Techjunkie
Enlightened
I received my 4500K SST-90 emitter from Avnet today. The packaging was laughable. It came in a box about the size of a shoe box. Buried under much bubble wrap was an anti moisture bag with very specific handling instrutions. Within that was a large bag of dessicant, a moisture indicator card and finally what I'd call a jewelry box if it was designed by the military.
I used Jo's BrightLumens deluxe SST-90 Mag D heatsink. It does the trick, but having done it once, I'm not looking forward to doing it again. I can't think of a better way to mount the emitter, but it was nerve racking working on a $50 LED that's so delicate and so hard to mount.
I used the fattest wire that would possibly fit through the holes, 20 gauge stranded copper, and ran into trouble getting the fitment right. I had the emitter heatsinked by resting it upside down on top of the jaws of my vise, and that made the solder cool immediately on the pads. The thick pos wire was encroaching on the space of the center just a hair, but it was enough to cause it to sit improperly in the heatsink. My iron wasn't hot enough to melt the solder again and in the end I carfully used a dremel with a cut-off disc to shave back some of the solder so it would sit right. The other complication is getting the thick wire laden with solder bent just the right way to position it so it goes straight down the hole from the corner of the LED. Pics below explain it better than I can here.
I used AAA in the wire troths and Arctic silver thermal adhesive under the main pad. Adding the AAA to heatsink the wires to the HS is something I always do to improve thermal management. It's that much more surface area to transfer heat with.
As for reflectors, I did what I usually do and searched for the cheapest possilble solution. I experimented with the DX P7 reflector (sku 12229) and managed to make quite a mess of it with the dremel. Even more scary than the soldering, I had to use the dremel to cut a camphered edge to the inside of the heatsink after the emitter was already monunted.
The DX P7 reflector actually gave a much better (tighter / more intense) spot than the KD V2 reflector that I opened to 16mm. I had to raise the heatsink about 2mm to get the short DX reflector to focus properly. I was surprised to see a spot about the same size as given to the smaller MC-E emitter by other 53mm reflectors I've used for MC-Es.
I played around with some other refelctor/optic options, but I'm not done yet. I tried the 40mm+ deep reflector that comes inside the Mag 2D LED after cutting off the cam and opening the hole to 16mm. That gave the best focus yet, but I couldn't screw the head down passed the O-ring. If I cut the lip off the heatsink, I might be able to drop it down the tube until it sits on the Mag switch. That's about a half an inch, which might be just enough to make that reflector fit.
I tried the KD aspheric but what I found to be even better (much to my surprise/delight) was the $4 DX aspheric sku 12834. It's glass, not plastic, only 18mm deep and while it does not perform as well with the CREE XR-E, it seems to be better suited to the SST-90. Perhaps the difference in the domes affects the performance with the optic.
Oh, by the way, until the Sector_Cleared driver is available, I'm running the emitter direct drive on 2 parallel Sony 26650VT with a slapped together carrier that I described here.
I had to add copper braid to the tail spring. Tailcap mearurements were only ~3.5A on the bare spring, but over 7A on the braid. The Vf of the emitter drops as it heats up, but the batteries sag after use. The highest I've measured so far at the tail was 7.95A. The warmer color SST-90 must have a pretty high Vf. Also, I tried some slightly used NiMH C cells. Three yielded a 5-6 amp draw and with four cells current draw shot up to 11A before I quickly disconnected the leads. Maybe a 4s2p pack of AAs might be the direct draw sweet spot for the SST-90, assuming 2pAA sags more than 1C. I may just have to find out.
(update: after replacing the spring/braid with a cut down "golden" mag spring from KD and using deoxit on all the contacts, I'm measuring 8.5A draw at the tail and 8+ through the spring)
Until some of my other cheap reflectors come in, I'm keeping th DX aspheric in it. It's freaky fun - much larger and also more intense than the KD Q5 + Aspheric kit. I'll add beamshots later this week.
