
AD6652
Preliminary Technical Data
Rev. PrC | Page 40 of 76
.
2,
rogrammed in Output
Port Control Registers 0x10:1–0 and 0x18:1–0. These averaged
samples are then decimated with decimation ratios programma-
ble from 1 to 4096. This 12-bit decimation ratio is defined in
ensate for the gain associated
with t ese operations, attenuation scaling is provided before the
CIC filter.
This scaling operation accounts for the division associated with
h in
ted as a bit shift
ttenuation scaling, SCIC, is programmable from 0 to
14 using 4 bits of 0x10 and 0x18 of the output port control
registers, and is given by
C
N
M
ceil
×
=
tion
2
(3000)). Because attenuation is
B
nd, therefore, prevents overflows in the AGC
loop. But it is also evident that the CIC scaling is inducing a
d
converted to rms signal samples by
applying a square root. This square root is implemented using a
he
e
lso compensate for error, if any, due
ed previously. Therefore, the
scaling constant over a frame of data (or a stream of symbols)
Because the number of average samples must be an integer
multiple of the decimation value, only the multiple number 1,
3, or 4 is programmed. This number is p
Registers 0x11 and 0x19.
The average and decimate operations are linked together and
implemented using a first-order CIC filter and FIFO registers.
The gain and bit growth associated with CIC filters depends on
the decimation ratio. To comp
h
the averaging operation as well as the traditional bit growt
CIC filters. Because this scaling is implemen
operation, only coarse scaling is possible. Fine scaling is imple-
mented as an offset in the request level, explained later in this
section. The a
)]
(
[log
2
avg
CIC
CI
S
where:
MCIC is the decimation ratio (1 to 4096).
Navg is the number of averaged samples programmed as a
multiple of decimation ratio (1, 2, 3, or 4).
Ceil is MathCad-speak for rounding up to the next whole
number.
For example, if a decimation ratio MCIC is 1000 and Navg is
selected to be 3 (decimation of 1000 and averaging of 3000
samples), then the actual gain due to averaging and decima
is 3000 or 69.54 dB ( = log
implemented as a bit shift operation, only multiples of 6.02 d
attenuations are possible. SCIC, in this case, is 12 corresponding
to 72.24 dB. This way, SCIC scaling always attenuates more than is
sufficient to compensate for the gain changes in average and
decimate sections a
gain error (difference between gain due to CIC and attenuation
provided) of up to 6.02 dB. This error should be compensate
for in the request signal level, as explained below.
Logarithm to the Base 2 is applied to the output from the
average and decimate section. These decimated power samples
(in logarithmic domain) are
simple shift operation. The rms samples so obtained are sub-
tracted from the request signal level, R, specified in Registers
(0x0B, 0x14), leaving an error term to be processed by the loop
filter, G(z).
Set this programmable request signal level, R, according to t
output signal level desired. The request signal level R is
programmable from 0 to 23.99 dB in steps of 0.094 dB. Th
request signal level should a
to the CIC scaling, as explain
request signal level is offset by the amount of error induced in
CIC, given by
02
.
6
)
(
log
20
10
×
×
=
CIC
avg
CIC
S
N
M
Offset
where the offset is in dB.
Continuing with the previous example, this offset is given by
Offset = 72.24 69.54 = 2.7 dB
So the request signal level is given by
094
.
0
094
.
0
)
(
×
=
Offset
DSL
ceil
R
where:
R is the request signal level.
DSL (desired signal level) is the output signal level that the
desires.
So, in the previous example, if the desired signal level is
13.8 dB, the request signal level, R, is programmed to be
16.54 dB.
The AGC provides a programmable second-order loop filter
The program
user
.
mable parameters, gain K and pole P, completely
The error term after
r
define the loop filter characteristics.
subtracting the request signal level is processed by the loop
filter, G(z). The open loop poles of the second-order loop filte
are 1 and P, respectively. The loop filter parameters, pole P and
gain K, allow adjustment of the filter time constant, which
determines the window for calculating the peak-to-average
ratio.
The open loop transfer function for the filter including the gain
parameter is as follows:
2
1
)
(
=
Kz
z
G
)
1
(
1
+
Pz
z
P
r
If the AGC is properly configured (in terms of offset in request
level), then there are no gains except the filter gain K. Unde
these circumstances, a closed loop expression for the AGC loop
is possible and is given by
2
1
)
1
(
1
)
(
1
)
(
)
(
+
+
=
+
=
Pz
z
P
K
Kz
z
G
z
G
z
G
closed