3D Mag SSR-90 DD - lots of pictures

This is a light that I had built for bigchelis. It started out as a simple 3D Rebel MagLED but we had much bigger plans for it. The meager 110 lumens, as measured by bigchelis, just wasn't going to do so I gutted it out and rebuilt it. Sorry, no build pictures but this was a fairly simple build and it is modular so you should get the idea once you've seen the pictures. I'm powering it with a trio of Saft brand NiMH D-sized cells that are spec'd at 9500mAH capacity.


SSR-90, I believe it is SSR-90-W65S-R11-WN. This is the same emitter that LambdaLights uses in his Vara2000 that was measured at 2100 lumens. The stock Rebel reflector was decammed and trimmed by about 1.5mm to adjust the focus.


Here is the business end of the light with the head removed.


No switch in the usual spot. So, where's the switch?




The LED slug. That's 2.25" of 6061 T6 aluminum turned down to make it a friction fit in the Mag tube. I cut a groove down the side of the slug to simplify the wiring. I used a dab of thermal paste between the star and the heatsink.


The other end of the LED module. Those are four 24G wires from the positive connector to the positive side of the star. If you don't think that those four thin wires can carry sufficient current to the emitter to drive this at over 10A then let me tell you that four batteries will turn the LED blue with this set up so I believe that it is sufficient for the task.


The tail cap switch.

Does this light get hot? Absolutely! After five minutes the head of the light is approaching about 130 degrees F. The light coming out the front of this light is warm, too, unlike other LED lights that I have had experience with (the dual K2 beam also felt warm but not to this degree!). I'm guessing that the batteries are good for about an hour of total run time, though I would not advise that anyone try to run this light for that length of time straight. We ran this light on BC's sphere and it dropped 40 lumens over a 4 minute span but two of the positive wires had popped off at some point so that data is invalid. It made over 1600 lumens on four cells, but, now that I've resoldered the loose wire and have broken in the emitter, the LED turns blue with four cells. I'll need to remeasure this at bigchelis' house this coming Sunday.

Beam shots
The wall is painted mint green so I don't want to hear any comments about the tints of the lights as being a bit green. The distance to the wall is about six feet. The pictures of the dual K2 were taken on a different day but the SSC P7 Mag and the SSR-90 Mag were taken at the same time. White balance was set to sunlight and ISO is at 640. The dual K2 2C Mag write-up is here and was measured at about 460 OTF lumens when these pictures were taken. The SSC P7 2D Mag is a C-Bin P7 driven by a KaiDomain three level driver at 2.8A and powered by two unprotected 18650 Li-Ion batteries salvaged from a laptop battery pack. It has a diffusing lens on it since the beam is pretty bad without it. Bigchelis measured this one at 500 lumens OTF.

1/500 second @ f/29

SSC P7 2D Mag


Dual K2 2C Mag


SSR-90 3D Mag

1/500 second @ f/5.6

SSC P7 2D Mag


Dual K2 2C Mag


SSR-90 3D Mag - that hot spot is about 7" across

1/250 second @ f/5.6

SSC P7 2D Mag


Dual K2 2C Mag


SSR-90 3D mag


Surefire G2 with ThruNite reflector, XP-G R5, and 700mA driver measured at 210 OTF lumens


Surefire 9N with TLS TX3 head and ThruNite 3-mode 1.5A XP-G pill measured at 270 OTF lumens

This light is really simple, modular, and really puts out a lot of light. It doubles as a hand warmer for the colder climates and the beam itself is quite warm so you can heat up your face while you are waiting for the heatsink to warm up. Just don't look into the beam :huh:

Re: 3D Mag SSR-90 DD - lots of pictures - Stage 2 is done

Remember how I said that this light is modular? That's because it was built with the intention of having three heatsinks to test out the thermal performance of aluminum, copper, and brass. I built it with aluminum first since copper is so expensive and I could mess up the design using aluminum and not feel bad if I did it wrong.

