Datasheet

AD9278 Data Sheet
Rev. A | Page 34 of 44
I/Q Demodulator and Phase Shifter
The I/Q demodulators consist of double-balanced passive mixers.
The RF input signals are converted into currents by transconduc-
tance stages that have a maximum differential input signal
capability matching the LNA output full scale. These currents
are then presented to the mixers, which convert them to base-
band (RF − LO) and twice RF (RF + LO). The signals are phase
shifted according to the codes programmed into the SPI latch
(see Table 16). The phase shift function is an integral part of the
overall circuit. The phase shift listed in Column 1 of Table 16 is
defined as being between the baseband I or Q channel outputs.
As an example, for a common signal applied to a pair of RF inputs
to an AD9278, the baseband outputs are in phase for matching
phase codes. However, if the phase code for Channel 1 is 0000
and that of Channel 2 is 0001, then Channel 2 leads Channel 1
by 22.5°.
Table 16. Phase Select Code for Channel-to-Channel Phase Shift
Φ Shift
I/Q Demodulator Phase
(SPI Register 0x2D[3:0])
0°
0000
22.5° 0001
45° 0010
67.5° 0011
90° 0100
112.5° 0101
135° 0110
157.5° 0111
180° 1000
202.5° 1001
225° 1010
247.5° 1011
270° 1100
292.5° 1101
315° 1110
337.5° 1111
Dynamic Range and Noise
Figure 63 is an interconnection block diagram of all eight
channels of the AD9278. More channels are easily added to that
summation (up to 32 when using an ADA4841 as the
summation amplifier) by wire-OR connecting the outputs as
shown. In beamforming applications, the I and Q outputs of a
number of receiver channels are summed. The dynamic range of
the system increases by the factor 10log
10
(N), where N is the
number of channels (assuming random uncorrelated noise).
The noise in the 8-channel example of Figure 63 is increased by
9 dB while the signal quadruples (18 dB), yielding an aggregate
SNR improvement of (18 − 9) = 9 dB.
The output-referred noise of the CW signal path depends on
the LNA gain and the selection of the external summing
amplifier and the value of R
FILT
. To determine the output
referred noise, it is important to know the active low-pass filter
(LPF) values, R
FILT
, and C
FILT
, shown in Figure 63. Typical filter
values for a single channel are 2 kΩ for R
FILT
and 0.8 nF for C
FILT
;
these values implement a 100 kHz single-pole LPF. In the case
where eight channels are summed, R
FILT
and C
FILT
are 250 Ω and
6.4 nF.
If the RF and LO are offset by 10 kHz, the demodulated signal is
10 kHz and is passed by the LPF. The single-channel mixing gain
from the RF input to the ADA4841 output (for example, I1´,
Q1´) is approximately the LNA gain for R
FILT
and C
FILT
of 2 kΩ
and 0.8 nF.
This gain can be increased by increasing the filter resistor while
maintaining the corner frequency. The factor limiting the
magnitude of the gain is the output swing and drive capability
of the op amp selected for the I-to-V converter, in this example,
the ADA4841. Because any amplifier has limited drive capability,
there is a finite number of channels that can be summed. The
channel-summing limit relates directly to the current drive
capability of the amplifier used to implement the active low-
pass filter and current-to-voltage converter. The maximum
sum, when the ADA4841 is used, is 32 channels of the AD9278;
that is, four AD9278s (4 × 8 = 32 channels) can be summed in
one ADA4841.