Designer’s Guide Community

New Member

Offline

Posts: 9

Switching mode RF PA
Feb 21^st, 2011, 6:52pm

Hi, everyone!
I was designing a current mode class-D PA which is a switching mode PA. I was wondering whether the input source should be a sine type or square. For swiching mode operation, I think the gate voltage of the main transistor should be a square wave。 Is it right? If it is right, how to set up the test? Is the input impedance matching network necessary for a square drive?
Thank you!

Back to top

cmcd_schematic.JPG

IP Logged

RFICDUDE

Community Fellow

Offline

Posts: 323

Re: Switching mode RF PA
Reply #1 - Feb 22^nd, 2011, 4:25am

I guess your question depends on how you intend to use the amplifier?
If you are going to design a driver stage to drive the final stage, then no you do not need to impedance match to the source. More importantly, you shouldn't try to use a 50ohm source, rather you should use something that is more representative of the circuit you intend to drive the final stage with.

The test signal should be square for testing the switching performance, but you are also going to need something like a pulse width modulated squarewave test signal to test the linearity of the Class-D amplifier (i.e. switching efficiency alone is not sufficient).

Back to top

IP Logged

jeffrey-peng

New Member

Offline

Posts: 9

Re: Switching mode RF PA
Reply #2 - Feb 22^nd, 2011, 5:38am

Hi DUDE,
I would like use this PA as a final stage and drive the athena. If I using a square source drive the PA directly without input matching, is the PAE result got from spectreRF's PSS simulation meaningful?

RFICDUDE wrote on Feb 22^nd, 2011, 4:25am:

I guess your question depends on how you intend to use the amplifier?
If you are going to design a driver stage to drive the final stage, then no you do not need to impedance match to the source. More importantly, you shouldn't try to use a 50ohm source, rather you should use something that is more representative of the circuit you intend to drive the final stage with.

The test signal should be square for testing the switching performance, but you are also going to need something like a pulse width modulated squarewave test signal to test the linearity of the Class-D amplifier (i.e. switching efficiency alone is not sufficient).

Back to top

IP Logged