Datasheet LTC1668 Datasheet (Linear Technology)

Page 1
Final Electrical Specifications
FREQUENCY (1.25MHz/DIV)
0.05
SIGNAL AMPLITUDE (dBm)
–45
–25
–5
1668 G01
–65
–85
–55
–35
–15
–75
–95
–105
6.3
12.55
f
CLOCK
= 25Msps
f
OUT
= 1.007MHz AMPLITUDE = 0dBFS = –8.5dBm SFDR = 86dBc
FEATURES
LTC1668
16-Bit, 50Msps DAC
February 2000
U
DESCRIPTIO
50Msps Update Rate
16-Bit Resolution
High Spectral Purity: 87dB SFDR at 1MHz f
Differential Current Outputs
30ns Settling Time
5pV-s Glitch Impulse
Low Power: 180mW from ±5V Supplies
TTL/CMOS (3.3V or 5V) Inputs
Small Package: 28-Pin SSOP
U
APPLICATIO S
Cellular Base Stations
Multicarrier Base Stations
Wireless Communication
Direct Digital Synthesis (DDS)
xDSL Modems
Arbitrary Waveform Generation
Automated Test Equipment
Instrumentation
The LTC®1668 is a 16-bit, 50Msps differential current output DAC implemented on a high performance BiCMOS process with laser trimmed, thin-film resistors. The com­bination of a novel current-steering architecture and a high performance process produces a DAC with excep­tional AC and DC performance. This is the first 16-bit DAC in the marketplace to exhibit an SFDR (spurious free dynamic range) of 87dB for an output signal frequency of 1MHz.
Operating from ±5V supplies, the LTC1668 can be con­figured to provide full-scale output currents up to 10mA. The differential current outputs of the DAC allow single­ended or true differential operation. The –1V to 1V output compliance of the LTC1668 allows the outputs to be con­nected directly to external resistors to produce a differ­ential output voltage without degrading the converter’s linearity. Alternatively, the outputs can be connected to the summing junction of a high speed operational amplifier, or to a transformer.
The LTC1668 is available in a 28-pin SSOP and is fully specified over the industrial temperature range.
, LTC and LT are registered trademarks of Linear Technology Corporation.
TYPICAL APPLICATIO
REFOUT
R
0.1µF
0.1µF
SET
C1
C2
0.1µF
2k
I
REFIN
COMP1
COMP2
V
SS
–5V
U
16-Bit, 50Msps DAC
5V
0.1µF
V
2.5V
REFERENCE
+
AGND DGND CLK DB15 DB0
0.1µF
DD
HIGH SPEED
CLOCK
INPUT
Information furnished by Linear Technology Corporation is believed to be accurate and reliable. However, no responsibility is assumed for its use. Linear Technology Corporation makes no represen­tation that the interconnection of its circuits as described herein will not infringe on existing patent rights.
16-BIT
DAC
16-BIT DATA
INPUT
LTC1668
I
OUT A
I
OUT B
LADCOM
1668 TA01
52.3
52.3 V
+
OUT
1V
P-P
DIFFERENTIAL
Single Tone SFDR
1
Page 2
LTC1668
WWWU
ABSOLUTE AXI U RATI GS
(Note 1)
Supply Voltage (VDD)................................................ 6V
Negative Supply Voltage (VSS) ............................... – 6V
Total Supply Voltage (VDD to VSS) .......................... 12V
Digital Input Voltage ....................–0.3V to (VDD + 0.3V)
Analog Output Voltage
(I
and I
OUT A
Power Dissipation............................................. 500mW
Operating Temperature Range
LTC1668C .............................................. 0°C to 70°C
LTC1668I........................................... –40°C to 85°C
Storage Temperature Range ................ –65°C to 150°C
Lead Temperature (Soldering, 10 sec).................. 300°C
) ........ (VSS – 0.3V) to (VDD + 0.3V)
OUT B
UU
W
PACKAGE/ORDER I FOR ATIO
TOP VIEW
DB13 DB12 DB11 DB10
DB9 DB8 DB7 DB6 DB5 DB4 DB3 DB2 DB1
DB0 (LSB)
1 2 3 4 5 6 7 8
9 10 11 12 13 14
G PACKAGE
28-LEAD PLASTIC SSOP
T
= 110°C, θJA = 100°C/W
JMAX
DB14
28
DB15 (MSB)
27
CLK
26
V
25
DGND
24
V
23
COMP2
22
COMP1
21
I
20
I
19
LADCOM
18
AGND
17
I
16
REFOUT
15
DD
SS
OUT A
OUT B
REFIN
Consult factory for Military grade parts.
