Hi,
Does anyone know what sort of technologies and circuit topologies are commonly used to make wideband LNAs with high IP3 such as the one below:
http://www.rflambda.com/pdf/lownoiseamplifier/RLNA00M45GA.pdf
I would assume HEMTs are involved but wasn't able to find related journal paper references on google scholar. Any such references would be highly appreciated.
Standard CMOS process can be used according to this reference http://www.seas.ucla.edu/brweb/papers/Journals/BRMar99.pdf. But the GaN process seems to be the industry favorites :)
http://www.qorvo.com/gan-for-dummies
http://www.diva-portal.org/smash/get/diva2:452317/FULLTEXT01.pdf
if you have a high voltage supply you use various kind of GaAs transistor such as HEMT,PHEMT,GAN,etc, with high linearity otherwise you most use other technology such as cmos with high linear configurations.
I think this might suit your application:
https://www.hittite.com/content/documents/data_sheet/hmc-c004.pdf
You can follow the below link:
http://www.diva-portal.org/smash/get/diva2:452317/FULLTEXT01.pdf
Hope it will help you.
Good Luck.
Sounik Kiran
Thanks, its a good introduction to PA designs but unfortunately not much info on noise.
The trouble with GaN seems to be the noise figure at low frequencies (below 1Ghz) shoots up almost exponentially. (Would be interesting if anyone knows exactly why this happens in GaN transistors)
I have recently found that the e-pHemt process can handle this bandwidth and the distributed amplifier architecture can cover this bandwidth. If anyone knows similar architectures that can cover this bandwidth it would be useful to know.
Chris,
I think looking at device models for e-pHemt and GaN in IEEE and adding noise sources could intuitively answer your question
Timothy, Thanks for replying. Yes the e-pHemt route is looking very promising. The GaN has significant problems below 1GHz which seems to be currently unaddressed in the literature.
- Badges
- Science topic