And some additional prototype work;
For the "Industrial Design" I am using the short fuse (7/8") while this standard size fuse works well for this implementation.
This is prototype #3. I am awaiting 2 new Tenergy 4200mah sub-C cells that I ordered direct from Tenergy and that is why I hadn't wired this up yet. Proto #2 is performing its duties so I won't disturb that one either.
I'm curious to know how many of you are both LB enthusiasts -and- have the wherewithal to follow a build such as this. Essentially I am throwing functional ideas out to the DIY crowd for the time being. Please do remember that you are working with a lot of energy in these cells. And I welcome questions and I am perfectly willing to share data.
As to the evolution implemented in proto #3, we have a good solid barrel contact although it is brass (polished stainless would be cool here). It is not soldered and it is not a continuous tube. The cell diameter is .56". The brass tube is a 1/2" OD standard hobby tube (.014" wall). I slit the tube. The gap is on the farside and is probably .18" wide when installed. The tube is exactly 1" long. All the edges are cleaned up and treated with chambers and such to remove snags.
The fuse I will be using is an AGC 30 amp. The one I implemented has two 22 awg stranded Teflon wires attached with solder. Bad idea, I know, because you can blow the fuse soldering it. The fuse's end caps also don't take solder readily (bright nickel) but it does get there eventually with enough flux and abrasion. My fuse got a bit cloudy inside but I did test it against the LB load with the other cell and it still works fine. I do think the fuse should remain unsoldered. This can be done with a small copper disk soldered to the wires and heaved into the back of the fuse's chamber where it cannot be pulled out of the wire-hole. Since we are pushing on the end of the housing, the disc also makes sure it always makes good contact with the fuse. In this case, the fuse would be trapped between the 2-wire contact and the rail in the LB. Shrink tubing is added to go all the way to the fuse/contact and is to be buried in the braid-wire as it leaves the dedicated wire hole in the body of the wand.
The braid measures .18" wide and .025" thick. This matters as between the brass and the housing is the spring tension of the brass tube, which is much stronger than one would first expect, to clamp onto the braid making many points of contact. The tube is very workable once installed. For preparation, the braid gets the end folded over. That end is placed and trapped in the pocket in the leading edge of the housing so it can't just be pulled out from under the tube. You want to start the two wires into the braid ASAP so as you dig for a location for a hole to bring those wires inside the braid, be sure to keep in mind all that you are aligning here. You want and exit hole for the wires about 1-3/4" when the braid is pulled 'long' on the opposite side from the entry access hole. This is where the positive cell's tab will be. I'll walks through the shrink wrap insulation of attaching the cell when they get here. Bottom line; make sure you keep insulation in mind and where to add shrink tubing.
The 3D print is PLA this time. The fuse is a snug fit and I modeled it based on my 3D printer's particular idiosyncrasies. A very quick deburr on the ID and the fuse is a snug fit. I've included an 1/8" hole in the thumb-end so you can push the fuse out when needed. The hole for the fuse is slightly barrel shaped so only a small move on the fuse will make it easy to remove if it is not wired in place. I did not bother with a sight hole for the fuse. Just wasn't worth the extra effort yet.
I am still enjoying the ability to charge the cell on the Opus charger. But this also means the unit is not pocketable. Both ends of the cell are exposed. This is easy enough to fix with end-caps and shrink tubing but then we're back to how to charge the unit. Maybe a cell-sock? Obviously a little more effort and we can put a contact on the housing for the stock charger. I'm not there yet.
Although I am printing these using a 3D printer, I am very much keeping in mind fabrication of this type of device from wood. Drill through for a 'push-hole'; drill to depth the fuse hole; turn the OD and contact OD; switch or move to the mill for the tough, pocket, and cable hole. Break sharp edges.
Assembly of this type of unit is a little involved I will admit. However, I can see the process of building these would be quite simple as long as all the parts met spec. Proper fit of the fuses and careful wire management could really make this solution a stunning offering. And if you have much less patience than I do, which is hard to imagine, then I support anyone taking on this type of LB portable power solution as a product. Again, I am not dying to get into yet another production venture from home. This type of development work is fun for me. The rest just boils down to actual work.