ORDER PART
NUMBER
LTC1668CG LTC1668IG
ELECTRICAL CHARACTERISTICS
temperature range, otherwise specifications are at TA = 25°C. VDD = 5V, VSS = –5V, LADCOM = AGND = DGND = 0V, I
The denotes specifications which apply over the full operating
= 10mA.
OUTFS
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS DC Accuracy (Measured at I
, Driving a Virtual Ground)
OUTA
Resolution 16 Bits Monotonicity 14 Bits
INL Integral Nonlinearity ±8 LSB DNL Differential Nonlinearity ±1 ±4 LSB
Offset Error 0.1 ±0.2 % FSR Offset Error Drift 5 ppm/°C
GE Gain Error Internal Reference, R
External Reference, V
= 2k 2 % FSR
IREFIN
= 2.5V, R
REF
= 2k 1 % FSR
IREFIN
Gain Error Drift Internal Reference 75 ppm/°C
External Reference 50 ppm/°C
PSRR Power Supply Rejection Ratio VDD = 5V ±5% ±0.1 % FSR/V
= –5V ±5% ±0.1 % FSR/V
V
SS
Analog Output
I
OUTFS
Full-Scale Output Current 110mA Output Compliance Range IFS = 10mA –1 1 V Output Resistance; R
IOUTA
, R
IOUTB
I
to LADCOM 0.7 1.1 1.5 k
OUTA, B
Output Capacitance 5pF
Reference Output
Reference Voltage REFOUT Tied to I
Through 2k 2.475 2.5 2.525 V
REFIN
Reference Output Drift 25 ppm/°C Reference Output Load Regulation I
= 0mA to 5mA 6 mV/mA
LOAD
2
Page 3
LTC1668
ELECTRICAL CHARACTERISTICS
temperature range, otherwise specifications are at TA = 25°C. VDD = 5V, VSS = –5V, LADCOM = AGND = DGND = 0V, I
The denotes specifications which apply over the full operating
= 10mA.
OUTFS
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS Reference Input
Reference Small-Signal Bandwidth IFS = 10mA, C
= 0.1µF20kHz
COMP1
Power Supply
V
DD
V
SS
I
DD
I
SS
P
DIS
Positive Supply Voltage 4.75 5 5.25 V Negative Supply Voltage –4.75 –5 –5.25 V Positive Supply Current IFS = 10mA, f Negative Supply Current IFS = 10mA, f Power Dissipation IFS = 10mA, f
IFS = 1mA, f
= 25Msps, f
CLK
= 25Msps, f
CLK
= 25Msps, f
CLK
= 25Msps, f
CLK
= 1MHz 35 mA
OUT
= 1MHz 33 40 mA
OUT
= 1MHz 180 mW
OUT
= 1MHz 85 mW
OUT
Dynamic Performance (Differential Transformer Coupled Output, 50 Double Terminated, Unless Otherwise Noted)
f
CLOCK
t
S
t
PD
Maximum Update Rate 50 75 Msps Output Settling Time To 0.1% FSR 30 ns Output Propagation Delay 8ns Glitch Impulse Single Ended 15 pV-s
Differential 5 pV-s
t
r
t
f
i
NO
Output Rise Time 4ns Output Fall Time 4ns Output Noise IFS = 10mA 50 pA/√Hz
IFS = 1mA 30 pA/√Hz
AC Linearity
SFDR Spurious Free Dynamic Range f
= 25Msps, f
CLK
OUT
= 1MHz
to Nyquist 0dB FS Output 78 87 dB
–6dB FS Output 87 dB –12dB FS Output 86 dB –18dB FS Output 80 dB
f
Spurious Free Dynamic Range f Within a Window f
THD Total Harmonic Distortion f
= 50Msps, f
CLK
f
= 50Msps, f
CLK
f
= 50Msps, f
CLK
f
= 50Msps, f
CLK
= 25Msps, f
CLK
= 50Msps, f
CLK
= 25Msps, f
CLK
f
= 50Msps, f
CLK
= 1MHz 84 dB
OUT
= 2.5MHz 80 dB
OUT
= 5MHz 77 dB
OUT
= 20MHz 65 dB
OUT
= 1MHz, 2MHz Span 86 96 dB
OUT
= 5MHz, 4MHz Span 88 dB
OUT
= 1MHz –84 –77 dB
OUT
= 5MHz –76 dB
OUT
Digital Inputs
V
IH
V
IL
I
IN
C
IN
t
DS
t
DH
t
CLKH
t
CLKL
Digital High Input Voltage 2.4 V Digital Low Input Voltage 0.8 V Digital Input Current ±10 µA Digital Input Capacitance 5pF Input Setup Time 8ns Input Hold Time 4ns Clock High Time 5ns Clock Low Time 8ns
Note 1: Absolute Maximum Ratings are those values beyond which the life of the device may be impaired.
