Originally these are single mode lights with potted internals and something like a P4/7 emitter in a plastic reflector & window.XP-L HI 5A3 4000K
MosX ceramic insulation 20mm MCPCB
OP aluminium reflector
ARC Glass
LD-A4 3A driver
Hand made brass pill & copper contact
Max 560lm (3A CC)
I used some brass tubing, plate and copper to make a custom pill and new positive contact for this light. The original guts were epoxied in. A major problem with this upgrade was how to rebuild it with a modern driver, pill, contact and reflector.The finished pill slides in snug and is designed to be removable for future upgrades. If I had a lathe some threading would have been nice here.The head has a copper sleeve soldered into the hole I drilled then a slightly longer copper bar with a small driver spring soldered on top is epoxied into the sleeve with JB weld, leaving a uniform 0.8mm gap around the positive contact.
Around 560lm OTF at 3A current controlled from the XP-L HI 4000K
Some led-swaps today. I wanted to use the blue and red Sofirn SP10B lights, but after finishing the blue one with a Samsung LH351D (4000K 90+CRI), I decided to not use the red SP10B for the second led, a Nichia E21A 4000K R9080 on a djozz-119board, but a Utorch UT10 because everything about it seems a tiny bit better to me, a tiny bit smaller, a tiny bit better UI, a tiny bit better output, and (more than a tiny bit) better mode spacing.
It took a while to center the E21A led to an acceptable level (slightly off-center but the beam is perfect) because there is no center piece for this led.
At this point, having used the LH351D in a variety of hosts, Iām not too enthousiastic using them in small TIR optics or small reflectors, the beam gets too floody to my liking. It was fine in the larger SP33 reflector (31 mm smooth) but in a E2L (Carclo triple TIR) and the SP10B (18mm OP reflector) the hotspot is too wide. In contrast, the E21A is perfect for these small hosts with its small emitting area (even smaller than a dedomed XP-G2).
Beams. Left blue SP10B, right UT01
The blue SP10B build is really nice but for me, the UT01 build with E21A wins today, comparable in output with the SP10B+LH351D (213lm for the Utorch, 270 for the Sofirn), but even nicer beam and tint.
Just made a led swap in a Nitecore SRT9 from cool white XHP50 to 5000K XHP50.2
3D printed a tool for the bezel:
And opened it up:
Replacing done:
And everytinhg back together:
Gained 227 lumens. Now it is making 2379 lumens.
And it has perfect hot spot and much nicer tint.
Making a 3D printed tool for the bezel. That is a skill/service a lot of members are waiting for.
I myself prefer to use the TEVA-method. Itās the smell of my feet, I think.
Zozz, when you cut down an MCPCB like that itās always a good idea to slightly bevel the edge. This pulls the power carrying traces away from the outer edge and gives some clearance that can help avoid a short circuit.
Dale is correct, the flare could be a hidden problem later. I find its easier to use 90Ā° single flute countersink end mill on a rotary table or simply a flat sharp diamond file.
Thanks! Done it. And checked with multimeter and with microscope too. The image is just not the final version. I also scraped off some solder mask to make wire soldering points bigger.
I also unmodded my car yesterday. Pulled out one alarm system and one immobilizer. Piece of wiring nightmare. I donāt know why so popular these aftermarket things here. They only make contact and electrical problems and making troubleshooting worse.
XP-L HI 5A3 4000K
The finished pill slides in snug and is designed to be removable for future upgrades. If I had a lathe some threading would have been nice here.
The head has a copper sleeve soldered into the hole I drilled then a slightly longer copper bar with a small driver spring soldered on top is epoxied into the sleeve with JB weld, leaving a uniform 0.8mm gap around the positive contact.
