2011-12-11

Personnel: SDV

Purpose: Today we are examining the performance of the amplifier and preamplifier.

Situation: Previously, Neil, Jason, and Heather had examined the circuitry and, by my recollection, had determined that we have proper transmission of signals from the Multichannel Systems headstages to the Multichannel Systems amplifier and into the National Instruments break-out box that steve modified when he was in grad school.

However, we seem to have a lot of cross-talk between channels when we record in LabView.

Our questions today are:

• 1. Can we design a connection system for the headstages so that it is easier to change channels and keep all channels either grounded or plugged in to the voltage divider test circuit?

  2. At the level of the National Instruments 6107e acquisition card, as assessed by acquiring through the Measurement and Automation Explorer tool, is there cross talk between signals?

  3. If there is no cross-talk between signals as assessed by the MAX tool, is there any cross-talk in our LabView acquisition program?

  4. If there is no cross-talk at the level of our acquisition program in LabView, is there something wrong with the output files we generate, or the code that we use to read these signals in Matlab?

To tackle the first issue, I went to Radio Shack and bought a bunch of jumper wires so we can more easily connect to the solderless breadboard. In addition, I found 2 half-DIP sockets that are double sided (pins coming out of both sides), and plugged it into the breadboard, so the NN16 connector from MCS can plug right into the breadboard, and the headstages right into the NN16 connector. I used the "Ground" lines on the breadboard for our signal ground, and used small jumpers to connect the "unused" channels to the ground lines of the breadboard. Each channnel plugs in very firmly, and there is little risk of channels becoming unplugged or touching each other. See picture here

To work quickly, I've set up the 4 channel BNC selector in a special way. Slots "A" and "B" will be plugged into the SPIKE channel and FIELD channel to be tested. Slots "C" and "D" will be used to vary the input channel across the other channels.

Testing with headstages A and B plugged in, A is mediating channels 1-8, B is 9-16.

For testing field potentials, I am using the NI-DAQmx Task "FieldChannelCheck" in MAX, with a 2Hz signal, -20dB setting on the signal generator turned OFF, and pretty high amplitude

For testing spike channels, I am using the NI-DAQmx Task "SpikeChannelCheck" in MAX, with a 2kHz signal, -20dB setting on the signal generator turned ON, pretty low amplitude

Note1: Channel 1 worked well on oscilloscope, but I didn't get any signal in LabView on ai0 (the field channel); seemed to be okay on the spike channel. I found that the connector was not plugged in well enough to the computer; unplugging and replugging the connector worked.

Conclusion: Head Stage B Channels 2 and 8 are dead.

LabView Testing Below: Head Stage Order A on Bottom B on Top