3
Page 4
LTC1668
UW
TYPICAL PERFOR A CE CHARACTERISTICS
Single Tone SFDR 2-Tone SFDR
–5 –15 –25 –35 –45 –55 –65 –75
SIGNAL AMPLITUDE (dBm)
–85 –95
–105
0.05 FREQUENCY (1.25MHz/DIV)
f
= 25Msps
CLOCK
f
= 1.007MHz
OUT
AMPLITUDE = 0dBFS = –8.5dBm SFDR = 86dBc
6.3
1668 G01
12.55
Integral Nonlinearity Differential Nonlinearity
5 4 3 2
1
0 –1 –2 –3
INTEGRAL NONLINEARITY (LSB)
–4 –5
16384
32768
DIGITAL INPUT CODE
49152
65535
1668 G03
–10 –20 –30 –40 –50 –60 –70 –80
SIGNAL AMPLITUDE (dBm)
–90 –100 –110
2.0
1.5
1.0
0.5
–0.5
–1.0
DIFFERENTIAL NONLINEARITY (LSB)
–1.5
–2.0
f
= 50Msps
CLOCK
=
f
OUT1
4.028MHz =
f
OUT2
4.419MHz
AMPLITUDE 1, 2 = –6dBFS = –14.5dBm SFDR > 77dBc
3.2
0
0
4.2
FREQUENCY (0.2MHz/DIV)
16384
32768
DIGITAL INPUT CODE
49152
5.2
1668 G02
65535
1668 G04
U
UU
PI FU CTIO S
REFOUT (Pin 15): Internal Reference Voltage Output. Nominal value is 2.5V. Requires a 0.1µF bypass capacitor to AGND.
I
(Pin 16): Reference Input Current. Nominal value is
REFIN
1.25mA for IFS = 10mA. IFS = I
REFIN
• 8.
AGND (Pin 17): Analog Ground. LADCOM (Pin 18): Attenuator Ladder Common. Normally
tied to GND.
I
(Pin 19): Complementary DAC Output Current. Full-
OUT B
scale output current occurs when all data bits are 0s.
I
(Pin 20): DAC Output Current. Full-scale output
OUT A
current occurs when all data bits are 1s.
4
COMP1 (Pin 21): Current Source Control Amplifier Com­pensation. Bypass to VSS with 0.1µF.
COMP2 (Pin 22): Internal Bypass Point. Bypass to V
SS
with 0.1µF. VSS (Pin 23): Negative Supply Voltage. Nominal value is
–5V.
DGND (Pin 24): Digital Ground. VDD (Pin 25): Positive Supply Voltage. Nominal value is 5V. CLK (Pin 26): Clock Input. Data is latched and the output
is updated on positive edge of clock. DB15 to DB0 (Pins 27, 28, 1 to 14): Digital Input Data Bits.
Page 5
BLOCK DIAGRA
LTC1668
0.1µF
V
REF
REFOUT
15
R
SET
2k
IREFIN
16
W
2.5V
REFERENCE
IFS/8
5V
25
ATTENUATOR
LADDER
LSB SWITCHES
0.1µF
SEGMENTED SWITCHES
FOR DB15–DB12
LADCOM
I
OUT A
I
OUT B
18
20
19
52.352.3
+ –
LTC1668
V
OUT
1V
P-P
DIFFERENTIAL
COMP121
22
0.1µF
COMP2
0.1µF
–5V
UWW
TI I G DIAGRA
+
23
V
SS
0.1µF
AGND17DGND
DB0
TO DB15
CLK
I
INT
24
N – 1
CLK DB0DB15 • • •
26 27 14
CLOCK
INPUT
CURRENT SOURCE ARRAY
• • • • • •
INPUT LATCHES
16-BIT
DATA INPUT
N N + 1
t
DS
t
DH
1668 BD
I
OUT A/IOUT B
t
CLKL
N – 1 N
t
CLKH
t
PD
t
ST
0.1%
1668 TD
5
Page 6
LTC1668
WUUU
APPLICATIO S I FOR ATIO
Theory of Operation
The LTC1668 is a high speed current steering 16-Bit DAC made on an advanced BiCMOS process. Precision thin film resistors and well matched bipolar transistors result in excellent DC linearity and stability. A low glitch current switching design gives excellent AC performance at sample rates up to 50Msps. The device is complete with a 2.5V internal bandgap reference and edge triggered latches, and sets a new standard for DAC applications requiring very high dynamic range at output frequencies up to several megahertz.
