sec.5.7.tex

\label{sec:5.7}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Channel crosstalk
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

While there is no formal specification regarding the channel crosstalk, we have measured crosstalk of $<0.5\%$ in LN$_2$ and $<1\%$ at RT. We have done a variety of measurements that suggest the crosstalk is due to kickback between the channels during the sampling phase that makes it so the SHAs are not completely settled between samples. When the channels are slowed down, the crosstalk is somewhat mitigated, which indicates that it is due to insufficient bandwidth in the SHA MUX. To improve this, we will lower the impedance of the MUX to allow faster settling.