Analog Devices ADRV9001 User Manual page 170

UG-1828
preLutScale
This value, given as a fixed point U2.1 number, sets the scaling factor before searching the LUT. The scaling factor can be set as 1, 1.5, 2,
2.5, 3 and 3.5. This allows the user to scale the input signal magnitude in order to cover close to the full range of the LUT for better DPD
performance. If the signal input to the compander is too small, then only part of the LUT is utilized. When the input signal is small, user
could try different scaling factors to increase the signal level which might improve the DPD performance. The scaling factor can be
determined according to the dBFS of the input data peak. As an example, if the signal peak power is less than -4dBFS, the scaling factor
3.5 could be applied. As shown in Figure 149 and Figure 150, the default value of "pre-LUT Scale" is 2, which could be further changed
by user.
DPD Post Initial Calibration Parameters Configuration
The second set of DPD parameters should be configured after initial calibration. It is defined by the following data structure:
typedef struct adi_adrv9001_DpdCfg
{
uint32_t numberOfSamples;
bool outlierRemovalEnable;
uint32_t outlierRemovalThreshold;
uint32_t additionalPowerScale;
uint32_t rxTxNormalizationLowerThreshold;
uint32_t rxTxNormalizationUpperThreshold;
bool immediateLutSwitching;
bool useSpecialFrame;
bool resetLuts;
} adi_adrv9001_DpdCfg_t;
Table 74 briefly summarizes all the DPD post initial calibration parameters described in the data structure.
Table 74. DPD Post Initial Calibration Parameters
Parameter
numberOfSamples
outlierRemovalEnable
outlierRemovalThreshold
additionalPowerScale
rxTxNormalizationLowerThreshold
rxTxNormalizationUpperThreshold
Type Description
uint32_t Specifies the number of
samples to use for each
iteration of DPD
computation.
uint8_t Enables an algorithm for
(bool) removing captured data that
may render DPD
computation unstable.
uint32_t If the above parameter is
enabled, defines a threshold
(U2.30)
for removing captured data
that may render DPD
computation unstable.
uint32_t Provides an estimate of the
standard deviation of the
modem input data
magnitude to scale the data
for internal DPD
computation.
uint32_t Signal power for the lower
(U2.30) threshold for the
normalization of the
magnitude and phase of the
RX and TX data
uint32_t Signal power for the upper
(U2.30) threshold for the
normalization of the
magnitude and phase of the
RX and TX data
Rev. PrA | Page 170 of 253
Preliminary Technical Data
Min Max Default
1024 4096 4096
FALSE
0
0 2 32-1 4
0 1.0 0.0031622776602
(−25 dBFS)
0 1.0 0.031622776602 (-
15 dBFS)
Note
The maximum
value is preferred.
Currently not
supported, so
user should set
"FALSE".
Currently not
used since outlier
removal
algorithm is not
supported.