Oscillator 95
Table 4.9: EM Extracted Parameters at 60 GHz for Symmetric
Loaded Transformer Implementation
L
1
(pH) L
2
(pH) M
12
(pH) k
153.1 136.9 44.6 0.308
optimized switch size can be found. Again, a different size of 64 µm/60 nm
is adopted for switch transistors in the symmetric implementation compared
with the size of 50 µm/60 nm in the asymmetric implementation. The size dif-
ferences between two designs come from technology and topology differences,
as we will discuss in Section 4.4.3.2.
With the tuning scheme proposed in Table 4.8, there are 6 sub-bands gen-
erated by the symmetric VCO. Compared to the asymmetric VCO intro duced
in Section 4.4.2 which can provide a wider FTR due to more sub-bands and
fewer loaded s w itches thus smaller parasitic capacitance , the symmetric VCO
significantly improves the phase noise performance with highly suppressed
phase noise variation and can still achieve a wide FTR.
4.4.2.2 Simulation and Measurement Results
The designe d symmetric 60-GHz VCO is implemented in Global Foundries
65-nm CMOS 1P8M technology. EM simulation (ADS-Momentum) is used
for circuit design and verification before the fabrication.
For a fair comparison to the asymmetric 60-GHz VCO presented it Sec-
tion 4.4.2, different transformer and switch sizes are designed for the two
fabrications to a chieve the same primary inductance L
1
as well as equivalent
Q-factors Q
eq
. In this way, similar oscillation frequency as well as similar loss
introduced by loaded transformer can be e nsured. The extracted parameters
for the transformer used for the symmetric VCO are summar ized in Table
4.9, and the equivalent circuit parameters under various band selection modes
are plotted in Figure 4.24. Also note that a square shape is adopted for the
transformer for ease of switch allocation. Though an octagonal shape is theo-
retically less lossy, the effect is minimal for single-loop transformer at 60 GHz
according to simulation.
Phase Noise Analysis
Similar to the a symmetric leaded transformer, percentage of noise contribu-
tion from switches o n loaded transformer to output is simulated and plotted
in Figure 4.29 for symmetric loaded transformers at offset frequencies of 1
MHz a nd 10 MHz. Compared with asymmetric implementation, less noise
contribution is o bserved for symmetric implementation. Although similar Q
eq
values in Figure 4.16 and Figure 4.24 indicate s imilar loss introduced by loaded
transformers, a smaller k value means less noise coupling from switches, which
gives better PN performance at the penalty of smaller FTR. As s uch, one can