Update: finally experimented with 4s2p AA - see post # 25 below
Update #2: New reflector experiments. Stopped-down photos show improvement with a reflector that actually fits, post #28
Update #3: the experimentation and aggressive burn-in has significantly reduced the Vf of the emitter see post #31
I used Jo's BrightLumens deluxe SST-90 Mag D heatsink. It does the trick, but having done it once, I'm not looking forward to doing it again. I can't think of a better way to mount the emitter, but it was nerve racking working on a $50 LED that's so delicate and so hard to mount.
I used the fattest wire that would possibly fit through the holes, 20 gauge stranded copper, and ran into trouble getting the fitment right. I had the emitter heatsinked by resting it upside down on top of the jaws of my vise, and that made the solder cool immediately on the pads. The thick pos wire was encroaching on the space of the center just a hair, but it was enough to cause it to sit improperly in the heatsink. My iron wasn't hot enough to melt the solder again and in the end I carfully used a dremel with a cut-off disc to shave back some of the solder so it would sit right. The other complication is getting the thick wire laden with solder bent just the right way to position it so it goes straight down the hole from the corner of the LED. Pics below explain it better than I can here.
I used AAA in the wire troths and Arctic silver thermal adhesive under the main pad. Adding the AAA to heatsink the wires to the HS is something I always do to improve thermal management. It's that much more surface area to transfer heat with.
As for reflectors, I did what I usually do and searched for the cheapest possilble solution. I experimented with the DX P7 reflector (sku 12229) and managed to make quite a mess of it with the dremel. Even more scary than the soldering, I had to use the dremel to cut a camphered edge to the inside of the heatsink after the emitter was already monunted.
The DX P7 reflector actually gave a much better (tighter / more intense) spot than the KD V2 reflector that I opened to 16mm. I had to raise the heatsink about 2mm to get the short DX reflector to focus properly. I was surprised to see a spot about the same size as given to the smaller MC-E emitter by other 53mm reflectors I've used for MC-Es.
I played around with some other refelctor/optic options, but I'm not done yet. I tried the 40mm+ deep reflector that comes inside the Mag 2D LED after cutting off the cam and opening the hole to 16mm. That gave the best focus yet, but I couldn't screw the head down passed the O-ring. If I cut the lip off the heatsink, I might be able to drop it down the tube until it sits on the Mag switch. That's about a half an inch, which might be just enough to make that reflector fit.
I tried the KD aspheric but what I found to be even better (much to my surprise/delight) was the $4 DX aspheric sku 12834. It's glass, not plastic, only 18mm deep and while it does not perform as well with the CREE XR-E, it seems to be better suited to the SST-90. Perhaps the difference in the domes affects the performance with the optic.
Oh, by the way, until the Sector_Cleared driver is available, I'm running the emitter direct drive on 2 parallel Sony 26650VT with a slapped together carrier that I described here.
I had to add copper braid to the tail spring. Tailcap mearurements were only ~3.5A on the bare spring, but over 7A on the braid. The Vf of the emitter drops as it heats up, but the batteries sag after use. The highest I've measured so far at the tail was 7.95A. The warmer color SST-90 must have a pretty high Vf. Also, I tried some slightly used NiMH C cells. Three yielded a 5-6 amp draw and with four cells current draw shot up to 11A before I quickly disconnected the leads. Maybe a 4s2p pack of AAs might be the direct draw sweet spot for the SST-90, assuming 2pAA sags more than 1C. I may just have to find out.
(update: after replacing the spring/braid with a cut down "golden" mag spring from KD and using deoxit on all the contacts, I'm measuring 8.5A draw at the tail and 8+ through the spring)
Until some of my other cheap reflectors come in, I'm keeping th DX aspheric in it. It's freaky fun - much larger and also more intense than the KD Q5 + Aspheric kit. I'll add beamshots later this week.
Update: finally experimented with 4s2p AA - see post # 25 below
Update #2: New reflector experiments. Stopped-down photos show improvement with a reflector that actually fits, post #28
Update #3: the experimentation and aggressive burn-in has significantly reduced the Vf of the emitter see post #31
Last edited:
