140
outputs of two of the channels of the Stim Project switching board (fig. 4). This allowed
resistors of different values to be connected to determine the optimal value of
resistance to use. The experiment described in the preceding subsection was again
repeated with varying resistances; values significantly above 1MΩ had no discernable
effect, with values approaching 0.5MΩ making a significant improvement. Requiring
such small values is unfortunate however; since many customers will be using
electrodes with resistances approaching or surpassing 1MΩ, such small grounding
resistors will result in a large signal loss. Figure 5 illustrates the effect that 540kΩ
grounding resistors have on reducing latency. Here we see that the addition of a
grounding resistor to the electrode input of channel 16 (second waveform from the top)
reduces the latency by almost a second when compared to channel 18 (third waveform
from the top), which does not have a grounding resistor attached. Attaching a
grounding resistor to the amplifier output of channel 14 (first waveform from the top)
reduces the latency by over two seconds. Figure 5 also includes the waveforms of
channels 14,16 and 18 prior to filtering (the fifth through eighth waveforms from the
top), which shows the level of voltage shift created by switching states (note the change
in scale), and how that voltage has to return to relatively close to baseline before
recording can begin.
Despite the large improvement gained by the addition of grounding resistors the
latency is still approximately a second, which is still orders of magnitude above a
desirable level so another solution must be applied. Simply reducing the resistance of
grounding resistors will too greatly reduce signal strength to be considered.