Referring to the Block Diagram, the DAC contains an array of current sources that are steered to I
OUTA
or I
OUTB
with NMOS differential current switches. The four most signifi­cant bits, DB15 to DB12 are made up of 15 current segments of equal weight. The lower bits, DB11 to DB0 are binary weighted, using a combination of current scaling and a differential resistive attenuator ladder. All bits and segments are precisely matched, both in current weight for DC linearity, and in switch timing for low glitch impulse and low spurious tone AC performance.
Setting the Full-Scale Current, I
The full-scale DAC output current, I
OUTFS
OUTFS
, is nominally 10mA, and can be adjusted down to 1mA. Placing a resistor, R sets I
OUTFS
, between the REFOUT pin, and the I
SET
as follows.
REFIN
pin
The internal reference control loop amplifier maintains a virtual ground at I source, I I
is a scaled replica of the DAC current sources and
INT
I
OUTFS
I
OUTFS
, to sink the exact current flowing into I
INT
= 8 • (I
= 8 • (I
), therefore:
INT
REFIN
For example, if R
2.5V, I
= 2.5/2k = 1.25mA and I
REFIN
by servoing the internal current
REFIN
) = 8 • (V
= 2k and is tied to V
SET
REF/RSET
) (1)
REF
= 8 • (1.25mA)
OUTFS
REFIN
= REFOUT =
.
= 10mA. The reference control loop requires a capacitor on the
COMP1 pin for compensation. For optimal AC perfor­mance, C
should be connected to VSS and be placed
COMP1
very close to the package (less than 0.1").
For fixed reference voltage applications, C
COMP1
should
be 0.1µF or more. The reference control loop small-signal bandwidth is approximately 1/(2π) • C for C
COMP1
= 0.1µF.
COMP1
• 80 or 20kHz
Internal Reference Output—REFOUT
The onboard 2.5V bandgap voltage reference drives the REFOUT pin. It is trimmed and specified to drive a 2k resistor tied from REFOUT to I
1.25mA load (I
= 10mA). REFOUT has nominal
OUTFS
, corresponding to a
REFIN
output impedance of 6, or 0.24% per mA, so it must be buffered to drive any additional external load. A 0.1µF capacitor is required on the REFOUT pin for compensa­tion. Note that this capacitor is required for stability, even if the internal reference is not being used.
DAC Transfer Function
The LTC1668 uses straight binary digital coding. The complementary current outputs, I current from 0 to I I
swings from 0mA when all bits are low (i.e., Code =
OUT A
OUTFS
. For I
OUTFS
OUT A
and I
OUT B
, sink
= 10mA (nominal),
0) to 10mA when all bits are high (i.e., Code = 65535) (deci­mal representation). I I
OUT A
I
OUT A
I
OUT B
and I
= I = I
are given by the following formulas:
OUT B
• (DAC Code/65536) (2)
OUTFS
• (65535-DAC Code)/65536 (3)
OUTFS
is complementary to I
OUT B
OUT A
.
In typical applications, the LTC1668 differential output currents either drive a resistive load directly or drive an equivalent resistive load through a transformer, or as the feedback resistor of an I-to-V converter. The voltage outputs generated by the I
OUT A
and I
output currents
OUT B
are then:
V
OUT A
V
OUT B
= I = I
OUT A
OUT B
• R
• R
LOAD
LOAD
(4) (5)
The differential voltage is:
V
DIFF
= V
= (I
OUT A OUT A
– V
– I
OUT B
OUT B
) • (R
LOAD
(6)
)
6
Page 7
WUUU
APPLICATIO S I FOR ATIO
LTC1668
Substituting the values found earlier for I I
:
OUTFS
V
= {2 • DAC Code – 65535)/65536} • 8 •
DIFF
(R
LOAD/RSET
) • (V
) (7)
REF
OUT A
, I
OUT B
and
From these equations some of the advantages of differen­tial mode operation can be seen. First, any common mode noise or error on I
OUT A
and I
is cancelled. Second, the
OUT B
signal power is twice as large as in the single-ended case. Third, any errors and noise that multiply times I I
, such as reference or I
OUT B
noise, cancel near
OUTFS
OUT A
and
midscale, where AC signal waveforms tend to spend the most time. Fourth, this transfer function is bipolar; e.g. the output swings positive and negative around a zero output at mid-scale input, which is more convenient for AC applications.
