AD9780 Datasheet, PDF (27/36 Pages) Analog Devices – Dual 12-/14-/16-Bit, LVDS Interface, 500 MSPS DACs

35
30
25
20
15
10
5
0
256
512
768
1024
DAC GAIN CODE
Figure 63. IFS vs. DAC Gain Code
DAC TRANSFER FUNCTION
Each DAC output of the AD9780/AD9781/AD9783 drives two
complementary current outputs, IOUTP and IOUTN. IOUTP provides
a near IFS when all bits are high. For example,
DAC CODE = 2^N − 1
where N = 12/14/16 bits for AD9780/AD9781/AD9783
(respectively), while IOUTN provides no current.
The current output appearing at IOUTP and IOUTN is a function of
both the input code, and IFS and can be expressed as
IOUTP = (DAC DATA/2^N) × IFS
(1)
IOUTN = ((2^N − 1) − DAC DATA)/2^N × IFS
(2)
where DAC DATA = 0 to 2^N − 1 (decimal representation).
The two current outputs typically drive a resistive load directly
or via a transformer. If dc coupling is required, IOUTP and IOUTN
should be connected to matching resistive loads (RLOAD) that are
tied to analog common (AVSS). The single-ended voltage
output appearing at the IOUTP and IOUTN pins is
VOUTP = IOUTP × RLOAD
(3)
VOUTN = IOUTN × RLOAD
(4)
Note that to achieve the maximum output compliance of 1 V at
the nominal 20 mA output current, RLOAD must be set to 50 Ω.
Also note that the full-scale value of VOUTP and VOUTN should
not exceed the specified output compliance range to maintain
specified distortion and linearity performance.
There are two distinct advantages to operating the AD9780/
AD9781/AD9783 differentially. First, differential operation
helps cancel common-mode error sources associated with IOUTP
and IOUTN, such as noise, distortion, and dc offsets. Second, the
differential code-dependent current and subsequent output
voltage (VDIFF) is twice the value of the single-ended voltage
output (VOUTP or VOUTN), providing 2× signal power to the load.
VDIFF = (IOUTP – IOUTN) × RLOAD
(5)
AD9780/AD9781/AD9783
ANALOG MODES OF OPERATION
The AD9780/AD9781/AD9783 use a proprietary quad-switch
architecture that lowers the distortion of the DAC by eliminating a
code-dependent glitch that occurs with conventional dual-switch
architectures. This architecture eliminates the code-dependent
glitches, but creates a constant glitch at a rate of 2 × fDAC. For
communications systems and other applications requiring good
frequency domain performance from the DAC, this is seldom
problematic.
The quad-switch architecture also supports two additional
modes of operation: mix mode and return-to-zero mode. The
waveforms of these two modes are shown in Figure 64. In mix
mode, the output is inverted every other half clock cycle. This
effectively chops the DAC output at the sample rate. This chop-
ping has the effect of frequency shifting the sinc roll-off from dc
to fDAC. Additionally, there is a second subtle effect on the output
spectrum. The shifted spectrum is also shaped by a second sinc
function with a first null at 2 × fDAC. The reason for this shaping
is that the data is not continuously varying at twice the clock
rate, but is simply repeated.
In return-to-zero mode, the output is set to midscale every
other half clock cycle. The output is similar to the DAC output
in normal mode except that the output pulses are half the width
and half the area. Because the output pulses have half the width,
the sinc function is scaled in frequency by two and has a first
null at 2 × fDAC. Because the area of the pulses is half that of the
pulses in normal mode, the output power is half the normal
mode output power.
INPUT DATA D₁ D₂ D₃ D₄ D₅ D₆ D₇ D₈ D₉ D₁₀
DAC CLK
QUAD-SWITCH
t
DAC OUTPUT
(f_S MIX MODE)
QUAD-SWITCH
DAC OUTPUT
t
(RETURN-TO-
ZERO MODE)
Figure 64. Mix Mode and Return-to-Zero Mode DAC Waveforms
The functions that shape the output spectrums for the three
modes of operation, normal mode, mix mode, and return-to-
zero mode, are shown in Figure 65. Switching between the
analog modes reshapes the sinc roll-off inherent at the DAC
output. This ability to change modes in the AD9780/AD9781/
AD9783 makes the parts suitable for direct IF applications. The
user can place a carrier anywhere in the first three Nyquist
zones depending on the operating mode selected. The perfor-
mance and maximum amplitude in all three Nyquist zones is
Rev. A | Page 27 of 36