I started this back in June when I glued up the 8 segments that make up the sides of the central octagon body. I had the general idea for some time. This took a long time to get to completion in part because I could not decide on several details. What type of cell; I changed back and forth between 14500, 16340, 18350, and 186500 or 18650 before I made the final choice of a 16650. Anyhow, I did get it completed in time to be given as a Christmas gift to my wife and best friend.
I did not take a lot of photos during the building process, but I have a few which I shall include here along with some description if anyone is interested.
First, the completed light will be shown….
Here it is with a cup of coffee for scale… (images are clickable to enlarge)
…and here is the MIDI along with the previously shown MINI…
The triple mcpcb with Samsung LH351D emitters were planned from the initial idea; 2700K in 90 CRI. I eventually decided on a simple 4*7135 Nanjg driver loaded with a custom 5 level firmware, Lo thru Hi, with memory because it is simple to understand and use.
Early on in the planning I decided to use one of the clear acrylic small jars I had leftover from my mod of an SRK into a lantern.
LINK TO THAT…… Inspired by DBSAR, my SRK-based Lantern
A heat sink was needed so a leftover from another project was planned to be used. I wanted the emitter facing downwards like the BLF LT1.
Where to place the driver took a long while to decide. The driver also needed to have heatsinking and the wood body was not very thermally efficient as a heatsink. I decided to place it in the upper portion of the head, above the heatsink along with the mcpcb. A test indicated the heat would travel into the finned aluminum heatsink even though the sink was below the heat producing parts.
The placement of the switch also took some thought. Possible locations were in the lamp base, the side of the body (like an LT1) or in the head. I chose the head location as it would be close to the driver and emitters and if located in the top surface would take a simple finger press to operate.
I did take some pictures of the head unit. The Wakefield heat sinks come with an assortment of extruded in place locations for mcpsb mounting screws. Those may be tapped with 4-40 threds where desired to mount things. They are meant for metric (M3 ?) fasteners but the 4-40 UNC work.
The short length of wood dowel in double stick taped to the Omten 1288 switch. It is used as a spacer to hold the switch in position.
The small acrylic jar used as a globe comes with an aluminum twist on lid. Holes were drilled and the mcpcb mounted to the upper surface of the hetsink with the lid. This is not a high powered light so I could have used a single emitter. I chose a triple as that left the center hole in the mcpcb for a screw to attach the Ledil diffusing optic I wanted to use. That optic was another left over part. I often buy parts like that in multiples just in case I mess up the first one.
Above is the trial fit with wood shim spacers I used when soldering the wires to the mcpcb.
The next image shows the Ledil optic fitted with a long screw through the mcpcb and lid into a hole drilled and tapped in the center of the heatsink.
Here’s the upper assembly of lid, mcpcb, driver, switch and heatsink.
That all fits inside the octagon top knob, but alas I did not take any photos. Let’s just say there was a lot of drilling out of wood as well as a lot of chiseling and carving knife use. A lot of trial and fitting to carve out hollows for the switcha nd driver and leave enough wood to be sturdy.
There is a big gap in time and photos of parts as they were made. I was not thinking ahead to showing the project. I was too engrossed im making the idea work and changing things as I went along.
The light body is pictured in the upper left of the above image. The main part is an octagon made from 8 slices of glued up stock that was left over from some other projects. The octagonal shaped base is glued up stock with a hole drilled in the center and then cut roughly to shape with a coping saw, then filed to final size. The white colored wood is a maple block that was drilled to accept a 16650 cell. The copper pieces are a mount and heat sink for the 1 amp charger board. The board is cemented to the ciopper with a small amount of my remaining Arctic Alumina epoxy.
The upper right assembly in that image is the top knob that houses the driver and switch assembly. The heatsink is not visible there. The disc is an octagonal shade, glued up from thin wood stock. I hole is drilled in the center and the shade slips over the heatsink.
The bottom of the acrylic jar was drilled out. Inside the jar is a small sheet of Shoji lamp diffuser material, which in the end product was not used. I like it but it made it difficult to see the charger indicator led’s as well as causing more light loss than I was happy with.
The disc in the lower left of the above image is a plate that will be screwed onto the bottom of the body unit to enclose the unner parts.
The wood part with the sroing and wire is the lower (negative) contact for the pictured 16650 cell.
The other smaller wood disc is a support piece for the globe. In the foreground of that image is a cup with screws and the USB-C port that will be used.
The next image shows the parts again with some repositioned for a different view.
The next two images show the assembled light with USB cord plugged in. First with the red charging indicator lit and then with the green ready indicator lit.
The final images are closeups.
and my username is an old one going back to when I worked in graphic design. I still miss all the gorgeous papers I got to work with. In my current work (medicine) everything is printed on the cheapest possible generic printer paper 
