A while back I was tasked to work under a sink replacing the strainer/tailpiece. Quite an older home and the built-in counter isn’t up to our present day standards and the ‘cupboard’ space being God awfully small and with a front panel.
Space was very constrained and neither a headlight nor a standard light gave me any illumination. What I needed was a side-mounted magnetic light with swivelling head, and not too bright as I’m really close to the work.
I had this BBQ light in my drawer (the forgotten ones). This was made entirely of glass-filled plastic (I believe polystyrene) and had a 3 Volt incandescent bulb (ugh!) and a 3 x AA cartridge. So naturally, I undertook the task of re-purposing the light to my needs.
I had many setbacks, so many that I’ve lost track. At one point I just put the project aside as wasn’t justified to put so much time and energy into a plastic device. But with much perseverance, it finally came to fruition.
I had an XML2 4000ºK mounted on a copper 20mm star that just fitted nicely in the head, a single driver of unknown origin that could be adapted to the neck, and much ingenuity to carry the negative terminal to the head.
Found a Sandoz Calcium pill container – thin wall aluminium and a flat bottom. Just fitted in the body cavity, minus removing about 11 mm of length. Next is a way to have a connection point that permits the rotation of the head, and yet is easily accessible for the cell. A circular copper wire within a groove that contacts with the aluminium can. Next is a plastic cylindrical spacer to accommodate a 26650 or 18650 battery. Due to the limited depth I opted for a conical spring soldered to an embossed thin steel washer. Phosphor bronze was too stiff.
Below the head is the driver, fixed vis a top plastic plate. I opted for a cylindrical steel spring with a button cover - this to permit the head assembly to pivot about the cell’s + pole.
The driver worked on the bench, but something went fowl when I first put everything together. The aluminized reflector made intermittent contact between the MCPCB leads. The simplest of designs, but nowhere could I find an equivalent transistor. After many failed attempts of various pulls from some discarded electronic boards, I re-worked the driver board for 3 x 7135. This should give me ~1 amp and ~400 lm. Quite adequate for my needs, and I have thermal consideration as this light can’t have any radiating aluminium. The design of the ball head which swivels up/down and makes the single contact switch limits any additional heat-sinking. Also. the plastic, low-temperature polystyrene doesn’t help here. The original design of mounting to the BBQ hood side handle with this is nonsensical. Some cost-cutting I would suspect.
Unfortunately, my first attempt never had the driver work. After much analysis, and de-populating the 7135 to individually verify the regulation, I found that the cut-off serrations had some solder spill, ever so slightly. As I had re-done the traces by cutting out some of the copper, the outer negative ring made contact with the B- of the drivers. So after cleaning everything up, made some bench tests and proceeded to the assembly.
Some items will have to be changed as the magnet could be stronger for all the added weight. Tested for some 15 minutes and no plastic reek nor much heat in the ball head or the plastic lens.
If I’m happy? No, this took too much effort and being a specialty flashlight, I don’t know if and when it will be used. It works but the beam is littered with artifacts due to the lens design. I can’t remove or replace it as it holds the LED star in place.
But I’ll chalk one up as finished.
Some additional pics:
Strengthening the retaining posts in the head. Had to re-build some broken pivots (white stuff).
Contact blades (.012” thick.) had to be drilled .075” dia.
Assembled head and inner tubes w/in body.
Got to use my extended boring bar.