an.tsouchlos/rfics-presentation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 2 Wiki Activity

Fix conversion gain -> P_IF

Browse Source
This commit is contained in:
Andreas Tsouchlos 2024-09-10 02:51:15 +02:00
parent be38796b99
commit 19bc5f09f7
4 changed files with 40 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
script.md
View File

@ -21,6 +21,8 @@
### OP: SQuad
1.
### OP: TIA
### OP: Buffer
@ -31,4 +33,8 @@
### Final circuit
- Weird behavior of LO curve appeared when lowering SQuad VBE. Lowering VBE
yielded higher conversion gain
-
## Discussion & Conclusion

2
sections/02_proposed_ideas.tex
View File

@ -100,7 +100,7 @@
\begin{itemize}
\item Usage of switching quad (SQuad) instead of conventional Gilbert cell for more voltage headroom
\item Mixer loaded by modified Cherry-Hooper \citereference{CH63} transimpedance amplifier (TIA)
\item Transmission-line based differential L-type matching networks for high bandwidth
\item Transmission line based differential L-type matching networks for high bandwidth
\item Signal fed using marchand baluns for high bandwidth
\end{itemize}
\end{minipage}%

47
sections/03_simulation_results.tex
View File

@ -310,23 +310,23 @@
\begin{frame}
\frametitle{Integration: SQuad \& TIA}
\vspace*{-6mm}
\vspace*{-4mm}
\begin{figure}
\begin{subfigure}{0.5\textwidth}
\begin{tikzpicture}
\begin{axis}[
width=\textwidth,
height=0.5\textwidth,
ylabel={Conversion Gain (dB)},
ylabel={$P_\text{IF}\ (\text{dBm})$},
xlabel={$P_\text{LO}\ (\text{dBm})$},
grid,
xtick={-50,-40,...,10},
ytick={-40,-30,...,10},
ytick={-80,-70,...,-20},
]
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=LOPow, y=ConvGain]
table[col sep=comma, x=LOPow, y expr=(\thisrowno{1}-40)]
{res/simulation/INT_SQuad_TIA_ConvGain_vs_LOPow.csv};
\addplot[mark=*] coordinates {(-5,8.969)} node[pin=-100:{OP}]{} ;
\addplot[mark=*] coordinates {(-5,-31.031)} node[pin=-100:{OP}]{} ;
\end{axis}
\end{tikzpicture}
\end{subfigure}%
@ -356,16 +356,16 @@
\begin{axis}[
width=\textwidth,
height=0.5\textwidth,
ylabel={Conversion Gain (dB)},
ylabel={$P_\text{IF}\ (\text{dBm})$},
xlabel={$P_\text{RF}\ (\text{dBm})$},
xtick={-60,-50,...,20},
ytick={-20,-10,...,30},
ytick={-60,-50,...,10},
grid,
]
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=RFPow, y=ConvGain]
table[col sep=comma, x=RFPow, y expr=(\thisrowno{1}-40)]
{res/simulation/INT_SQuad_TIA_ConvGain_vs_RFPow.csv};
\addplot[mark=*] coordinates {(-20,28.073)} node[pin=-80:{$P_{\SI{1}{dB}}$}]{} ;
\addplot[mark=*] coordinates {(-20,-11.927)} node[pin=-80:{$P_{\SI{1}{dB}}$}]{} ;
\end{axis}
\end{tikzpicture}
\end{subfigure}%
@ -416,23 +416,23 @@
\begin{frame}
\frametitle{Integration: SQuad, TIA \& Buffer}
\vspace*{-6mm}
\vspace*{-4mm}
\begin{figure}
\begin{subfigure}{0.5\textwidth}
\begin{tikzpicture}
\begin{axis}[
width=\textwidth,
height=0.5\textwidth,
ylabel={Conversion Gain (dB)},
ylabel={$P_\text{IF}\ (\text{dBm})$},
xlabel={$P_\text{LO}\ (\text{dBm})$},
grid,
xtick={-50,-40,...,10},
ytick={-20,-10,...,40},
ytick={-60,-50,...,0},
]
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=LOPow, y=ConvGain]
table[col sep=comma, x=LOPow, y expr=(\thisrowno{1}-40)]
{res/simulation/INT_Buffer_ConvGain_vs_LOPow.csv};
\addplot[mark=*] coordinates {(-5,37.774)} node[pin=-100:{OP}]{} ;
\addplot[mark=*] coordinates {(-5,-2.226)} node[pin=-100:{OP}]{} ;
\end{axis}
\end{tikzpicture}
\end{subfigure}%
@ -458,21 +458,20 @@
\end{subfigure}%
\begin{subfigure}{0.5\textwidth}
\hspace{1.5mm}
\begin{tikzpicture}
\begin{axis}[
width=\textwidth,
height=0.5\textwidth,
ylabel={Conversion Gain (dB)},
ylabel={$P_\text{IF}\ (\text{dBm})$},
xlabel={$P_\text{RF}\ (\text{dBm})$},
xtick={-60,-50,...,20},
ytick={-10,0,...,40},
ytick={-50,-40,...,10},
grid,
]
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=RFPow, y=ConvGain]
table[col sep=comma, x=RFPow, y expr=(\thisrowno{1}-40)]
{res/simulation/INT_Buffer_ConvGain_vs_RFPow.csv};
\addplot[mark=*] coordinates {(-40,37.774)} node[pin=-80:{$P_{\SI{1}{dB}}$}]{} ;
\addplot[mark=*] coordinates {(-40,-2.226)} node[pin=-80:{$P_{\SI{1}{dB}}$}]{} ;
\end{axis}
\end{tikzpicture}
\end{subfigure}%
@ -514,7 +513,7 @@
\begin{frame}
\frametitle{Final Circuit}
\vspace*{-6mm}
\vspace*{-4mm}
\begin{figure}
\begin{subfigure}{0.5\textwidth}
\begin{tikzpicture}
@ -582,20 +581,20 @@
\begin{axis}[
width=\textwidth,
height=0.5\textwidth,
ylabel={Conversion Gain (dB)},
ylabel={$P_\text{IF}\ (\text{dBm})$},
xlabel={$P_\text{RF} / P_\text{LO} \ (\text{dBm})$},
legend pos = south west,
xtick = {-90,-80,...,10},
ytick = {-60,-40,...,40},
ytick = {-100,-80,...,0},
grid,
]
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=RFPow, y=ConvGain]
table[col sep=comma, x=RFPow, y expr=(\thisrowno{1}-40)]
{res/simulation/final_ConvGain_vs_RFPow.csv};
\addlegendentry{RF}
\addplot+[mark=none, line width=1pt]
table[col sep=comma, x=LOPow, y=ConvGain]
table[col sep=comma, x=LOPow, y expr=(\thisrowno{1}-40)]
{res/simulation/final_ConvGain_vs_LOPow.csv};
\addlegendentry{LO}
\end{axis}

