The name is absent

130

must not be allowed to drive the sensitive NSA into saturation during stimulation since it
can take several seconds for it to recover, which is almost an eternity when recording
neural signals. Additionally if there are any capacitances in the system these large
currents can leave residual charges that can likewise saturate the NSA or obscure the
neural signal. The following describes how I explored and addressed these issues.

Cross-TalkandEMlSusceptibiIity

The switching board appears to be well shielded on the side that the signal cable
connects to, providing a high degree of immunity from channel-to-channel crosstalk
when the board is connected to a load. Although the channel-to-channel cross-talk
could reach over 5% on a neighboring channel when floating (a maximum 257mV_pk was
recorded with a 5V_pk input stimulus), when the board was connected to a microarray
suspended in saline the signal on neighboring channels was not measurably higher than
that created by purely channel-to-channel conduction through the saline medium.
Likewise, this shielding also provides a high level of protection from EMI in the stimulus
cable. As shown in figure 1A, an electrode carrying a 5V_pk signal brought close to the
stimulus cable did not provide a measurable amount of interference. Figure IB shows
that bringing the electrode close to the board on the side of the cable also does not
produce a recordable amount of EMI. However, when that same electrode is brought to
the opposite side of the board it creates a signal of over SmV_pk (fig. 1C). This can cause a
problem if the cable comes into close proximity with the switching board as shown in
figure ID. As a result any user will have to take care that the switching board is used in

<<   138   139   140   141   142   143   144   >>

More intriguing information