
PRODUCT SPECIFICATION
TMC1173A/TMC1273
6
P
Notes:
1. TMC1273 Only.
2. TMC1173A Only.
Power
V
DDA
14
2
, 15
2
,
18
11, 13
17
2
, 18
2
,
21
13, 16
+3.3V
Analog Supply Voltage.
These should originate from a common
+3.3V source and be decoupled to A
GND
.
Digital Supply Voltage.
+3.3 Volt power inputs. These should
originate from a common +3.3V power source and be decoupled
to A
GND
.
Analog Ground.
Connect to the system analog ground plane.
Digital Ground.
Connect to the system analog ground plane.
V
DDD
+3.3V
A
GND
D
GND
No Connect
N/C
20, 21
2, 24
24, 25
3, 28
0.0V
0.0V
1, 8, 15,
22
open
Not Connected.
Pin Description
(continued)
Pin Name
Pin Number
N2, M7
Pin Type
Pin Function Description
R3
Specification Notes
Bandwidth
The specification for bandwidth of an A/D converter is some-
what different from the normal frequency-response specifi-
cation used in amplifiers and filters. An understanding of the
differences will help in selecting converters properly for par-
ticular applications.
A/D conversion comprises two distinct processes: sampling
and quantizing. Sampling is “grabbing” a snapshot of the
input signal and holding it steady for quantizing. The quan-
tizing process is approximating the analog input, which may
be any value within the conversion range, with its nearest
numerical value. While sampling is a high-frequency pro-
cess, quantizing operates on a dc signal, held steady by the
track/hold circuit. Therefore, the sampling process is what
relates to the dynamic characteristics of the converter.
Sampling involves an aperture time, the time during which
the track/hold is trying to capture the input signal and settle
on a dc value to hold. It is analogous to the shutter speed of a
camera: the shorter the aperture (or faster the shutter) the less
the signal will be blurred, and the less uncertainty there will
be in the quantized value.
For example, a 10 MHz sinewave with a 1V peak amplitude
(2Vp-p) has a maximum slew rate of 2
p
fA at zero crossing,
or 62.8V/
m
s. With an 8-bit A/D converter, q (the quantization
step size) = 2V/255 = 7.8mV. The input signal will slew one
LSB in 124ps. To limit the error (and noise) contribution due
to aperture effects to 1/2LSB, the aperture must be shorter
than 62ps.
This is the primary reason that the signal to noise ratio drops
off as full scale frequency increases. Note, also, that the slew
rate is directly proportional to signal amplitude, A. A/Ds will
handle lower-amplitude signals of higher bandwidth.
All this is of particular interest in applications such as digi-
tizing analog VGA RGB signals, or the output of a CCD
imaging chip. These data are effectively pre-sampled: there
is a period of rapid slewing from one pixel value to another,
followed by a relatively stable dc level before the signal
slews to the next pixel value. The goal is, of course, to sam-
ple on these pixel values, not on the slewing between pixels.
During the aperture time, the A/D sees essentially a dc sig-
nal, and classic bandwidth considerations are not important.
As long as the input circuit can slew and settle to the new
value in the prescribed period, an accurate conversion will be
made.
The TMC1173A/1273 is capable of slewing a full 2V and
settling between samples taken as little as 25ns apart, mak-
ing it ideal for digitizing analog VGA and CCD outputs.