Note that the term (R
LOAD/RSET
) appears in both the differential and single-ended transfer functions. This means that the Gain Error of the DAC depends on the ratio of R
to R
LOAD
temperature tracking of R the absolute tempco of R
, and the Gain Error tempco is affected by the
SET
with R
LOAD
is very critical for DC
LOAD
. Note also that
SET
nonlinearity. As the DAC output changes from 0mA to 10mA the R
resistor will heat up slightly, and even a
LOAD
very low tempco can produce enough INL bowing to be significant at the 16-bit level. This effect disappears with medium to high frequency AC signals due to the slow thermal time constant of the load resistor.
Analog Outputs
The LTC1668 has two complementary current outputs, I
and I
OUT A
impedance of I
(see DAC Transfer Function). The output
OUT B
OUT A
and I
OUT B
(R
IOUT A
and R
IOUT B
) is
typically 1.1k to LADCOM. (See the Equivalent Analog Output Circuit, Figure 1.) The LADCOM pin is the com­mon connection for the internal DAC attenuator ladder. It usually is tied to analog ground, but more generally it should connect to the same potential as the lead resistors on I current to VSS of approximately 0.32 • (I current that flows from I R
IOUT A
OUT A
and I
and R
. The LADCOM pin carries a constant
OUT B
OUTFS
IOUT B
OUT A
resistors.
and I
through the
OUT B
), plus any
R
IOUT A
1.1k
LADCOM
I
OUT A
I
OUT B
5pF
V
18
20
52.3
19
52.3
SS
–5V
23
1668 F01
LTC1668
R
IOUT B
1.1k
5pF
Figure 1. Equivalent Analog Output Circuit
The specified output compliance voltage range is ±1V. The DC linearity specifications, INL and DNL, are trimmed and guaranteed on I
into the virtual ground of an
OUT A
I-to-V converter, but are typically very good over the full output compliance range. Above 1V the output current will start to increase as the DAC current steering switch impedance decreases, degrading both DC and AC linear­ity. Below – 1V, the DAC switches will start to approach the transition from saturation to linear region. This will de­grade AC performance first, due to nonlinear capacitance and increased glitch impulse. AC distortion performance is optimal at amplitudes less than ±0.5V I
due to nonlinear capacitance and other large-signal
OUT B
P-P
on I
OUT A
and
effects. At first glance, it may seem counter-intuitive to decrease the signal amplitude when trying to optimize SFDR. However, the error sources that affect AC perfor­mance generally behave as additive currents, so decreas­ing the load impedance to reduce signal voltage amplitude will reduce most spurious signals by the same amount.
The LTC1668 is specified to operate with full-scale output current, I
, from the nominal 10mA down to 1mA.
OUTFS
This can be useful to reduce power dissipation or to adjust full-scale value. However, that the LTC1668 DC and AC accuracy is specified only at I AC accuracy will fall off significantly at lower I At I
= 1mA, INL and DNL typically degrade to the 14-
OUTFS
= 10mA, and DC and
OUTFS
OUTFS
values.
bit to 13-bit level, compared to 16-bit to 15-bit typical accuracy at 10mA I
. Increasing I
OUTFS
from 1mA, the
OUTFS
7
Page 8
LTC1668
WUUU
APPLICATIO S I FOR ATIO
accuracy improves rapidly, roughly in proportion to 1/I reducing I
. The AC performance tends to be less affected by
OUTFS
, except for the unavoidable affects on
OUTFS
SFDR and THD due to increased INL and DNL.
Output Configurations
Based on the specific application requirements, the LTC1668 allows a choice of the best of several output configurations. Voltage outputs can be generated by ex­ternal load resistors, transformer coupling or with an op amp I-to-V converter. Single-ended DAC output configu­rations use only one of the outputs, preferably I
OUT A
, to produce a single-ended voltage output. Differential mode configurations use the difference between I I
to generate an output voltage, V
OUT B
, as shown in
DIFF
OUT A
and
equation 7. Differential mode gives much better accuracy in most AC applications. Because the DAC chip is the point of interface between the digital input signals and the analog output, some small amount of noise coupling to I
OUT A
and I
is unavoidable. Most of that digital noise
OUT B
is common mode and is canceled by the differential mode circuit. Other significant digital noise components can be modeled as V
REF
or I
noise. In single-ended mode,
OUTFS
I
noise is gone at zero scale and is fully present at full
OUTFS
scale. In differential mode, I
noise is cancelled at
OUTFS
midscale input, corresponding to zero analog output. Many AC signals, including broadband and multitone communications signals with high peak to average ratios, stay mostly near midscale.
