Concerning the Convoy S2+ SST20 4000K 95CRI we discussed several posts ago.
I got notification that it āShippedā today. Hopefully it really didā¦ā¦ ā¦
In the free software world, itās pretty normal for things to be created based on what the community wants. Frequently, the community just makes its own stuff, scratching its own itches. BLF has a similar āmake what you wantā culture, but it involves physical goods instead of just data, and physical goods canāt yet be copied over the internet. So several of us have been trying to get manufacturers to participate in this process, in order to make cool things available to anyone who wants them.
Cory Doctorow based some of his main writing themes on free software culture. Itās a major plot element in his āLittle Brotherā book, and is applied to a hardware context in āMakersā, and is applied to artistic works in āPirate Cinemaā. Heās pretty into this stuff.
He writes about it as sci-fi, usually, butā¦ itās not fiction. It has been an ongoing and growing movement since the mid-1900s, and is now the code foundation for the majority of the internet.
Realize Iām a few weeks behind here, but this caught my eye:
Having done mechanical design in a past career where we outsourced most production and and concept-design-production workflows with relatively simple CNC (laser cutters) as a hobbyist I feel like point 3 is a bit simplified.
Thereās appreciable effort that goes into the design itself, for sure. Someone that produces their own finished product will have some experience with whatās feasible to produce and bake that into the design document.
But having a good design - a production-ready design - isnāt the same thing as having production plans.
Whenever I design a tabbed box for laser cutting, translating the design file into a work file takes appreciable time - layout for material efficiency and optimizing geometry for cutting efficiency being the key tasks.
When dealing with machining metal parts the design to work document process may well be more involved. Tool choices, speed/feed considerations far more involved than laser cutter power/feed, material selection, jigs/workholding, production sequencing (since intermediate parts may be needed between workstations/schedules), QC processes, production scheduling concerns, negotiating slight dimensional changes that could make/break the concept for your capabilities and/or the clientās needs, etc.
Suspect that the prospect for future sales of a design thatās expected to be successful beyond the initial discounted batches - and presumably without encumbrances - is the true carrot.
With the discussion of potentially switching to the 95 CRI SST-20, I just want to double down on the LH351D.
The SST-20 is slightly higher CRI, but at 700 lm the LH351D is around 45% more efficient at the same colour temperature from what Iām seeing, and at 300 lm the LH351D is around 25% more efficient at the same colour temperature.
Thatās a big difference in battery drain and heat for the same output.
edit: this is based on maukkaās tests of the SST-20 and the LH351C (my understanding is that the LH351D is slightly more efficient than the LH351D at 1000 lm, although you lose a bit of efficiency by going from 5000k to 4000k)
edit 2: actually, thatās got me thinking, after reading through the pre-production review fully, do we possibly want to look at the LH351C? If weāre expecting the light to be 900 lumens or lower (300 lumens per emitter) most of the time, then I think weāll be sitting right in the LH351Cās target current. Yeah, weād lose a bit of maximum lumens in turbo (all 20 seconds of it), but the better efficiency in regular use would help the light work better in regular use, and we might be able to source them more quickly.
^ with the SST-20 you loose efficiency/output but you gain a brighter hotspot (and a somewhat higher CRI). And a bit of the 25% efficiency loss is gained back because the smaller die SST-20 gives less optical losses in the TIR.