vbat/4 looks stable, is functional
External input works
PTAT does not look so good, seems stuck at 255
some issues with determining absolute voltage
There may be some offset that needs to be characterized per-chip if we need to know absolute voltage. For situations that just need deltaV, we don't need that calibration.
offset is different by chip in part because ADC LDO varies per chip
@Titan Yuan's setup had a strange short between the LDO and external input
Typically 5ish LSB variance around samples. Not good but acceptable for the intended applications