Status of Sentinel 65X: It more or less works. Some stuff needs testing that there's no code yet to test, but I see no reason to think those things won't work. The design itself is proved sane.
So here's my best idea for a plan to actually sell Sentinel 65X boards to devs who want in early.
I cannot fund a run of 50 on my own, which is required to bring the cost of a unit down to an acceptable level. So I have to take orders and money before I have the product to deliver. That sucks, I hate it, but it is what it is.
However, before that, the SMD prototype now being finished will be fully tested and assured working as designed, or at least have the bugs worked out in a revision to be used for the big batch. Say, for example, I might have picked the wrong part for chip foo, or the wrong component size for part bar. Specific example, the tolerances on the QFP-100 footprint in KiCAD are super tight, so I've already added 0.5mm to each pad to fix that.
So let's assume I have a design that works, production ready. Next, I take the orders and collect the money. Once I have the money, I go into JLCPCB's thing and order all the parts into my private reserve. That will take time to fill, sadly, but there's little to be done about that.
Once all the parts are available, then I can put in the actual PCB+Assembly order.
I tell you, if Olimex wanted to handle this for me, I'd endorse it in a second, I hate this entire process. I am terrified I'm going to screw it up for everyone.
I'm still actively seeking a desktop-style Turbo XT/AT clone case, if anyone has one to sell.
Perfectly fine if the guts are busted, battery bombed, even a bit rusty.
