Some questions about PA
时间：04-11
整理：3721RD
点击：
I am a newbie on rf pa design, only read some books&papers without any actual experience,
Here are my questions
1. Some pa use this structure:
rf input -> bandpass sigma-delta modulator -> Class S amplifier(PWM) -> bandpass filter
As above, if input is audio frequency, this method is called Class-D PA, right?
In rf, some amplifiers with above structure is still a nonlinear pa.
that means, nonconstant-envelope signal can not use this method to maintain its envolope, is it right?
(Since i have no idea about sigma-delta modulator)
2. BPSK modulation in 11b is linear modulation, right? what about CCK? More generally, all the modulations which adopted raised cosine squared filter is linear modulations, right?
3. In my opinion, the load-pull theory which shows Ropt=2*Vdd/Imax is based on the fully utility of the power transistor because it's most expensive in distributed realization. When we use CMOS or BiCMOS to realize a monolithic PA, is the 'fully utility' or 'sufficient utility' necessary?
(Maybe the reason is to reduce dc power?)
Here are my questions
1. Some pa use this structure:
rf input -> bandpass sigma-delta modulator -> Class S amplifier(PWM) -> bandpass filter
As above, if input is audio frequency, this method is called Class-D PA, right?
In rf, some amplifiers with above structure is still a nonlinear pa.
that means, nonconstant-envelope signal can not use this method to maintain its envolope, is it right?
(Since i have no idea about sigma-delta modulator)
2. BPSK modulation in 11b is linear modulation, right? what about CCK? More generally, all the modulations which adopted raised cosine squared filter is linear modulations, right?
3. In my opinion, the load-pull theory which shows Ropt=2*Vdd/Imax is based on the fully utility of the power transistor because it's most expensive in distributed realization. When we use CMOS or BiCMOS to realize a monolithic PA, is the 'fully utility' or 'sufficient utility' necessary?
(Maybe the reason is to reduce dc power?)