The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Switched tuning LC oscillators
https://designers-guide.org/forum/YaBB.pl?num=1205929447

Message started by neoflash on Mar 19^th, 2008, 5:24am

Title: Switched tuning LC oscillators
Post by neoflash on Mar 19^th, 2008, 5:24am

Normally, people use MIM or finger metal capacitor for high linearity. Anyone ever see paper use varactor or moscap as switched tuning capacitor array?

Pls suggest reference, thanks.

Title: Re: Switched tuning LC oscillators
Post by ACWWong on Mar 19^th, 2008, 6:05am

lots of people use moscap for switching.... better Q can be acheived (depending on circumstances of course) using MOS rather than a switch in series with a MIM. Nearly all papers i've seen of late for CMOS switched LC VCO use MOS switched caps.

If you google it i'm sure you'll find loads of papers, but i do like this earlier example from RFIC2000:
"A LOW PHASE NOISE MONOLITHIC VCO IN SiGe BiCMOS"
Jean-Marc Mourant, Jim Imbornone, and Ted Tewksbury*

http://ieeexplore.ieee.org/iel5/6896/18556/00854418.pdf?isnumber=18556&prod=CNF&arnumber=854418&arSt=65&ared=68&arAuthor=Mourant%2C+J.-M.%3B+Imbornone%2C+J.%3B+Tewksbury%2C+T.&htry=5

Title: Re: Switched tuning LC oscillators
Post by neoflash on Mar 19^th, 2008, 7:11am

use direct mos cap, controlled by digitally biasing of the backside might suffer big supply noise pushing.

If there is no voltage regulator for the digital control signal, the conversion from power noise to phase noise should be high.

Title: Re: Switched tuning LC oscillators
Post by neoflash on Mar 19^th, 2008, 7:14am

any people ever try put varactor in series with switch, for switched tuning?

Title: Re: Switched tuning LC oscillators
Post by jlee on Mar 23^rd, 2008, 7:05pm

Another reference: "Design of wide-band CMOS VCO for multiband wireless LAN application" by N. H. W. Fong.

For varactor/moscap based switch cap, biasing voltage noise (LDO, current biasing) is a big concern. Try PXF simulation, and you will find the flicker noise upconversion gain is normally increased by 10dB.

Actually, well designed mimcap based switch cap has little effect on phase noise and its cmax/cmin (>4) is really higher than varactor based switch cap (<3). Q is not always the criterion for phase noise performance. :D

Regards

Jason

Title: Re: Switched tuning LC oscillators
Post by neoflash on Mar 25^th, 2008, 7:29am

The binary biased varactor always make me surprised. Without good regulator, the supply noise can raise the noise floor to absurd level. However, I often see authors of Jssc using this design without regulator highlighted. Then, I just keeps being surprised.