This is what $72 of C110 round copper bar stock looks like:


The fraternal twins:

The light can be taken apart using the 2mm Allen wrench pictured plus a tool that I had made to use to push the heatsink in or out of the Mag tube since both heatsinks are a tight slip fit. The light engine part of it, the SSR-90 star wired to the positive contact, can be removed and swapped between the copper and the aluminum heatsink in a few minutes. Alternatively, the second sink can be set up with a different emitter as well. Not pictured, but also delivered with the light is a sleeve to allow C cells to be used, though the heatsink is too long to allow four C cells to fit using the modified tailcap. Still, with 3 NiMH D cells this light is very bright, in the 1400 lumen range, and will stay upwards of 1200 lumens until it becomes too hot to hold.

impressive heatsink!

is this light pretty much fixed focus? how much spill is there?

just curious as i am interested in building an SSR-90 mag myself.

Most everyone will tell you that a Mag is only good for one focus and that is as tight as possible.

The spill probably is in the 100 lumens range. I was shining this light at buildings after work the other day and I was able to light up buildings easily over 100 yards away but it lights up everything else, too! I'd get outdoor beamshots but it's been foggy at my house for the past few weeks so I can't really do it.

When we tested the light last week it was still doing a steady 1550~1600ish for a complete 4 minute runtime.

It gets hot as expected and we saw 4.88A at the tail with 4 NiMH C cells, which as PCC mentioned we will probably not be able to do anymore cause of the tint turning blue.

The SSR-90 emitters: Kevin from Lamdalights sent to me. I couldn't get these SSR-90 65k anywhere, but alot of 57k alternatives.

With the huge heatsink the mag switch had to be omitted as you saw.

OTF lumens will be pending Now

Those are great pictures of your build. I am trying to get parts together to make a similar light. Could you please tell me where I could purchase the rebel reflectors?

Starlight said:

Those are great pictures of your build. I am trying to get parts together to make a similar light. Could you please tell me where I could purchase the rebel reflectors?

Click to expand...

The Maglight started life as a 3D Rebel LED Mag that did 110 OTF lumens. So, go to Costco and get the $29 package which includes a mini mag too.

Well, I got this one with the light. I'm not sure where you would get one separately.

Second heatsink made and the light has been delivered. If I ever get a large brass slug to use for a third heatsink then I'll make one more for further testing.

SWEET!! Nice work :thumbsup:

And let's not forget - ......

Hi all,

I used a 2D Mag Rebel hosts instead of the black 3D you see in the pictures. This was due to keep the total power under 30 WATTS.

• 2D Mag Rebel hosts
• Stock rebel reflector cut to fit
• Copper heatsink
• 3 Tenergy NiMH C cells topped off at 1.43V each.
• 6.6A at the tail
• small magnet as tailcap/spring = less resistance (i think)
• AS5 as thermal paste

Here are the Copper readings given the test conditions outlined above:

1 SEC = 2247.7
30 SEC = 2217.9
1 MIN = 2133.0
2 MIN = 2025.2
3 MIN = 2013.8
4 MIN = 1970.2

:twothumbs:twothumbs:twothumbs:twothumbs:twothumbs

@bigchelis

What is the result of your Emitter swap experiment?

aurum said:

@bigchelis

What is the result of your Emitter swap experiment?

Click to expand...

So far I just have the aluminum and copper heatsinks in my possesion.

Yesterday I was trying to do the readings with 4 NiMH C cells, but the current was almost 10A (50 WATT, no load) at the tail, so I backed off and decided to stick with just 3 NIMH cells and I got alot better OTF results.

Next:

I have to test the OTF lumens on the aluminum copper. Then I have to get some brass and maybe PCC can make some time to build that one too, which will be the 3rd replica.

bigC

Nice build!! How much does the copper heatsink weigh?

@ bigchelis: Also nice results, seems like you got a LED from the top end of the highest brightness bin. But before swapping to the Al heatsink don't forget to measure current while measuring brightness .

Der Wichtel said:

Nice build!! How much does the copper heatsink weigh?

@ bigchelis: Also nice results, seems like you got a LED from the top end of the highest brightness bin. But before swapping to the Al heatsink don't forget to measure current while measuring brightness .

Click to expand...

