MAX2038
Ultrasound VGA Integrated
with CW Octal Mixer
______________________________________________________________________________________ 19
Applications Information
Mode Select Response Time
The mode select response time is the time that the
device takes to switch between CW and VGA modes.
One possible approach to interfacing the CW outputs to
an instrumentation amplifier used to drive an ADC is
shown in Figure 2. In this implementation, there are four
large-value (in the range of 470nF to 1µF) capacitors
between each of the CW_IOUT+, CW_IOUT-,
CW_QOUT+, CW_QOUT- outputs and the circuitry they
are driving. The output of the CW mixer usually drives
the input of an instrumentation amplifier made up of op
amps whose input impedance is set by common-mode
setting resistors.
There are clearly both a highpass corner and a lowpass
corner present in this output network. The lowpass cor-
ner is set primarily by the 115Ω mixer pullup resistors,
the series 50Ω resistors, and the shunt 0.022µF capaci-
tor. This lowpass corner is used to filter a combination
of LO leakage and upper sideband. The highpass cor-
ner, however, is of a larger concern due to the fact that
it is dominated by the combination of a 1µF DC-block-
ing capacitor and the pair of shunt 31.6kΩ resistors.
If drawn, the simplified dominant highpass network
would look like Figure 3.
The highpass pole in this case is at f
P
= 1/(2 x pi x RC)
~ 5Hz. Note that this low highpass corner frequency is
required in order to filter the downconverted clutter tone,
which appears at DC, without interfering with CWD imag-
ing at frequencies as low as 400Hz. For example, if one
wanted to use CWD down to 400Hz, then a good choice
for the highpass pole would be at least a decade below
this (< 40Hz) as not to incur rolloff due to pole. Remember,
if the highpass pole is put at 400Hz, the response is 3dB
down at that corner frequency. The placement of the high-
pass pole at 5Hz in the above example is between the DC
and 40Hz limitations just discussed.
The bottom line is that any reasonably sized DC block
between the output of the mixer and the instrumentation
amplifier will pose a significant time constant that slows
the mode select switching speed.
An alternative solution to the approach in Figure 2,
which enables faster mode select response time, is
shown in Figure 4.
In Figure 4, the outputs of the CWD mixers are DC-
coupled into the inputs of the instrumentation ampli-
fiers. Therefore, the op amps must be able to accom-
modate the full compliance range of the mixer outputs,
which is a maximum of 11V when the mixers are dis-
abled, down to the 5V supply of the MAX2038 when the
mixers are enabled. The op amps can be powered from
11V for the high rail and 5V for the low rail, requiring a
6V op amp.
115Ω115Ω
1μF
1μF
0.022μF
31.6kΩ
31.6kΩ
50Ω
CW_IOUT+
CW_IOUT-
Figure 2. Typical Example of a CW Mixer’s Output Circuit
1μF
31.6kΩ
Figure 3. Simplified Circuit of Highpass Pole
+11V
+5V
Figure 4. Improved Mode Select Response Time Achieved with
DC-Coupled Input to Instrumentation Amplifier
MAX2038
Ultrasound VGA Integrated
with CW Octal Mixer
20 ______________________________________________________________________________________
Serial Interface
The serial interface of the MAX2038 programs the LO for
16, 8, or 4 quadrature phases using a serial shift register
implementation. Data is shifted into the device on DIN.
The serial shift register clock is applied to the CLK input.
The serial shift register has 5 bits per channel. The first 4
bits are for phase programming, and the fifth bit enables
or disables each channel of the mixer array.
Each mixer can be programmed to 1 of 16 phases;
therefore, 4 bits are required for each channel for pro-
gramming. The master high-frequency mixer clock is
applied to differential inputs LO_LVDS+ and LO_LVDS-
(for modes 1 and 2) and LO_ (for modes 3 and 4). The
LOAD input is provided to allow the user to load the
phase counters with the programming values to gener-
ate the correct LO phases. The input signals for mixing
are applied to the eight differential inputs, CWIN_+ and
CWIN_-. The summed I/Q baseband differential outputs
are provided on CW_IOUT+/- and CW_QOUT+/-.
CW_M1 and CW_M2 are used to select one of the four
possible modes of operation. See Table 1.
The serial interface is designed to allow multiple
devices to be easily daisy chained in order to minimize
program interface wiring. DOUT is available for this
daisy-chain function.
Programming the Beamformer
During normal CWD operation, the mixer clock at LO_ or
LO_LVDS+/- is on and the programming signals on DIN,
