let's collect some information on noise removal strategies, maybe just as a QA format first.
Please add more information and links to other resources on the topic!
GND vs. REF on the Headstage
The gnd and ref connectors on the headstage aren't the same.
Most EIBs and adapters that we've been using tie them together, but it's still good to understand the difference.
(add more info here)
Extra shielding
A simple and fast solutions to some otherwise unsolvable noise might be to add even a partial faraday cage to your recording sites. This is true even with implants that already have a complete shielding - as long as part of your animal is not shielded, it can still pick up noise.
Line Noise caused by USB ground
Sometimes, the USB cable that connects the aq board to the PC (which is grounded to building ground via the power supply) isnt good enough - grounding the external shield of the USB cable that connects the FPGA to the computer can help.
Line noise caused by different power circuits
Make sure that all equipment, including the recording computer, is plugged into the same circuit and grounded to the same ground. This is especially crucial in buildings with separate building and earth grounds.
Be careful when bridging two ground circuits - sometimes they can carry a large potential difference.
Dont use ungrounded laptops with no extra ground
When using a laptop on battery power, make sure to ground the acquisition board to the building/table ground, otherwise the recording will pick up a lot of line noise. The best way to do this is with the "GND" screw terminals at the side of the board. You can, for now, also connect the shield (but not the center) of the back-panel "clock" BNC to your favorite ground. This may not be a good idea, since we might someday do something with the clock pin.
Beware of bad power supplies
Sometimes, aftermarket power supplies, mostly for laptops, but also for LED lights, or any other power brick, can cause a lot of line noise - try individually turning off equipment to isolate the offending one.
Also, the supplied wall wart is not great at rejecting line noise - if 60Hz is a president issue, its worth trying powering the acquisition system from a high quality bench DC supply (5V) - this can significantly reduce noise in some cases.
Avoiding ground loops
add info here