do not have an idea what made you leave TLF (I have not much impression on what happens over there) but I’m certainly delighted to see your thorough and well presented led tests here on BLF! :slight_smile:

Done :slight_smile:
Sounds like I’m going to have to order some new LEDs from mouser again…lol
Are you planning to test the DMLS too?

:slight_smile: thank you!

The DMLS is also on my to-do list, and I got this LED already. Also I measured this type, but there aren’t surprises in luminance or light flux. The thermal resistance limits the possible maximum power so I can not recommend this emitter.

But I haven’t such an amount of time to write a complete test of all SYNIOS, but I’ll see what I can do. :slight_smile:

If DMLN31.SG is about 200 lumens and 290 cd/mm^2, this new led is a real improvement for throw.

Anxious to read the test! :+1:

Can’t wait to read the test too. My very rough estimate of this LED with Khatod PL1670UN shows 35 kcd. Killer for the size.
Good to have you here koef3, welcome. :slight_smile: