ADL5372

Rev. 0 | Page 13 of 24

OPTIMIZATION

The carrier feedthrough and sideband suppression performance

of the ADL5372 can be improved by using optimization

techniques.

Carrier Feedthrough Nulling

Carrier feedthrough results from minute dc offsets that occur

between each of the differential baseband inputs. In an ideal

are equal to zero, which results in no carrier feedthrough. In a real

modulator, those two quantities are nonzero; and, when mixed

with the LO, they result in a finite amount of carrier feedthrough.

The ADL5372 is designed to provide a minimal amount of carrier

feedthrough. Should even lower carrier feedthrough levels be

while the Q-channel offset is varied until a minimum carrier

feedthrough level is obtained. The Q-channel offset required to

achieve this minimum is held constant, while the offset on the

I-channel is adjusted until a new minimum is reached. Through

two iterations of this process, the carrier feedthrough can be

reduced to as low as the output noise. The ability to null is

sometimes limited by the resolution of the offset adjustment.

Figure 26 shows the relationship of carrier feedthrough vs. dc

offset as null.

–60

–88

–84

–80

–76

–72

–68

–64

–300 –240 –180 –120 –60

0

60

120

180

240

300

Figure 26. Carrier Feedthrough vs. DC Offset Voltage at 1900 MHz

Note that throughout the nulling process, the dc bias for the

baseband inputs remains at 500 mV. When no offset is applied

The same applies to the Q channel.

It is often desirable to perform a one-time carrier null calibra-

tion. This is usually performed at a single frequency. Figure 27

shows how carrier feedthrough varies with LO frequency over a

range of ±50 MHz on either side of a null at 1900 MHz.

LO FREQUENCY (MHz)

–80

–75

–70

–65

–60

–55

–50

–45

–40

–35

–30

1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950

Figure 27. Carrier Feedthrough vs. Frequency After Nulling at 1900 MHz

Sideband Suppression Optimization

Sideband suppression results from relative gain and relative

phase offsets between the I-channel and Q-channel and can

be suppressed through adjustments to those two parameters.

Figure 28 illustrates how sideband suppression is affected by

the gain and phase imbalances.

0dB

0.0125dB

0.025dB

0.05dB

0.125dB

0.25dB

0.5dB

1.25dB

2.5dB

0

–10

–20

–30

–40

–50

–60

–70

–80

–90

0.01

0.1

1

10

100

PHASE ERROR (Degrees)

Figure 28. Sideband Suppression vs. Quadrature Phase Error for

Various Quadrature Amplitude Offsets

Figure 28 underlines the fact that adjusting only one parameter

improves the sideband suppression only to a point, unless the

other parameter is also adjusted. For example, if the amplitude

offset is 0.25 dB, improving the phase imbalance better than 1°

does not yield any improvement in the sideband suppression. For

optimum sideband suppression, an iterative adjustment

between phase and amplitude is required.

The sideband suppression nulling can be performed either through

adjusting the gain for each channel or through the modification

of the phase and gain of the digital data coming from the digital

signal processor.