The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> CMOS LC tank with vs without current source
https://designers-guide.org/forum/YaBB.pl?num=1271829457

Message started by casual on Apr 20^th, 2010, 10:57pm

Title: CMOS LC tank with vs without current source
Post by casual on Apr 20^th, 2010, 10:57pm

typical cmos LC tank (pmos + nmos cross coupled) can oscillate well without current source at the bottom or at the top, the output swing can reach rail2rail without problem.

I just wonder why some designs, current source is added at the bottom, making it current limiting. Besides this, it needs a ac bypass cap for ac ground.

The only reason i could think of is to make it insensitive to vdd since current is defined?

thank you.

Title: Re: CMOS LC tank with vs without current source
Post by rfcooltools.com on Apr 22^nd, 2010, 11:36am

Typically the use of the PMOS and NMOS both (Complimentary Gm) can be the most efficient power, but up to a point. The limit is that the tank has a smaller voltage limit due to a dual constraint of PMOS in the up direction and NMOS in the down direction, with that the best phase noise is more limited than say a PMOS or NMOS only. When a NMOS or PMOS only design is made the Tank is usually connected to the supply or ground thus the VCO amplitude on the Tank can be twice or more of the since it can swing above supply or below ground in the PMOS case. But as a result the dBcPnoise/IC efficiency.

Note using a tank to ground as in the PMOS gm case can cause the parasitic diodes to turn on momontarrily and this can cause some problems, which may not be modelled properly.

Title: Re: CMOS LC tank with vs without current source
Post by love_analog on Apr 22^nd, 2010, 4:17pm

If there is no current source the tank Q is degraded. With a current source, you will mantain the tank Q

That is why you will see that people will put a inductor in series with the current source (if the current source was shunted by a cap to reduce the downconverted 2*fref noise)

Title: Re: CMOS LC tank with vs without current source
Post by casual on Apr 28^th, 2010, 6:12am

thanks for your replies.

Quote:

If there is no current source the tank Q is degraded. With a current source, you will mantain the tank Q

i think the above reply is refering to pmos/nmos only LC tank right?
How about CMOS LC tank case? does it need current source too?

Title: Re: CMOS LC tank with vs without current source
Post by love_analog on May 1^st, 2010, 8:42am

This is independent of cmos or nmos/pmos. You could make a cmos without current source and leave with the Q degradation. Your choice. Simulate it.

Title: Re: CMOS LC tank with vs without current source
Post by raja.cedt on May 4^th, 2010, 7:35am

hi casual
i think with out current source you will get very good phase noise due no 2*fo up conversion as well as more swing. Coming to supply noise if you put good supply regulator probably you would end up with less phase noise with less power.

Thanks,
Rajasekhar.

Title: Re: CMOS LC tank with vs without current source
Post by casual on May 19^th, 2010, 12:11am

The Q will degrade without current source?
I do not see any degradation on Q since Q is fixed by the inductor and other parasitic components.