On sale now.
See OP for pricing. 
The emitters I needed showed up today.
I made a first attempt at putting everything together, and ZOMG thereās not much room available! How do you get everything to fit?
Iām guessing Iāll have to use basically the shortest wires possible from the driver to the emitters, and will need to route them so that they donāt overlap any of the chips at all. I may also need to remove the spring from the driver or cut it way down.
As for emitters, one of my four carclo optics actually fits onto domed XP-Ls, though the other three donāt. Itās still just slightly raised, but might work. Otherwise, a de-domed setup should be fine (though the tint is uncomfortably warm).
The triple XP-L has pre-soldered leads, and so does the BLF17DD driver. So far I tried cutting them both pretty short and splicing them, with kapton tape over the joint. It still doesnāt quite fit though. The driver floats a bit above the pill, and I wasnāt able to get the light screwed together fully with a battery inside. I did at least confirm that my firmware works though, and that I didnāt have any electrical shorts.
How did other people get everything to fit?
I may of missed it but which Host are you using? I found there was plenty of room in my Ti-XC Host. I wouldnāt bother with splicing the wires for the driver to emitters, just remove the wires on the driver and solder the wires from the emitter straight on to it.
Hi,
I had a bit of a time getting my light with a Tri-EDC (with triple and BLF17DD) put together, and I posted a thread about it with info that I hope might help:
If you are using a Tri-EDC host, it definitely is not easy, as the pill is very short.
Good luck,
Jim
Thanks, ohaya. It looks like you ran into exactly the same issues I did.
I already cut out a groove on each side of the pill to screw it in with, and I routed the wires separately around the driver chips, but I still need to reduce the spring as far as possible. That should buy me 2-3mm, I think, which is probably enough.
Cutting down the spring worked to get parts to fit in this Cypreus Tri-EDC. I used a dremel with a cutting stone to cut about two thirds of the spring off, then re-attached the wire braid at an appropriate spot. Now it goes together fairly easily without denting the battery. When compressed, the spring is just barely thicker than the other components on the driver.
Itās too bad the de-domed XP-L 2C is so yellow. Also, every Carclo optic Iāve tried except the 10511 (wide, frosted) has quite a lot of beam artifacts at 120 degree intervals. Iām using a 10507 (narrow spot, smooth) for now, the one which isnāt flat on the front. Brightest spot, but also the most beam artifacts.
One of the four Carclo optics actually worked on a domed triple XP-L 3D, but the tint wasnāt much better. The tint varied throughout the beam, probably because the optics squashed the domes a bit and left the base exposed. So, I switched to de-domed 2C instead.
In any case, I just did a side-by-side comparison of my custom modded SRK (~3000 lumens or so) and this just-assembled Cypreus Tri-EDC. They appear to be roughly the same brightness. However, my light box canāt seem to measure either one of them. It freaks out about the tint or something, even at fairly low levels like 500 lumens, and gives me an error.
On the low end, I have the Cypreus spitting out about 0.3 or 0.4 lumens on moon on a full cell, using fast PWM=0 with a PFM level of 40. So, fast PWM=0/40. Itās significantly dimmer at 0/255, or brighter at 0/10, but 0/40 seems about right for a moon mode. And when the voltage drops, the moon mode gets dramatically less brightā¦ like 0.001 lumen at 3.6V.
I put my brand new custom off-time firmware on the MCU, and it gives me the following modes: (mode spacing chosen to be visually linear on a cube-root scale except for moon, but I havenāt measured yet)
- 1: Moon: ~0.3 lm
- 2: Low: ~11 lm
- 3: Med:
- 4: High:
- 5: Higher:
- 6: Maximum: ~3000 lm (guessing, not yet measured)
- 7: Moon-med 1Hz flasher (āstutterā at med, then rest at moon, repeat)
- 8: Low-High 1Hz flasher
- 9: Med-Higher 1Hz flasher
- 10: High-Max 1Hz flasher (should probably remove this)
- 11: Heartbeat beacon, two brief flashes in a 1Hz cycle
- 12: 10Hz strobe (motion-freezing, not tactical)
- 13: 24Hz strobe (motion-freezing)
- 14: 60Hz strobe (motion-freezing)
- 15: Battery check mode
- ā1: Maximum / āimpressā mode
- ā2: 24Hz strobe
- ā3: Battery check mode
A short press advances the mode by one, a medium press goes backward one mode, and a long press resets to the first mode. The ānegativeā modes can be accessed only by going backward from moon. This lets me access turbo very quickly (click, medium press), and also gets me my favorite strobe and battery check modes without going through all the others. A short press from any ānegativeā mode goes back to moon.
I left the driver loose so I can reflash it without too much hassle. This is just a first draft of the firmware.
