AD6676EBZ datasheet(30/90 Pages) AD

30 / 90 page

AD6676

Data Sheet

Rev. A | Page 30 of 90

Some applications may benefit from a reduced IDAC1FS setting

because a reduction in the PIN_0dBFS levels results in a decibel

per decibel reduction in the gain and linearity (P1dB, IIP3)

requirements of the front-end driver. This enables a lower

power RF line-up with the possibility of 3.3 V operations.

Alternatively, it can allow a greater IF AGC operation range

from the AD6676 when the previous stages output (P1dB) level

is set by its power supply setting. Carefully evaluate the trade-

off in the ac performance of the AD6676 when deciding to

operate at reduced IDAC1FS settings.

Using the MRGN Parameter to Optimize NTF

The MRGN application parameters provide an additional degree

of freedom when trying to optimize the NTF for a particular

application. This feature is particularly useful when the AD6676

operates with a low oversampling ratio where the quantization

noise contribution begins to limit the NSD performance. In such

cases, the default MRGN settings may not be adequate, resulting

in regions of the pass band (typically at the edges) where the

worst-case NSD is higher than in other regions. For these cases,

the NTF can be optimized by adjusting the Σ-Δ ADC resonator

frequencies in such a way that that result in a more optimally

distributed NSD over the entire pass band.

The MRGN_L, MRGN_U, and MRGN_IF parameters are

located in Register 0x107 through Register 0x109. MRGN_L

and MRGN_U specify the number of megahertz by which the

lower and upper edges of the target pass band are extended,

whereas MRGN_IF specifies the resonance frequency offset of

RESON1 from the center of the target pass band. The maximum

setting in these registers must be in the range of 10 MHz to

20 MHz because higher offset settings can adversely affect the

STF. The MRGN parameter is represented as an array equal to

[MRGN_L, MRGN_U, MRGN_IF].

The following example using a low oversampling ratio of 10

highlights the effects of the MRGN parameters on the NTF and

STF. In this example, the goal is to optimize the worst-case NSD

performance across a 160 MHz pass band region with FADC =

3.2 GHz and IF = 300 MHz while trying to preserve a flat STF.

Figure 84 shows the corresponding NTF performance for different

MRGN settings, and Table 8 lists the resonant frequencies of

RESON1, RESON3, and RESON3 that pertain to these settings.

Note that the default setting of [5 5 0] results in the upper half of

the pass band having the worst NSD (−141 dBFS/Hz at 380 MHz).

Symmetrical MRGN settings of [10 10 0] and [15 15 0] are shown

to highlight how the NTF varies as only the resonant frequencies

of RESON2 and RESON3 are increasingly offset symmetrically

about the IF center of 300 MHz. To improve on the default setting

of [5 5 0], an asymmetrical setting of [8 16 2] that is weighted

towards the upper half of the pass band region was found to

achieve a more distributed worst-case NSD of −145 dBFS/Hz.

Table 8. Resonator Frequencies vs. MRGN Settings

(FADC = 3.2 GHz, FIF = 300 MHz, BW = 160 MHz)

MRGN_L

MRGN_U

MRGN_IF

RESON2

(MHz)

RESON1

(MHz)

RESON3

(MHz)

233

298

365

229

299

370

227

298

373

230

306

374

–140

–155

–150

–145

220

240

260

280

300

320

340

360

380

INPUT FREQUENCY (MHz)

MRGN = [5 5 0]

MRGN = [15 15 0]

MRGN = [10 10 0]

MRGN = [8 16 2]

Figure 84. NSD Performance for MRGN Settings Shown in Table 8

Maintaining a flat STF across the pass band is also desirable

when modifying the MRGN settings. Figure 84 shows how each

of the different MRGN settings affects the STF. Note that the

asymmetrical MRGN setting of [8 16 2] results in an STF that is

slightly skewed above IF center but still maintains ±0.5 dB flatness.

0.25

–1.25

–0.75

–0.25

–1.00

–0.50

220

240

260

280

300

320

340

360

380

FREQUENCY (MHz)

MRGN = [8 16 2]

MRGN = [10 10 2]

MRGN = [15 15 2]

MRGN = [5 5 0]

Figure 85. STF for Four Different MRGN Settings

Whereas the previous example represents an extreme case, other

cases having higher oversampling ratio can also potentially benefit

from optimization. After the values of fCLK, IF, and BW have been

determined for a particular application, it may be advantageous

to explore whether a different MRGN setting yields any

improvement. It is important to note that this sort of

optimization is based on an iterative trial and error method.

However, after the MRGN setting has been determined, both

the STF and NTF remain repeatable.

AD6676EBZ 数据表(PDF) 30 Page - Analog Devices

AD6676EBZ 数据表(HTML) 30 Page - Analog Devices

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

数据表下载

链接网址

AD6676EBZ 数据表(PDF) 30 Page - Analog Devices

AD6676EBZ 数据表(HTML) 30 Page - Analog Devices

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

数据表 下载

链接网址

数据表下载