The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> HELP! about linear-in-dB for RFCMOS
https://designers-guide.org/forum/YaBB.pl?num=1158328045

Message started by nnync on Sep 15^th, 2006, 6:47am

Title: HELP! about linear-in-dB for RFCMOS
Post by nnync on Sep 15^th, 2006, 6:47am

if i want to design analog RF Varialbe gain amplifier with large dynamic range(more than 50dB )for RF chain,how to realize linear-in-dB(or close to dB linearity) with RFCMOS structure. as we know, it is easy to use biplar with its exponential characteritic. some paper mentioned using sub-threshold MOSFET or Talor approximation.which methods is doable for cmos process? or maybe someone can give me some suggestion ?Thank u.

Title: Re: HELP! about linear-in-dB for RFCMOS
Post by loose-electron on Sep 15^th, 2006, 12:20pm

50db of gain range for an RF amplifier is not too common. Most architectures may have on the order of 20 db of gain variance on the RF side, and then put lots of gain variance after the mixer down conversion.

I am sure others have different perceptions on this, but thats my experience from the cell phone world.

Do you want variable gain control (think gilbert cell multiply) or stepped (think switches) ??

Jerry

Title: Re: HELP! about linear-in-dB for RFCMOS
Post by nnync on Sep 17^th, 2006, 7:35am

thank u for ur viewpoint.yes ,as u said ,most gain variation can be realized after down mixer for tranceiver,but for transmitter,maybe we need high dynamic range for RF chain,just say,how to realize RFVGA for transmitter?

Title: Re: HELP! about linear-in-dB for RFCMOS
Post by loose-electron on Sep 17^th, 2006, 6:22pm

there are all kinds of ways to use switches and selective turning on and off of bias to do switched gain.

variable gain - think Gilbert multiply.

All of the above needs to answer to linearity and additive noise (spectral spreading issues) problems.

also if there is a RF-PA downstream, often the amplitude is fixed going in there and the variable gain happens there.

For gain calibration, you may need to implement some form of RSSI or log amp detector and use that for calibration.

just some thoughts, there are other approaches as well.

Title: Re: HELP! about linear-in-dB for RFCMOS
Post by yora on Sep 28^th, 2006, 1:24am

What's your frequency?? And how many dB for a single step?
I think all the spec is important in toppology choosing.