Digital Signal Processing Reference
In-Depth Information
Fig. 4.20
Measured EVM (OFDM, 2.4 GHz). EVM limit of 3% is also shown (
dashed line
)
power of 11.1 dBm (12.9 mW), which is lower than that of the constant supply PA—
12.4 dBm (17.4 mW). Both output power levels are lower than the target 16 dBm
(40 mW).
Figure
4.21
shows the efficiency performance for the OFDM measurement. It
reveals that the dynamic supply PA presents a higher efficiency for all the output
power levels measured. At low output power levels, a relative improvement in effi-
ciency of a factor of 2.4 (PAE ratio of 240%) is achieved. At an equal high linear
output power (less than 3% EVM), 11 dBm (12.6 mW) for example, the absolute
efficiency improvement is 3.2%. Figure
4.22
presents a plot comparing the PAE of
the dynamic and constant supply PAs in terms of EVM. It clearly shows that at 3%
EVM the dynamic supply RF PA presents a higher efficiency.
Table
4.6
summarizes the OFDM measurement results at 2.4 GHz. The target lin-
ear output power of 16 dBm (40 mW) could not be attained by neither the dynamic
supply nor the constant supply PA. We can conclude from this that the
35 dBc
IMD3 limit adopted in the design is not enough as a linearity requirement for WLAN
application. It is worth to note, however, that the envelope detection and process-
ing blocks were implemented with discrete off-the-shelf components and, hence,
they contributed with some level of distortion. This explains the lower linear output
power level reached with the dynamic supply. In its application within a modern
transceiver, the envelope information would be readily available in the baseband
circuit and the envelope processing could be implemented much more effectively in
a DSP. This would isolate a possible linearity problem to be an issue related only
to the modulator and the PA. However, in our test environment, each of the blocks
used in the envelope path contributed with some distortion level.
−
Search WWH ::
Custom Search