Differential transformer-coupled output configurations usually give the best AC performance. An example is the AC Characterization Setup circuit, Figure 2. The advan­tages of transformer coupling include excellent rejection of common mode distortion and noise over a broad frequency range and convenient differential-to-single­ended conversion with isolation or level shifting. Also, as much as twice the power can be delivered to the load, and impedance matching can be accomplished by selecting the appropriate transformer turns ratio. The center tap on the primary side of the transformer is tied to ground to provide the DC current path for I distortion, the DC average of the I
OUT A
OUT A
and I
and I
OUT B
OUT B
. For low
currents must be exactly equal to avoid biasing the core. This is especially important for compact RF transformers with small cores. The circuit in Figure 2 uses a Mini-Circuits T1-1T RF transformer with a 1:1 turns ratio. The load
0.1µF
0.1µF
5V
0.1µF
V
REFOUT
R
SET
2k
I
REFIN
2.5V
REFERENCE
+
COMP1
C1
C2
0.1µF
COMP2
V
SS
–5V
LOW JITTER
CLOCK SOURCE
AGND DGND CLK DB15 DB0
0.1µF
CLK IN
PULSE GENERATOR
OUT 1 OUT 2
HP8110A DUAL
DD
16-BIT
HIGH SPEED
DAC
CLK IN
LTC1668
I
OUT A
I
OUT B
LADCOM
16
DIGITAL DATA
HP1663EA
LOGIC ANALYZER WITH
PATTERN GENERATOR
1668 F02
50
50
110
MINI-CIRCUITS
T1–1T
TO HP3589A SPECTRUM ANALYZER 50 INPUT
Figure 2. AC Characterization Setup
8
Page 9
WUUU
APPLICATIO S I FOR ATIO
resistance on I differential resistor of 50, and the 1:1 turns ratio means the output impedance from the transformer is 50. Note that the load resistors are optional, and they dissipate half of the output power. However, in lab environments or when driving long transmission lines it is very desirable to have a 50 output impedance. This could also be done with a 50 resistor at the transformer secondary, but putting the load resistors on I since it reduces the current through the transformer. At signal frequencies lower than about 1MHz, the trans­former core size required to maintain low distortion gets larger, and at some lower frequencies this becomes impractical.
A differential resistor loaded output configuration is shown in the Block Diagram. It is simple and economical, but it can drive only differential loads with impedance levels and amplitudes appropriate for the DAC outputs.
The recommended single-ended resistor loaded configu­ration is essentially the same circuit as the differential resistor loaded, case—simply use the I referred to ground. Rather than tying the unused I output to ground, it is preferred to load it with the equiva­lent R
LOAD
of I
waveform complementary to I Adding an op amp differential to single-ended converter
circuit to the differential resistor loaded output gives the circuit of Figure 10.
This circuit complements the capabilities of the trans­former-coupled application at lower frequencies, since available op amps can deliver good AC distortion perfor­mance at signal frequencies of a few MHz down to DC. The optional capacitor adds a single real pole of filtering, and helps reduce distortion by limiting the high frequency signal amplitude at the op amp inputs. The circuit swings ±1V around ground.
Figure 3 shows a simplified circuit for a single-ended output using I-to-V converter to produce a unipolar buffered voltage output. This configuration typically has the best DC linearity performance, but its AC distortion at higher frequencies is limited by U1’s slewing capabilities.
OUT A
OUT A
and I
OUT B
. Then I
is equivalent to a single
and I
OUT A
will still swing with a
OUT B
.
OUT A
is preferred
OUT B
OUT A
output,
OUT B
LTC1668
C
OUT
R
FB
I
OUTFS
10mA
I
OUT A
LTC1668
I
OUT B
LADCOM
Figure 3. Unipolar Buffered Voltage Output
200
Digital Interface
The LTC1668 has 16 parallel inputs that are latched on the rising edge of the clock input. They accept CMOS levels from either 5V or 3.3V logic and can accept clock rates of up to 50MHz.
