Hi Hans,
Thanks for your query on E2E forum.
An ideal transfer function for a ADC has a transition point at LSB/2 voltage, the steps are designed to have transitions such that the midpoint of each step corresponds to the point on this ideal line. Here is an example:
Assume an ideal 3 bit ADC with bipolar input range = -4V to 4V:
FSR = 4-(-4) = 8V
ADC codes = 2^3 = 8
LSB size = 1V
LSB/2 = 0.5V
If we plot the transfer curve for the above ADC, the transition points will be as per table below.
As you can notice, at NFS the region is (NFS+0.5LSB) and at PFS this is (PFS-1.5LSB)
|
Code |
Transition Point (V) |
Voltage Range (V) |
|
-4 |
-4 |
-4V to -3.5V |
|
-3 |
-3.5 |
-3.5V to -2.5V |
|
-2 |
-2.5 |
-2.5V to -1.5V |
|
-1 |
-1.5 |
-1.5V to -0.5V |
|
0 |
-0.5 |
-0.5V to +0.5V |
|
1 |
0.5 |
+0.5V to +1.5V |
|
2 |
1.5 |
+1.5V to +2.5V |
|
3 |
2.5 |
+2.5V to +4V |
Another example for unipolar input ADC shown as below, the first transition happens at 1/2 LSB, and the last transition happens at Vref-1.5LSB (Full scale range is Vref).
Hopefully this is helpful for you to understand, thanks.
Best regards
Dale