Hey Sinner, Iām interested in the 80mm tri-edc but I donāt have a paypal acct. Can I use a credit card to make a payment to your paypal acct?
See, ToyKeeper, you stuff all that into the MCU and then itās so fat the light canāt get itās pants onā¦
Glad it all worked out ToyKeeper, I have been fixing things lately and this was one of the details i had to go through and re-machine some hosts , since there are only 4 maybe 5 hosts left of this design, I will be doing standard copper pill instead of a lower profile ones and also some minor upgrades for my next tri-edc 18650 light, more like 113.5mm seriesā¦lol
@drillbitz11 : I can accept paypal at this time, i am pretty sure you can pay with credit cards if send you a paypal invoice.
thanks.
This one is 102mm long. I think it should still be do-able in that size, but the pill seems about 1mm to 2mm short. Iād just extend the pill by about that far, and make the driver shelf a bit deeper. Possibly reduce the thickness of the bezel and tailcap by about 1mm each to give more room inside without making the light any longer.
Cutting the spring isnāt really an issue, considering that RMMās FET drivers come with a pretty huge spring. Itās just hard to get the pill together (even with short leads) since itās shallow. Gotta route the wires around the chips as much as possible instead of above them, which can only be done to a small extent since they need to go through the middle hole. And by default, the driverās leads were soldered to go toward each other. I had to reverse that so they aimed apart from each other.
In any case, it all worked out. Maybe I just stuffed the MCU too full after all.
Edit: Oh, it would be nice to have room for a lens though. These plastic optics are easier to scratch/dent, and the 10507 picks up dirt easily since itās not flat on the front.
Tell us how you like it? Bad boy rocks donāt it!
Tight for a single pair of wires?! Lol.
I fit 5 LED leadās in mine to the individually addressable 20mm quad MCPCB- an 18AWG positive and 4 22AWG - leadās, talk about a tight fit. And all in the standard pill with no extension / spacer.
C-K your intellect surpases the genius-level, besides, legend says only a handful will have the ability to have harnessed the power of cereals , oatmeals in particular. 
Five leads on a 100mm Tri-EDC 18650 host? The pill on those is particularly thin. It sounds like the other hosts have more space in the pill.
On mine, the threads also seem pretty loose, so Iām planning to fill as much of the gap as I can with thermal grease when I get the optics and firmware and such near a final revision. Fortunately I donāt have to take it apart much, since I have an off-time dev host for most of the testing.
Itās pretty sweet. I need to take a picture to show how much it rocks. Iām picturing a Skyray King next to the Cypreus, and an equal sign between them. It really is a SRK in host about 1/5th the size.
(133mm * 59mm * 40mm) / (102mm * 24mm * 24mm) = 5.34, SRK is 5.34X the size of the Tri-EDC
Iām still debating if I want to use 3x219B instead of 3xXP-L, but I think the de-domed 2Cās ~4000K yellow tint is actually appropriate coming from a solid chunk of copper. I prefer ~5000K for general purposes, but perhaps I can make an exception this time. I mean, itās a photon grenadeā¦ the finer points of tint snobbery arenāt very relevant.
Besides I can use my triple-219B Roche F6 for that; it puts out 1800 lumens and is more durable for EDC purposes.
Yes and not to forget it was passed to you from your grand parents on your birthday.
Hey Sinner, She likes it!
How could you not? The sheer output from something that small, that shiny and gorgeous, and even the bit of work it took to cram it all together into one deviceā¦ya just gotta love it!
Haha, I need to get my story straight. I think, when my grandfather Cyrus McBadass died, he give me and my sister his tools for killing the undead. She got his holy flintlock revolver and I got his portable sun. Weāll work together to continue the family legacy as vampire hunters.
Bad news for meā¦ I hate yellow-ish tint. My triples are arriving tomorrow and I already feel Iāve dumped $28 ($20 + $8 dedome service).
Should I order another one and use CKās optic cutting method to fit the XP-L?
Geez, at least try it first.
All you have to do is put the optic on, hold the leads up to the ends of a small weak li-ion cell, and check how the beam will look.
My de-domed 2C appears to be in the range of 4000K to 4200K color temperature. Significantly warmer than my 4500K Nichia 219As, and just a hair warmer and less green than my ZL H51w (specād at 4200K). Compared to incandescent lights it still looks quite cool.
However, IIRC, all methods of dome reduction end up warming up the tint. I was just surprised that it warmed up by so much ā about 1500K warmer after de-doming. I suppose that should be expected though, since a dedomed 1A turns into something like a 3B. It makes sense that a 2C would end up somewhere near 5B or 5S, but I hadnāt thought about it.