CLK, and LOAD are off. (LOAD = high, CLK = low, and
DIN = don’t care, but fixed to a high or low). To start the
programming sequence, turn off the mixer clock. Data is
shifted into the shift register at a recommended 10MHz
programming rate or 100ns minimum data clock
period/time. See Figure 5 for timing details.
After the shift registers are programmed, pull the LOAD
bus to logic-low and then back to logic-high to load the
internal counters into I/Q phase divider/selectors with
the proper values. LOAD must remain low for a mini-
mum time of t
CLH
. The user turns on the mixer clock to
start beamforming. The clock must turn on such that it
starts at the beginning of a mixer clock cycle.
DIN
CLK
LOAD
t
DCLKPWH
t
DCLKPWL
t
LD
t
CLH
t
DCLK
t
DSU
t
HLD
t
LDMIXCLK
MIXER
CLOCK ON
MIXER
CLOCK ON
MIXER
CLOCK OFF
MIXER
CLOCK ON
MIXER
CLOCK OFF
MIXER
CLOCK OFF
MIXER
CLOCK ON
Figure 5. Shift Register Timing Diagram
MAX2038
Ultrasound VGA Integrated
with CW Octal Mixer
______________________________________________________________________________________ 21
CW Mixer Output Summation
The maximum differential current output is typically
3mA
P-P
and the mixer output compliance voltage
ranges from 4.75V to 12V per mixer channel. The mixer
common-mode current in each of the differential mixer
outputs is typically 3.25mA. The total summed current
would equal N x 3.25mA in each of the 115Ω load resis-
tors (where N = number of channels). In this case, the
quiescent output voltage at +V
SUM
and -V
SUM
outputs
would be 11V - (N x 3.25mA x 115) = 11V - (8 x 3.25mA
x 115) = 8.05V. The voltage swing at each output, with
one channel driven at max output current (differential
3mA
P-P
) while the other channels are not driven, would
be 1.5mA
P-P
x 115Ω or 174mV
P-P
and the differential
voltage would be 348mV
P-P
. The voltage compliance
range is defined as the valid range for +V
SUM
and -
V
SUM
in this example.
External Compensation
External compensation is required for bypassing inter-
nal biasing circuitry. Connect as close as possible a
4.7µF capacitor from EXT_C1, EXT_C2, and EXT_C3
(pins 13, 14, 15) to ground.
External Bias Resistor
An external resistor at EXT_RES is required to set the
bias for the internal biasing circuitry. Connect, as close
as possible, a 7.5kΩ (0.1%) resistor from EXT_RES (pin
38) to ground.
Analog Input and Output Coupling
In typical applications, the MAX2038 is being driven from
a low-noise amplifier (such as the MAX2034) and the
VGA is typically driving a discrete differential anti-alias fil-
ter into an ADC (such as the MAX1436 octal ADC). The
differential input impedance of the MAX2038 is typically
240Ω. The differential outputs of the VGA are capable of
driving a differential load capacitance to GND at each of
the VGA differential outputs of 60pF, and differential
capacitance across the VGA outputs is 10pF, R
L
=
1kΩ. The differential outputs have a common-mode
bias of approximately 3.75V. AC-couple these differen-
tial outputs if the next stage has a different common-
mode input range.
Ultrasound-Specific IMD3 Specification
Unlike typical communications specifications, the two
input tones are not equal in magnitude for the ultra-
sound-specific IMD3 two-tone specification. In this
measurement, f
1
represents reflections from tissue and
f
2
represents reflections from blood. The latter reflec-
tions are typically 25dB lower in magnitude, and hence
the measurement is defined with one input tone 25dB
lower than the other. The IMD3 product of interest (f
1
-
(f
2
- f
1
)) presents itself as an undesired Doppler error
signal in ultrasound applications. See Figure 6.
Board Layout
The pin configuration of the MAX2038 is optimized to
facilitate a very compact physical layout of the device
and its associated discrete components. A typical
application for this device might incorporate several
devices in close proximity to handle multiple channels
of signal processing.
The exposed pad (EP) of the MAX2038’s TQFP-EP
package provides a low thermal-resistance path to the
die. It is important that the PCB on which the MAX2038
is mounted be designed to conduct heat from the EP.
In addition, provide the EP with a low-inductance path
to electrical ground. The EP MUST be soldered to a
ground plane on the PCB, either directly or through an
array of plated via holes.
-25dB
ULTRASOUND
IMD3
f
1
- (f
2
- f
1
)f
2
+ (f
2
- f
1
)f
1
f
2
Figure 6. Ultrasound IMD3 Measurement Technique

MAX2038CCQ+

Mfr. #:
Manufacturer:
Maxim Integrated
Description:
Instrumentation Amplifiers Ultrasound Variable Gain Amplifie
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union

Products related to this Datasheet