Luminus SBT-90 gen2 5700K tested (SBT-90-WDS-F72-SA600 SA-H4, from Neal)

I wonder how it compares to a dedomed 70.2 on output and throw?

I placed a small smooth reflector on top of the bare led that made a decent hotspot on my ceiling. I measured the tint of the hotspot at 4 currents.

current CCT duv CRI
0.1 A 5100 K +0.0164 67
0.5 A 5140 K +0.0152 67
2 A 5270 K +0.0122 67
10 A 5620 K +0.0061 67

The colour temperature is fairly neutral, at lower currents even significantly lower than Neal listed, and despite the high duv at lower currents I do not judge the tint as disturbingly green. The tint at 10 A looks quite perfect.

So these numbers are from different LEDs?

Yes…

GEN2 datasheet

Oh, it’s AR coated all right, both sides…

Thanks for the DS Jared, appreciate it!

thank you. i sent an inquiry to lantend. i’d really like a couple of these to try out. my concern was the position of the tint bin which is depicted in the data sheet as well above the bbl.

Lantend charges you the shipping, DHL at $35. So unless you’re getting several of them the shipping makes the price point pretty high. And of course, China is shut down between Chinese New Year and the Corona Virus, it would be very surprising if you got any response before Feb 8th as the government has instructed businesses to stay closed until then. Getting a package out of China may take quite a while with all that’s going on over there…

Neal has them as well, shipping should be more reasonable:

Neal is out of them, expecting to restock after Chinese New Years.

Any chance you took before and after readings?

I made other changes that pretty much make a before/after irrelevant, couldn’t know how much effect the window had as compared to the other changes.

For example, I filed off the corners of the base of the reflector to be sure the reflector didn’t short against the lead wires. I changed from 18ga Teflon coated lead wires to new 18ga Superworm leads. The missing frame and window allowed the reflector to seat further down and rest on the substrate, changing the whole scheme of things.

FWIW, it made 352Kcd before, now it makes 358Kcd. It was making 4240 lumens at start, now it shows a very brief 5180 before falling. This is in the 53mm reflector of the Crelant 7G5. It makes 4200 at 30 seconds now, cell down to 4.05V.

I removed the window from another 90.2 that’s in my SS/Cu X6, switched to an Eagle Eye X6 reflector (not threaded as this lights reflector was). When I built it the lumens at start was 5190 from 25.5A at the tail (Samsung 25S cell, Bistro driver) Lux was 106Kcd.

Now it makes 5630 lumens at start through an UCLp lens and 4880 lumens at step-down (almost 30 seconds) hot little mamma jama coming off the light box! About to get a lux reading, cell is recharging…

Edit: Now reads 116.5Kcd at start, 112.5Kcd at step-down. Serious heat being made, the big copper head on this X6 do get HOT!

So it is safe to assume that removing the window will add ±10% lumens and throw?
Is there any side effects to removing the window?

Side effect: you are going to be really careful not making any wrong movements.

Any suggestions for a 20-30A buck driver that can take 3s?

It has always been that way, when removing the dome from a normal die the bond wires are exposed and easily broken. The difference here is that a normal LED has 2 bond wires, these have 24! I have crashed the bond wires on a $70 SBT-70 before and wanted to cry, it takes just one small slip.

So yes, once that thin window is gone and the surrounding plate is removed, the bond wires are out there in the open and easily damaged. djozz has a good idea in cutting the reflector to seat around the plate (the frame the window sits on), as it really helps to focus the beam profile in most reflectors. But again, it requires some care not to twist the glued assembly off and still arrive at the same result, damaged bond wires. But this level of care is always necessary, whether it’s your first mod or your thousandth….

So I did do the mod, a D80V2 was the host, and as Dale predicted, the reflector sitting around that frame got me the correct focus, it makes a very pretty beam now.

Link to the WDYMT thread

No…not at 3volts…

I asked Lexel, for a 4S or 2S2P High current (20-25amp) 3v driver. He is working on a 2S2P 20-25amp 3v driver, he said 2S2P would be easier/better……

Anyone know if the ~10% bump is about right for de-windowing?

I did my first - modded a Luxstone X30 with a SBT90.2 de-windowed. almost sliced thru a bond wire or 2, but they appeared to be bent not broken. Wow! It's tricky. What happened to me was I used a surgical knife and seemed to be making progress, then it just gave way on a corner - should have supported the blade better - it was unexpected.

The X30 is supposedly 10 amps to the LED - had an original luminus LED in it, not sure what model. Was an easy mod, keeping the stock 2 mm thick rectangular MCPCB.

Lumens is about 3800 at start. The driver had a bank of 4 R025 1206 resistors in parallel just before the LED- wire. I added one R030 resistor, but probably didn't do all that much. Think'n of just jumping them out.

The X30 has a 2S3P battery pack using Samsung 2600 cells, so it's a buck driver. The LED wires are a nice 16 AWG. Was think'n of upgrading to 14 AWG, but not sure it will make a difference. The X30 is built like a tank, over built actually, here on Amazon, Robo819 review here. It was $600 originally!

My #'s before:

Hi: 2,632 @start, 2,560 @30 secs, throw: 152 kcd (780 meters) taken at 5 meters

Lo: 228 lumens

My #'s after:

Hi: 3,820 @start, 3,650 @30 secs, throw: 530 kcd (1456 meters) taken at 5 meters

Lo: 340 lumens

Not bad but could be much better at full power for a 94 mm head diameter light. It's a big reflector for sure.