
Analog to Digital (ATD) Module
MC1321x Reference Manual, Rev. 1.1
20-6
Freescale Semiconductor
large CAS is used and sufficient time to recharge CAS is provided between samples. In order to achieve
accuracy specified under the worst case conditions of maximum
V
SAMP and minimum CAS, RAS must
be less than the maximum value of 10 k
. The maximum value of 10 k for R
AS is to ensure low sampling
error in the worst case condition of maximum
V
SAMP and minimum CAS.
20.3.3
Analog Input Multiplexer
The analog input multiplexer selects one of the eight external analog input channels to generate an analog
sample. The analog input multiplexer includes negative stress protection circuitry which prevents
cross-talk between channels when the applied input potentials are within specification. Only analog input
signals within the potential range of VREFL to VREFH (ATD reference potentials) will result in valid ATD
conversions.
20.3.4
ATD Module Accuracy Definitions
Figure 20-3 illustrates an ideal ATD transfer function. The horizontal axis represents the ATD input
voltage in millivolts. The vertical axis the conversion result code. The ATD is specified with the following
figures of merit:
Number of bits (N) — The number of bits in the digitized output
Resolution (LSB) — The resolution of the ATD is the step size of the ideal transfer function. This
is also referred to as the ideal code width, or the difference between the transition voltages to a
given code and to the next code. This unit, known as 1LSB, is equal to
1LSB = (VREFH – VREFL) / 2
N
Eqn. 20-5
Inherent quantization error (EQ) — This is the error caused by the division of the perfect ideal
straight-line transfer function into the quantized ideal transfer function with 2N steps. This error is
± 1/2 LSB.
Differential non-linearity (DNL) — This is the difference between the current code width and the
ideal code width (1LSB). The current code width is the difference in the transition voltages to the
current code and to the next code. A negative DNL means the transfer function spends less time at
the current code than ideal; a positive DNL, more. The DNL cannot be less than –1.0; a DNL of
greater than 1.0 reduces the effective number of bits by 1.
Integral non-linearity (INL) — This is the difference between the transition voltage to the current
code and the transition to the corresponding code on the adjusted transfer curve. INL is a measure
of how straight the line is (how far it deviates from a straight line). The adjusted ideal transition
voltage is:
Eqn. 20-6
Zero scale error (EZS) — This is the difference between the transition voltage to the first valid code
and the ideal transition to that code. Normally, it is defined as the difference between the actual and
(Current Code - 1/2)
2N
Adjusted Ideal Trans. V =
* ((VREFH + EFS) - (VREFL + EZS))