Is it possible to design a 1.8 GHz VCO in 0.18 CMOS?

VCO

hi all!

Is it possible to design a VCO with tuning range almost 1.8 GHZ in 0.18 CMOS process ( from around 400 MHz to 2.2 GHz) ? Thanks in advance.

Re: VCO

Generally VCO can be constructed with two mechanisms:

1. Using oscillation swing to change frequency, such as Maneatis's self-bais structure. In this case if you use 3.3v power supply, you will have approximate 2.2v (70%) voltage headroom in control voltage for linear VCO gain. Then 1.8GHz/2.2v=818MHz/V. It is realizable in 0.18um process. If you use 1.8v power supply, you will have 1.26v (maybe lower due to velocity saturation) headroom. Then you get 1.4GHz/V. This is hard to realized in 0.18um.

2.Using ICO. Its frequency change with current. Although voltage headroom is limited the same as above case, you can make gm large enough to get adquate V-I converter gain. This may help you to enlarge VCO working range.

One thing to remember it is difficult to keep the same VCO gain in that large range. Gain variation will lead to instability of the PLL loop. So, the best thing is to divide the range into two segment and configure VCO to suit different segments.