, So I don't know if they had that problem when shipped or not. I did have a lot fail on the last 2017 batch, but I had a lot of issues with solder paste on the last run of purple boards, so put it down to that. Those failed boards just went in the bin anyway. The blue ones are all fully checked now that the time advances, There was like 80 of them made up and not one failed in any way. So I don't know why one batch would work perfectly, and another lot fail. Again, I put it down to the paste, but its only a guess. It can't be ruled out that something on your machines has killed them either.
I don't think there was many that batch, looking back there was 25 built, but I remember a lot failed, more failed than passed for sure. So probably about 10 ended in stock, you guys probably went through 5 of them.. But of course some problems were down to no capacitor on the falcon motherboard causing some faults as well.
Looking back again, it was likely Aug 24, 2017 batch. The next purple boards were Jan 30, 2018, I think those were the "in between" batches where I started fitting the capacitor (which your boards didn't have). Some later purple boards were stamped 2018. It doesn't mean of course that "all" boards during that batch had issues as the "Cause" of the failure isn't known.
Perhaps the tool by Uwe Seimet is good for pre-programmed NVRAM but not "fresh" ones where I guess all contents are 0xFF? If you look at the datasheet there is talk about bits needing to be set to make the oscillator run etc.