Referring to the Timing Diagram and Block Diagram, the data inputs go to master-slave latches that update on the rising edge of the clock. The input logic thresholds, VIH =
2.4V min, VIL = 0.8V max, work with 3.3V or 5V CMOS levels over temperature. The guaranteed setup time, tDS, is 8ns minimum and the hold time, tDH, is 4ns minimum. The minimum clock high and low times are guaranteed at 6ns and 8ns, respectively. These specifications allow the LTC1668 to be clocked at up to 50Msps minimum.
For best AC performance, the data and clock waveforms need to be clean and free of undershoot and overshoot. Clock and data interconnect lines should be twisted pair, coax or microstrip, and proper line termination is impor­tant. If the digital input signals to the DAC are considered as analog AC voltage signals, they are rich in spectral components over a broad frequency range, usually in­cluding the output signal band of interest. Therefore, any direct coupling of the digital signals to the analog output will produce spurious tones that vary with the exact digital input pattern.
Clock jitter should be minimized to avoid degrading the noise floor of the device in AC applications, especially where high output frequencies are being generated. Any noise coupling from the digital inputs to the clock input will
200
U1
®
LT
1812
+
V
OUT
0V TO 2V
1668 F03
9
Page 10
LTC1668
WUUU
APPLICATIO S I FOR ATIO
cause phase modulation of the clock signal and the DAC waveform, and can produce spurious tones. It is normally best to place the digital data transitions near the falling clock edge, well away from the active rising clock edge. Because the clock signal contains spectral components only at the sampling frequency and its multiples, it is usually not a source of in band spurious tones. Overall, it is better to treat the clock as you would an analog signal and route it separately from the digital data input signals. The clock trace should be routed either over the analog ground plane or over its own section of the ground plane. The clock line needs to have accurately controlled imped­ance and should be well terminated near the LTC1668.
Printed Circuit Board Layout Considerations— Grounding, Bypassing and Output Signal Routing
The close proximity of high frequency digital data lines and high dynamic range, wide-band analog signals makes clean printed circuit board design and layout an absolute necessity. Figures 5 to 9 are the printed circuit board layers for an AC evaluation circuit for the LTC1668. Ground planes should be split between digital and analog sections as shown. All bypass capacitors should have minimum trace length and be ceramic 0.1µF or larger with low ESR.
Bypass capacitors are required on VSS, VDD and REFOUT, and all connected to the AGND plane. The COMP2 pin ties to a node in the output current switching circuitry, and it requires a 0.1µF bypass capacitor. It should be bypassed to VSS along with COMP1. The AGND and DGND pins should both tie directly to the AGND plane, and the tie point between the AGND and DGND planes should nominally be near the DGND pin. LADCOM should either be tied directly to the AGND plane or be bypassed to AGND. The I I
traces should be close together, short, and well
OUT B
matched for good AC CMRR. The transformer output ground should be capable of optionally being isolated or being tied to the AGND plane, depending on which gives better performance in the system.
Suggested Evaluation Circuit
