Measurement Results for the Dynamic Supply CMOS RF Power Amplifie - Linear CMOS RF Power Amplifiers for Wireless Applications - page 51

gain, which is almost constant. Its flatness suggests that the dynamic supply PA
should present an extended linear range in comparison with the constant 2.5 V sup-
ply operation.
If we compare Figs. 4.10 , 4.13 , and 4.14 ,we can notice a difference between
the different power gain measurements. The S-parameter power gain is 4.6 dB at
5.2 GHz (measured with an input power of

25 dBm), whereas the single-tone
and 2-tone power gains are close to 5 dB. The explanation for this difference is
that for the single- and 2-tone measurements the instruments used for the test are a
signal generator and a spectrum analyzer, whereas for the S-parameter it is a network
analyzer. 5 Hence, in changing from one setup to the other, some measurement errors
may have been introduced.

−

In the figures that will follow, the results for the dynamic supply RF power
amplifier are always compared to those for the same RF PA operating at a
constant supply voltage of VDD = 2 . 5V.

Figure 4.15 depicts the IMD3 as a function of the output power. For low output
power levels, the dynamic supply PA generates higher distortion than with a constant

Fig. 4.15 Measured 2-tone IMD3 at 5.2 GHz. IMD3 limit of − 35 dBc is also shown ( dashed line )

5 A sweep in the input power from

10 to 10 dBm was made with the network analyzer at 5.2 GHz.
It revealed that the value of S 21 was also approximately 4.6 dB from

−

10 to 0 dBm. This value
decreased for higher input power levels as the amplifier began to compress.

−