ADL5336ACPZ-R7 datasheet(23/32 Pages) AD

23 / 32 page

Data Sheet

ADL5336

Rev. B | Page 23 of 32

EFFECT OF SETPOINT ON EVM

While in AGC mode, the EVM can degrade depending on the

output setpoint each VGA is set to. There is a strong relationship

between the output setpoint of VGA2 and EVM performance

while the output setpoint of VGA1 is held constant. Conversely,

the EVM does not change much while the output setpoint of the

VGA2 is held constant and the output setpoint of VGA1 is

changed. This behavior can be seen in Figure 66 where several

different setpoints of both VGAs were tested. This example uses

a 16 QAM modulated signal at 4.5 Msym/sec using a pulse

shaping filter and an alpha of 0.35.The frequency used was 140

MHz and CAGC = 0.1 μF. Both VGAs were set to maximum gain

codes of 11.

Figure 66. EVM vs. RF Input Power over Several Setpoints

Figure 67. EVM vs. RF Input Power While VGA1 Setpoint Held Constant to

250 mV rms and VGA2 Setpoint Swept; VGA1/VGA2 Gain Code = 11

CASCADED VGA/AGC PERFORMANCE

The ADL5336 is designed for easy cascading of the two VGAs.

Cascading VGAs decreases the overall noise figure by keeping

as much gain as possible before the final gain stage/noise source. A

single X-AMP has constant output referred noise. For an 8 dB

NF amplifier, with 36 dB maximum gain, in a 200 Ω matched

system, output referred noise VN, RTO = 144 nV/√Hz. RTO, the

noise contribution from the source, is the constant source noise

multiplied by the gain (as the gain is reduced, the noise contribution

from the source decreases). Measuring noise figure as 20 × log10

(total noise/noise from source), the dB-for-dB degradation in

NF typical of this architecture can be seen.

When the gain is partitioned into two VGAs, consider 18 dB

each. If each has an 8 dB NF, then each has an RTO noise of

18 nV/√Hz, including the source noise, and 16.5 nV/√Hz,

excluding the source noise. At maximum gain, the total RTO

noise is 145 nV/√Hz. As overall gain is decreased, the gain of

VGA2 is decreased first. When the gain of VGA2 is decreased

by 6 dB, the noise contributions from the source and VGA1

both decrease by 6 dB for an overall RTO noise of the system

that falls to 74 nV/√Hz.

When VGA1 and VGA2 are cascaded and operating in AGC

mode, setpoint programming affects dynamic range. The noise

measured at the output of VGA1 is relatively constant across

gain, which is a feature common to X-AMP VGAs. However,

measured at the output of VGA2, the noise contribution from

VGA2 is constant, but the noise contribution from VGA1 depends

on the gain of VGA2. For a given overall gain (VGA1 and VGA2),

the gain partitioning between VGA1 and VGA2 controls total

RTO noise and distortion.

To illustrate, consider the case where both VGAs are programmed

to a maximum gain of 14 dB and the setpoint of VGA2 is 101,

or 353 mV rms. Gain and signal levels can also be looked at

when the setpoint of VGA1 is programmed to 011, 101, and

111, 176 mV rms, 353 mV rms, and 707 mV rms (see Table 9).

Table 9. Total Cascaded Output Noise

Vi (mV rms)

AV1 (dB)

VO1 (mV rms)

AV2 (dB)

VO (mV rms)

nTOTAL

176

353

22.4

176

353

14.1

176

707

−6

353

11.2

–50

–45

–40

–35

–30

–25

–20

–15

–10

–5

–65

–55

–45

–35

–25

–15

–5

RF INPUT POWER (dBm)

VGA1 88mV rms, VGA2 250mV rms

VGA1 250mV rms, VGA2 250mV rms

VGA1 707mV rms, VGA2 250mV rms

VGA1 250mV rms, VGA2 88mV rms

VGA1 250mV rms, VGA2 500mV rms

VGA1 250mV rms, VGA2 125mV rms

VGA1 250mV rms, VGA2 176mV rms

0.5

1.0

1.5

2.0

2.5

–50

–45

–40

–35

–30

–25

–20

–15

–10

–5

–65

–55

–45

–35

–25

–15

–5

RF INPUT POWER (dBm)

VGA2 88mV rms

VGA2 125mV rms

VGA2 176 mV rms

VGA2 250mV rms

VGA2 500mV rms

VGAIN2 250/88

VGAIN2 250/125

VGAIN2 250/176

VGAIN2 250/250

VGAIN2 250/500

VGAIN1

ADL5336ACPZ-R7 数据表(PDF) 23 Page - Analog Devices

ADL5336ACPZ-R7 数据表(HTML) 23 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

数据表下载