Figure 4 is the schematic and Figures 5 to 9 are the circuit board layouts for a suggested evaluation circuit, DC245A. The circuit can be programmed with component selection and jumpers for a variety of differentially coupled trans­former output and differential and single-ended resistor loaded output configurations.
OUT A
and
10
Page 11
WUUU
APPLICATIO S I FOR ATIO
OUT
J4
V
4 T1 3
2
TP4
OUT A
J2
I
C18
0.1µF
JP2
TP3
R3
1.91k
R4
WHT
TESTPOINT
C3
0.1µF
0.1%
JP4
R5 R6
JP3
C17
0.1µF
1516
REFOUTREFIN
LTC1668-28
5V
TESTPOINT WHT C4
R7
110
TP5
20
19
OUT AIOUT B
I
DB15 (MSB)
DB14
DB13
DB12
123456789
27
28
C5
R8
6
MINI-
1
J5
JP5
C8
21
22
18
TESTPOINT WHT
COMP1
LADCOM
DB11
DB10
DB9
T1–1T
CIRCUITS
JP8
C9
OUT B
I
0.1µF
C8
0.1µF
C7
0.1µF
COMP2
DB8
DB7
JP7
JP6
0.1µF
–5V
5V
23
25
SS
V
V
DB6
DB5
C12
22pF
R10
50
0.1%
R9
50
0.1%
C12
22pF
C11
0.1µF
C10
0.1µF
17
24
DD
AGND
DGND
DB4
DB3
DB2
1011121314
1668 F04
DB1
DB0
26
CLK
TIE POINT
AGND DGND
GROUND PLANE
J6
EXTCLK
1%
R12
49.9
123
JP9
TP8
TESTPOINT RED
–5V
J9
C16
C20
C22
LTC1668
10µF
25V
+
0.1µF
0.1µF
Figure 4. Suggested Evaluation Circuit
TP10
TESTPOINT BLK
TP1
R1
10
J1
EXTREF
C1
0.1µF
R2
200
REF
TP2
2.5V
TESTPOINT WHT
246
JP1
135
6
2
LT1460DCS8-2.5
5V
C15
10µF
+5VD
16151413121110
RN5
1234567
+5VD
4
OUT
V
GND
IN
V
C2
0.1µF
13579
AMP
2468101214161820222426283032343638
102159-9
SIP
(NOT
PULL-UP/
OPTIONAL
RESISTORS
INSTALLED)
PULL-DOWN
16151413121110
9
RN6
22
1234567
8
SIP
(NOT
PULL-UP/
OPTIONAL
RESISTORS
INSTALLED)
PULL-DOWN
111315171921232527293133353739
9
22
8
TP7
TESTPOINT RED
+5VA
TP6
TESTPOINT RED
40
+5VD
25V
+
C23
C21
J8
C14
TP9
0.1µF
0.1µF
J11
10µF
TESTPOINT BLK
25V
+
C19
0.1µF
J7
J10
11
Page 12
LTC1668
WUUU
APPLICATIO S I FOR ATIO
Figure 5. Suggested Evaluation Circuit Board—Silkscreen
12
Figure 6. Suggested Evaluation Circuit Board—Component Side
Page 13
WUUU
APPLICATIO S I FOR ATIO
LTC1668
Figure 7. Suggested Evaluation Circuit Board—GND Plane
Figure 8. Suggested Evaluation Circuit Board—Power Plane
13
Page 14
LTC1668
WUUU
APPLICATIO S I FOR ATIO
Figure 9. Suggested Evaluation Circuit Board—Solder Side
14
Page 15
PACKAGE DESCRIPTIO
U
Dimensions in millimeters (inches) unless otherwise noted.
G Package
28-Lead Plastic SSOP (0.209)
(LTC DWG # 05-08-1640)
10.07 – 10.33*
(0.397 – 0.407)
2526 22 21 20 19 181716 1523242728
LTC1668
7.65 – 7.90
(0.301 – 0.311)
5.20 – 5.38** (0.205 – 0.212)
° – 8°
0
0.13 – 0.22
(0.005 – 0.009)
NOTE: DIMENSIONS ARE IN MILLIMETERS
*
DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH SHALL NOT EXCEED 0.152mm (0.006") PER SIDE
**
DIMENSIONS DO NOT INCLUDE INTERLEAD FLASH. INTERLEAD FLASH SHALL NOT EXCEED 0.254mm (0.010") PER SIDE
0.55 – 0.95
(0.022 – 0.037)
12345678 9 10 11 12 1413
0.65
(0.0256)
BSC
0.25 – 0.38
(0.010 – 0.015)
1.73 – 1.99
(0.068 – 0.078)
0.05 – 0.21
(0.002 – 0.008)
G28 SSOP 1098
15
Page 16
LTC1668
TYPICAL APPLICATIO
U
5V
0.1µF
0.1µF
0.1µF
R
0.1µF
SET
V
REFOUT
2k
I
REFIN
2.5V
REFERENCE
+
COMP1
COMP2
V
–5V
SS
AGND DGND CLK DB15 DB0
0.1µF
DD
CLOCK INPUT
16-BIT
HIGH SPEED
DAC
16-BIT DATA
INPUT
Figure 10. High Speed Buffered V
LTC1668
I
OUT A
I
OUT B
LADCOM
1668 F10
OUT
DAC
C
OPT
52.3
200
200
52.3
500
+
500
LT
®
1812
V
OUT
±1V 10dBm
RELATED PARTS
PART NUMBER DESCRIPTION COMMENTS
LTC1406 8-Bit, 20Msps ADC Undersampling Capability Up to 70MHz Input LTC1414 14-Bit, 2.2Msps ADC 84dB SFDR at 1.1MHz f LTC1420 12-Bit, 10Msps ADC 72dB SINAD at 5MHz f LTC1604 16-Bit, 333ksps ADC 16-Bit, No Missing Codes, 90dB SINAD, –100dB THD
Linear Technology Corporation
16
1630 McCarthy Blvd., Milpitas, CA 95035-7417
(408) 432-1900 ● FAX: (408) 434-0507
www.linear-tech.com
IN
IN
1668i LT/TP 0200 4K • PRINTED IN USA
LINEAR TECHNOLOGY CORPORATION 2000
Loading...