3rd Annual Scratch Build - Machine Built Division DB Custom, part 1

Ah, did I mention that it has a driver built by me as well? It’s the FET+1 with ToyKeepers A6 firmware. 2 groups… 7 levels/4 levels… reversing with a long press and a hidden group with Strobe, batt check and bicycle strobe, the first of the hidden modes is Turbo so when the light is turned on at moon a single longer press takes it directly to turbo. Very useful, my favorite UI for quite a while now.

Even the switch that I got from an A6 has been modified with an Omten and a pcb pass through with a single 18ga wire from the top of the spring directly to the switch itself. The driver spring has a copper disc soldered to the spring bypass wire for a solid flat plane connection to the cell, regardless of flat or button top.

Guess I need that beamshot now… about to go out for the Thursday night walk to take the trash can to the road so I’ll get that shot first. :slight_smile:

I’ll be back…

Standard set, Canon G1X at 1600 ISO f/5.6 at 1/2 second exposure, manual focus with manual exposure settings, 2 second timer on a tripod. Red oil drum is at 97 yds in the first pic, which is at 28mm equivalent. Second shot shows the white barn, same settings except for f/5.8 and zoomed in to 112mm equivalent.

That looks to be the best thrower pics I’ve seen of that drum.

I was thinking the same thing.

Well done Dale! Looks like big things have happened while I took time off from flashlights. Feel free to hit me up if you need any advice.


Thanks Buck, appreciate it.

Nice to see you back, Buck. I’ve missed your videos.

Thanks RBD!

Here's one I just made this weekend. Imagine the possibilities :)

I was under the impression that typically one imports a 3D model into a CAM program configured to suit one’s cnc, specify offsets, preset cutting tools and a machining strategy, hits enter and have G-code in seconds, are people actually hand writing G-code for anything other than most basic operations in 2.5D machining ?

Very nice work BTW.

Who creates the 3D model? Who proofs it? Who transfers it to code and proofs that to ensure all will flow smoothly? Who loads the cutting tools into the machine? How easy is it to get that wrong? I mean, if the tool is not in the machine at the proper depth… And what happens if a flute breaks off a carbide bit? Who’s monitoring that? Will the new tool be an exact fit for the broken tool? Are those people (all or one) sweating bullets while waiting to hear the ominous crash signifying someone missed something? That nightmarish sound that tells these folks they will soon be looking for a new job because that expensive machinery just crashed? There is honest sweat and there is nervous sweat, and I can truthfully say that an honest sweat is much easier to deal with. Air conditioning or not.

It’s beautiful when it all works together, but parts fail, chuck jaws even break on occasion. Crash! Bearings sieze, Crash! The lubricant gets clogged in the delivery hose, or those little individual knubs on the coolant hose can pop apart with no warning, spewing coolant everywhere and the part is suddenly getting hot, carbide snaps like peanut shells on a steakhouse floor. Murphy was a smart man, anything that Can go wrong, Will go wrong. I don’t want to be there when stuff like that hits the big proverbial fan…

I didn’t like it when my 3” piece of 6061 flung out and fell in the chip tray. Plenty of drama for me, with no big crash involved at all!

Here i was thinking you were talking about someone writing g-code for hours while you were talking about the whole workflow from 3D design to finished product, how stupid i can be, very much it seems :nerd_face:

Dan (TexasLumens) writes the code for custom lights he makes. He has told me it can take 17-18 hours just to write the code for a complex light.

David, I write my own G-code for some of the things I do. I have CAD and CAM (sort of), but for many of the simpler operations it is more expedient to just write the code by hand. Also, CAM for true 4th axis machining is REALLY expensive.

When I design a part, I am doing just that. As a hobbyist, I don't have a CAD library to pull drawings from. So, everything is from scratch. I'm sure that CAD guys who do it day in and day out might be just as fast or faster drawing the part, importing it into CAM then exporting it to Gcode than I am at hand coding. But, I'm only an occasional CAD/CAM user for now- so often it's just simpler to do it the old fashioned way. Believe it or not, I have an actual paper notebook that I sketch everything in with dimensions etc... My lathe is a manual machine, so no code for that :) The mill is CNC, so I make code for it one way or another. Whichever way seems like it will take the least time.

Here is an example of a guy who did his own Gcode in a fairly elegant way to groove a pulley. I doubt if it took him 5 minutes to write it.

Watched the video just now. I think if you move circles out they will intersect the curves instead. Looks pretty dang cool and fun to watch.

What is it like to make smooth curves?

Fritz, I can do curves as well. I hope to do a few experimental pieces with them in the near future. I think intersecting circles would look neat.

This isn’t a flashlight, but it is a shape that could be made on an NC lathe.
Or with radius control also:

Flashlights aren’t that advanced, but I hope it won’t take thousands of years to catch up.
These are the best shapes I have found so far:

The thing is, when you back up to where the design has to start then you end up with the same or similar to what’s out there. These are hand held lights, design from the hand forward and you get

Sure, it can be embellished, but I went by how a light laid in my hand, the reverse flare at the heel of the hand to aid grip, the taper inside to accommodate the flesh of the hand, the “tube” at the base of the pill for finger/thumb grip, essentially all necessary for proper fitment in the hand. Knurl strategic areas and there you have it.

Flower looking lights have been done for Mag D cells, to me the aesthetics just looked too artsy fartsy and so I never bought in. Function first, form as a result of function, then the guts to make it useful. Someone can probably make something that hasn’t been done before, sure, but seriously, a LOT has been done! Make the head and reflector a single piece, the head is a parabola, form following function. Thread a light engine in from behind, the focus would be adjustable, power it up and there you go.

Very nice!
The two lights I posted pictures of above are very functional, as is most pottery. The head of the 5039 is fatter than it needs to be in stock form, but the extra metal helps cool it when I increase the power.