The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> a question about linear pa.
https://designers-guide.org/forum/YaBB.pl?num=1242982455

Message started by liuqing2050 on May 22^nd, 2009, 1:54am

Title: a question about linear pa.
Post by liuqing2050 on May 22^nd, 2009, 1:54am

i am a new designner of linear pa. now, i have some problems in how to simulation the linearity of pa.
evm depicts the in-band distortion and acpr depicts the out-band distordion, what abot the am-am am-pm means? for in-band or out-band?
i find a strange problem, i design two pa. the simulation am-am and am-pm of the two pa are nearly similar. but the simulaton im3 has 10 dbc difference. why the im3 has so lage difference?

Title: Re: a question about linear pa.
Post by liuqing2050 on May 22^nd, 2009, 1:55am

thank you for your answers!

Title: Re: a question about linear pa.
Post by pancho_hideboo on May 22^nd, 2009, 5:35am

waseda-rfic wrote on May 22^nd, 2009, 1:54am:

what abot the am-am am-pm means? for in-band or out-band?

AM/AM and AM/PM affects both in-band and out-band characteristics.

waseda-rfic wrote on May 22^nd, 2009, 1:54am:

the simulation am-am and am-pm of the two pa are nearly similar.
but the simulaton im3 has 10 dbc difference. why the im3 has so lage difference?

You described too few informations, so I can't say something surely.

The following is one possiblity for your case.

Maybe AM/AM characteristics of one PA is dominated by output nonlinearity rather than input nonlinearity.
On the other hand, AM/AM characteristics of the other PA is dominated by input nonlinearity.

Title: Re: a question about linear pa.
Post by liuqing2050 on May 22^nd, 2009, 7:44pm

thank you very much for your advise. In my simulation, the AM-PM distortion is 4 degree at 16 dbm input (one stage pa) and IM3 is -10 DBc, in another PA (one stage pa) the AM-PM distortion is 4 degree at 12dbm input and IM3 below -20 DBc. the two PA output p1dbs are nearly the same. so what causes this difference?
if as you says,Maybe AM/AM characteristics of one PA is dominated by output nonlinearity rather than input nonlinearity.On the other hand, AM/AM characteristics of the other PA is dominated by input nonlinearity.
i want to ask a question. in a cmos pa, whether the nonlinearity of Cgs dominats the input nonlinearity and the nonlinearity of Rds dominats the output nonlinearity and which is more important?

Title: Re: a question about linear pa.
Post by liuqing2050 on May 23^rd, 2009, 7:40am

pancho_hideboo wrote on May 22^nd, 2009, 5:35am:

Maybe AM/AM characteristics of one PA is dominated by output nonlinearity rather than input nonlinearity.
On the other hand, AM/AM characteristics of the other PA is dominated by input nonlinearity.

you mean input nonlinearity is the nonlinearity of pa?

Title: Re: a question about linear pa.
Post by liuqing2050 on May 23^rd, 2009, 6:54pm

hi pancho_hideboo
thank s for your help!

Title: Re: a question about linear pa.
Post by liuqing2050 on May 23^rd, 2009, 7:33pm

hi pancho_hideboo
thanks for your help!
by the way, do you know transforme based power combining? my simulation IM3 is asymmetry. may be some problems in my transforme?

Title: Re: a question about linear pa.
Post by pancho_hideboo on May 23^rd, 2009, 9:30pm

waseda-rfic wrote on May 23^rd, 2009, 7:33pm:

my simulation IM3 is asymmetry. may be some problems in my transforme?

It seems you are trying to implement some IM3 canceling technique.
http://www.designers-guide.org/Forum/YaBB.pl?num=1243131271

So a frequency characteristics exists in your PA.
Asymmetrical IM3 is quite natural result when there is a frequency characteristics.

http://cp.literature.agilent.com/litweb/pdf/5989-8889EN.pdf

Title: Re: a question about linear pa.
Post by liuqing2050 on May 25^th, 2009, 1:44am

Hi pancho_hideboo
thanks!