17
sections/05_conclusion.tex
View File

@ -12,15 +12,18 @@
\item Differential to single-ended conversion $\rightarrow$ dense chip-to-package transition
\end{itemize}
\bigskip
% \item Own simulations
% \begin{itemize}
% \item Much higher conversion gain $\leftarrow$ technology with higher $f_\text{t}$ and $f_\text{max}$, no stability considerations
% \end{itemize}
% \bigskip
\item Own simulations
\begin{itemize}
\item Better results to be expected (technology with higher $f_\text{t}$, $f_\text{max}$, stability not considered)
\item Further investigation needed to determine whether unusual LO power behavior is problematic
\item Maybe better results by using current mirrors to set operating points of buffer instead of resistors
\item Maybe better results by replacement of discrete component matching networks by transmission line based ones
\end{itemize}
\bigskip
\item Applications of this design
\begin{itemize}
\item SiGe HBT technology integrable with CMOS $\rightarrow$ scalable, suitable for mixed-signal ICs
\item Ideal for electronic beam stearing in mm-Wave applications
\item SiGe HBT technology integrable with CMOS $\rightarrow$ scalable, suitable for mixed-signal ICs
\item Ideal for electronic beam stearing in mm-Wave applications (because of small size, moderate noise figure)
\end{itemize}
\end{itemize}
\end{frame}