I have a Mag Tailcap with the bottom opened up where I can wire the multi-meter to the springs while its under load. I was going to do it last night, but the 2D Mag is top heavy and I need another set of hands. I will see if I can get help. Another issue I have is that I didn't use a spring in a 2D Mag. I used a little magnet cause the 3rd NiMH C cell fits inside the D tailcap.

I can do this if I switch back to a 3D Mag and use a spring.

Wow, that's really really high for only 6.6A.

The Vara2000/Copper I bought from you that you had similar OTF measurement previously, I measured the tail cap current @ 9.87 AMP.

You got lucky with a super emitter or something?

Congrats!

bigchelis said:

Hi all,

I used a 2D Mag Rebel hosts instead of the black 3D you see in the pictures. This was due to keep the total power under 30 WATTS.

• 2D Mag Rebel hosts
• Stock rebel reflector cut to fit
• Copper heatsink
• 3 Tenergy NiMH C cells topped off at 1.43V each.
• 6.6A at the tail
• small magnet as tailcap/spring = less resistance (i think)
• AS5 as thermal paste

Here are the Copper readings given the test conditions outlined above:

1 SEC = 2247.7
30 SEC = 2217.9
1 MIN = 2133.0
2 MIN = 2025.2
3 MIN = 2013.8
4 MIN = 1970.2

:twothumbs:twothumbs:twothumbs:twothumbs:twothumbs

Click to expand...

ma_sha1 said:

Wow, that's really really high for only 6.6A.

The Vara2000/Copper I bought from you that you had similar OTF measurement previously, I measured the tail cap current @ 9.87 AMP.

You got lucky with a super emitter or something?

Congrats!

Click to expand...

The current was not that great in the beginning. Overdrive it for a couple cold/hot cycles which will drastically change the vF of that emitter.

When the emitter was fresh it was only doing 4.88A at the tail off 4 NiMH Tenergy C cells. Now; with the very same 4 NiMH C cells it does 10A +/- . The new super high current happened after PCC and I broke in the emitter with both low current and high current. I actually got 2400ish OTF off 4 NIMH C cells and from 1~3 minutes the lumens were 1500ish and the light was way way way way too hot to hold. This is why I backed off to 3 NiMH C cells.

Please note: driving these SSR-90's at 50 watts was a risk I was willing to take because I felt it had to be done. Your light with less heatsink will or might turn-blue and bye bye SSR-90

When building this light I used a trio of used D NiMH cells that had been abused in a previous life. They work well enough to give about 3A of output to the SSR-90 but not much more than that. With better NiMH cells you can get better results as BC has shown.

In initial testing at BC's place using four Tenergy C NiMH cells that light got really hot. Hotter than I would be comfortable running any light at knowingly. Making this light was educational for me as well.

I really want to see some results on the copper vs aluminum heatsink. I suspect the only difference you will notice is the weight

Mac

Der Wichtel said:

Nice build!! How much does the copper heatsink weigh?

Click to expand...

I didn't weigh it. I know that it is about 1.34" X 2.25" C110 copper with some chunks milled out of it. That would give you about 3.17 cubic inches. C110 copper weighs 0.323 pounds per cubic inch (http://www.suppliersonline.com/propertypages/C11000.asp) so that's about one pound, or 2.205kg. I'd say that it's probably closer to 15oz or about 2.05kg but these are guesses.

Mac, I suspect that both heatsinks will give the same OTF lumens until the heatsinks become heat saturated and the lumens start to drop because of the heat. I believe that the copper will take longer to reach this stage than aluminum but this is only a guess. I know that it seemed to me that this light became warm to the touch after running only one minute with those beat up D cells at about 3A using the aluminum heatsink. When I finished ther copper heatsink I ran it for about two or three minutes and it didn't seem to be warm but I could have run it for less time than I thought or the batteries may have had less charge and initial voltage, resulting in lower output than my previous tests. None of that matters now that the LED has been baked a few more times and the Vf of it has changed. The most scientific way to test it, of course, is to use a bench power supply to power the LED at a regulated voltage and current then measure the output from the light to see how long it takes for the output to drop due to heat, all the while measuring the temperature of the light. Neither of us has a regulated bench power supply so we can't do this but I do have an infrared thermometer that I forgot to bring over to BC's house yesterday.