ADC question

CodingChimp

                    fltl    pin
                    wrpin   ##( P_ADC | P_ADC_GIO ), pin
                    wxpin   #%00_1101,              pin
                    wypin   #0,                     pin
                    dirh    pin
                    waitx   ##( 8192 << 2 )
                    rdpin   callo_tmp,              pin

                    fltl    pin
                    wrpin   ##( P_ADC | P_ADC_VIO ), pin
                    wxpin   #%00_1101,              pin
                    wypin   #0,                     pin
                    dirh    pin
                    waitx   ##( 8192 << 2 )
                    rdpin   calhi_tmp,              pin

I used this bit of code to read the 0v and 3.3v on one of my analog inputs, but the level measured for 0v was around 2-3K counts which I found odd. And the 3.3V level was around 13K counts. Why is the spread between these so small when using 14 bit SINC2 sampling mode?

Circuitsoft

14 bits, max count is 2^14=16384. The scaling resistors allow for reading a little beyond the rails, so the 0v and 3.3v inputs are provided to calibrate the input levels. Being a sigma/delta type, the ADC is not a "multiplying" ADC, so it doesn't scale relative to supply voltage.

evanh

Here's a write-up I did - https://forums.parallax.com/discussion/comment/1484327/#Comment_1484327
It's not exactly how the ADC works but is a rough analogue. Basically there is a virtual ground at VIO/2 with scaling resistors. At x1, the scaling is roughly a 5.0 Volt span. There is measurable current pulling to VIO/2, and is higher at higher gains. The exact circuit is here - https://forums.parallax.com/discussion/comment/1510744/#Comment_1510744

This block diagram might also be of interest - https://forums.parallax.com/discussion/171420/smartpin-diagram-now-with-p-p-bit-mode-table

PS: And this too - https://forums.parallax.com/discussion/174441/measuring-floating-pin-voltage-adc-from-just-the-built-in-comparator/p1